From nobody Tue Dec 16 20:04:54 2025 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 7A993CDB47E for ; Fri, 13 Oct 2023 15:15:12 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S232334AbjJMPPL (ORCPT ); Fri, 13 Oct 2023 11:15:11 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:39880 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S232304AbjJMPPF (ORCPT ); Fri, 13 Oct 2023 11:15:05 -0400 Received: from mail-wr1-x436.google.com (mail-wr1-x436.google.com [IPv6:2a00:1450:4864:20::436]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 14D5BBE for ; Fri, 13 Oct 2023 08:15:03 -0700 (PDT) Received: by mail-wr1-x436.google.com with SMTP id ffacd0b85a97d-32d895584f1so1991020f8f.1 for ; Fri, 13 Oct 2023 08:15:03 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linaro.org; s=google; t=1697210101; x=1697814901; darn=vger.kernel.org; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:from:to:cc:subject:date :message-id:reply-to; bh=2RLmkvVPICM5LOyUR384F6DeLTrYTDxoe421oPZGgvI=; b=KzwSINusDHVd7pqbdy3wAi5pBcwK7xGyiPMLseynp4n60WZQM9S8kGN1xznv15z3YS qCPRcPNL7wuTA9Lqh0NkPdBO7D5wF1OmrLnRqou6Ogao2UZ8IKj1nbPOnxVU5rvvUVE6 8nDmsbAYr0p9QAGDPhqjVJX1+jsMDB5fkSQ1KJT5AeP68qVKs3hGi1EOdJAELbVzyQFq ladocASudvJo0heYhjY9xHOaP9yPKshUKGSOZWz1TB6rzFf88acBlJQ3m0ixaITPAvCF 2bJh71Rpf6+W1Zabh5tHFvGOdkEkUU7VwvS7kdgYD27QIIQ6yOKsHkq9ZIdUREqnUgFJ UW0Q== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1697210101; x=1697814901; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=2RLmkvVPICM5LOyUR384F6DeLTrYTDxoe421oPZGgvI=; b=pkE8HyJLKaGTXYPzKmc/cuxTzPsDtVbsiont17aKiwGAqlMenOh9zvyMfoyvk10IKF nbXKrC0Uj+yP8+R5Zi14I/SP+51diEaybam0MxNKDWFAvKNLEgSQ0CGY+m2ftibpaG8A T0RtOA4lbhXsJ+aHVv+V1iKRzPXFsyZ/D9Ye7KBYwq78UR/+agEkr1MEaWhfRI08CZ/Z Dy/5vLiafK7U7SlnhBoBkvPzz8SDDVY30GsPJn/rw27HsmuB3mFfgAX4Q+XxoGbHLu84 P+MtNB9YCgqO1hbnoTAuBCtdu7hn8UPmyjJZheD0d+gDDgXX9u9SupwV+0w6+MdR+ddG 8lVA== X-Gm-Message-State: AOJu0YyBS6AJjAwlnw0FpF3RQkhsgjvRexwpTMxfCKgfjHgYlIJtUqkh L2SbuY7cztc9wo7MlA41EWGQHA== X-Google-Smtp-Source: AGHT+IGOdxrhJB+CsfPMtSUREeCxtkr97BD8VGwdQK5h++iNIqLNVTi096PiB/Z/h0OBU0mkfz/N9A== X-Received: by 2002:adf:fc09:0:b0:31c:7ada:5e05 with SMTP id i9-20020adffc09000000b0031c7ada5e05mr21361390wrr.51.1697210101051; Fri, 13 Oct 2023 08:15:01 -0700 (PDT) Received: from vingu-book.. ([2a01:e0a:f:6020:38a9:4647:7a2a:5b13]) by smtp.gmail.com with ESMTPSA id d9-20020a5d6dc9000000b00327b5ca093dsm20814975wrz.117.2023.10.13.08.14.59 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Fri, 13 Oct 2023 08:15:00 -0700 (PDT) From: Vincent Guittot To: mingo@redhat.com, peterz@infradead.org, juri.lelli@redhat.com, dietmar.eggemann@arm.com, rostedt@goodmis.org, bsegall@google.com, mgorman@suse.de, bristot@redhat.com, vschneid@redhat.com, rafael@kernel.org, viresh.kumar@linaro.org, qyousef@layalina.io, linux-kernel@vger.kernel.org, linux-pm@vger.kernel.org Cc: lukasz.luba@arm.com, Vincent Guittot Subject: [PATCH 1/2] sched/schedutil: rework performance estimation Date: Fri, 13 Oct 2023 17:14:49 +0200 Message-Id: <20231013151450.257891-2-vincent.guittot@linaro.org> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231013151450.257891-1-vincent.guittot@linaro.org> References: <20231013151450.257891-1-vincent.guittot@linaro.org> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" The current method to take into account uclamp hints when estimating the target frequency can end into situation where the selected target frequency is finally higher than uclamp hints whereas there are no real needs. Such cases mainly happen because we are currently mixing the traditional scheduler utilization signal with the uclamp performance hints. By adding these 2 metrics, we loose an important information when it comes to select the target frequency and we have to make some assumptions which can't fit all cases. Rework the interface between the scheduler and schedutil governor in order to propagate all information down to the cpufreq governor. effective_cpu_util() interface changes and now returns the actual utilization of the CPU with 2 optional inputs: - The minimum performance for this CPU; typically the capacity to handle the deadline task and the interrupt pressure. But also uclamp_min request when available. - The maximum targeting performance for this CPU which reflects the maximum level that we would like to not exceed. By default it will be the CPU capacity but can be reduced because of some performance hints set with uclamp. The value can be lower than actual utilization and/or min performance level. A new sugov_effective_cpu_perf() interface is also available to compute the final performance level that is targeted for the CPU after applying some cpufreq headroom and taking into account all inputs. With these 2 functions, schedutil is now able to decide when it must go above uclamp hints. It now also have a generic way to get the min perfromance level. The dependency between energy model and cpufreq governor and its headroom policy doesn't exist anymore. eenv_pd_max_util asks schedutil for the targeted performance after applying the impact of the waking task. Signed-off-by: Vincent Guittot --- include/linux/energy_model.h | 1 - kernel/sched/core.c | 85 ++++++++++++-------------------- kernel/sched/cpufreq_schedutil.c | 43 ++++++++++++---- kernel/sched/fair.c | 22 +++++++-- kernel/sched/sched.h | 24 +++------ 5 files changed, 91 insertions(+), 84 deletions(-) diff --git a/include/linux/energy_model.h b/include/linux/energy_model.h index b9caa01dfac4..adec808b371a 100644 --- a/include/linux/energy_model.h +++ b/include/linux/energy_model.h @@ -243,7 +243,6 @@ static inline unsigned long em_cpu_energy(struct em_per= f_domain *pd, scale_cpu =3D arch_scale_cpu_capacity(cpu); ps =3D &pd->table[pd->nr_perf_states - 1]; =20 - max_util =3D map_util_perf(max_util); max_util =3D min(max_util, allowed_cpu_cap); freq =3D map_util_freq(max_util, ps->frequency, scale_cpu); =20 diff --git a/kernel/sched/core.c b/kernel/sched/core.c index a3f9cd52eec5..78228abd1219 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -7381,18 +7381,13 @@ int sched_core_idle_cpu(int cpu) * required to meet deadlines. */ unsigned long effective_cpu_util(int cpu, unsigned long util_cfs, - enum cpu_util_type type, - struct task_struct *p) + unsigned long *min, + unsigned long *max) { - unsigned long dl_util, util, irq, max; + unsigned long util, irq, scale; struct rq *rq =3D cpu_rq(cpu); =20 - max =3D arch_scale_cpu_capacity(cpu); - - if (!uclamp_is_used() && - type =3D=3D FREQUENCY_UTIL && rt_rq_is_runnable(&rq->rt)) { - return max; - } + scale =3D arch_scale_cpu_capacity(cpu); =20 /* * Early check to see if IRQ/steal time saturates the CPU, can be @@ -7400,45 +7395,36 @@ unsigned long effective_cpu_util(int cpu, unsigned = long util_cfs, * update_irq_load_avg(). */ irq =3D cpu_util_irq(rq); - if (unlikely(irq >=3D max)) - return max; + if (unlikely(irq >=3D scale)) { + if (min) + *min =3D scale; + if (max) + *max =3D scale; + return scale; + } + + /* The minimum utilization returns the highest level between: + * - the computed DL bandwidth needed with the irq pressure which + * steals time to the deadline task. + * - The minimum bandwidth requirement for CFS. + */ + if (min) + *min =3D max(irq + cpu_bw_dl(rq), uclamp_rq_get(rq, UCLAMP_MIN)); =20 /* * Because the time spend on RT/DL tasks is visible as 'lost' time to * CFS tasks and we use the same metric to track the effective * utilization (PELT windows are synchronized) we can directly add them * to obtain the CPU's actual utilization. - * - * CFS and RT utilization can be boosted or capped, depending on - * utilization clamp constraints requested by currently RUNNABLE - * tasks. - * When there are no CFS RUNNABLE tasks, clamps are released and - * frequency will be gracefully reduced with the utilization decay. */ util =3D util_cfs + cpu_util_rt(rq); - if (type =3D=3D FREQUENCY_UTIL) - util =3D uclamp_rq_util_with(rq, util, p); - - dl_util =3D cpu_util_dl(rq); - - /* - * For frequency selection we do not make cpu_util_dl() a permanent part - * of this sum because we want to use cpu_bw_dl() later on, but we need - * to check if the CFS+RT+DL sum is saturated (ie. no idle time) such - * that we select f_max when there is no idle time. - * - * NOTE: numerical errors or stop class might cause us to not quite hit - * saturation when we should -- something for later. - */ - if (util + dl_util >=3D max) - return max; + util +=3D cpu_util_dl(rq); =20 - /* - * OTOH, for energy computation we need the estimated running time, so - * include util_dl and ignore dl_bw. - */ - if (type =3D=3D ENERGY_UTIL) - util +=3D dl_util; + if (util >=3D scale) { + if (max) + *max =3D scale; + return scale; + } =20 /* * There is still idle time; further improve the number by using the @@ -7449,28 +7435,21 @@ unsigned long effective_cpu_util(int cpu, unsigned = long util_cfs, * U' =3D irq + --------- * U * max */ - util =3D scale_irq_capacity(util, irq, max); + util =3D scale_irq_capacity(util, irq, scale); util +=3D irq; =20 - /* - * Bandwidth required by DEADLINE must always be granted while, for - * FAIR and RT, we use blocked utilization of IDLE CPUs as a mechanism - * to gracefully reduce the frequency when no tasks show up for longer - * periods of time. - * - * Ideally we would like to set bw_dl as min/guaranteed freq and util + - * bw_dl as requested freq. However, cpufreq is not yet ready for such - * an interface. So, we only do the latter for now. + /* The maximum hint is a soft bandwidth requirement which can be lower + * than the actual utilization because of max uclamp requirments */ - if (type =3D=3D FREQUENCY_UTIL) - util +=3D cpu_bw_dl(rq); + if (max) + *max =3D min(scale, uclamp_rq_get(rq, UCLAMP_MAX)); =20 - return min(max, util); + return min(scale, util); } =20 unsigned long sched_cpu_util(int cpu) { - return effective_cpu_util(cpu, cpu_util_cfs(cpu), ENERGY_UTIL, NULL); + return effective_cpu_util(cpu, cpu_util_cfs(cpu), NULL, NULL); } #endif /* CONFIG_SMP */ =20 diff --git a/kernel/sched/cpufreq_schedutil.c b/kernel/sched/cpufreq_schedu= til.c index 458d359f5991..8cb323522b90 100644 --- a/kernel/sched/cpufreq_schedutil.c +++ b/kernel/sched/cpufreq_schedutil.c @@ -47,7 +47,7 @@ struct sugov_cpu { u64 last_update; =20 unsigned long util; - unsigned long bw_dl; + unsigned long bw_min; =20 /* The field below is for single-CPU policies only: */ #ifdef CONFIG_NO_HZ_COMMON @@ -143,7 +143,6 @@ static unsigned int get_next_freq(struct sugov_policy *= sg_policy, unsigned int freq =3D arch_scale_freq_invariant() ? policy->cpuinfo.max_freq : policy->cur; =20 - util =3D map_util_perf(util); freq =3D map_util_freq(util, freq, max); =20 if (freq =3D=3D sg_policy->cached_raw_freq && !sg_policy->need_freq_updat= e) @@ -153,14 +152,38 @@ static unsigned int get_next_freq(struct sugov_policy= *sg_policy, return cpufreq_driver_resolve_freq(policy, freq); } =20 +unsigned long sugov_effective_cpu_perf(int cpu, unsigned long actual, + unsigned long min, + unsigned long max) +{ + unsigned long target; + struct rq *rq =3D cpu_rq(cpu); + + if (rt_rq_is_runnable(&rq->rt)) + return max; + + /* Provide at least enough capacity for DL + irq */ + target =3D min; + + actual =3D map_util_perf(actual); + /* Actually we don't need to target the max performance */ + if (actual < max) + max =3D actual; + + /* + * Ensure at least minimum performance while providing more compute + * capacity when possible. + */ + return max(target, max); +} + static void sugov_get_util(struct sugov_cpu *sg_cpu) { - unsigned long util =3D cpu_util_cfs_boost(sg_cpu->cpu); - struct rq *rq =3D cpu_rq(sg_cpu->cpu); + unsigned long min, max, util =3D cpu_util_cfs_boost(sg_cpu->cpu); =20 - sg_cpu->bw_dl =3D cpu_bw_dl(rq); - sg_cpu->util =3D effective_cpu_util(sg_cpu->cpu, util, - FREQUENCY_UTIL, NULL); + util =3D effective_cpu_util(sg_cpu->cpu, util, &min, &max); + sg_cpu->bw_min =3D map_util_perf(min); + sg_cpu->util =3D sugov_effective_cpu_perf(sg_cpu->cpu, util, min, max); } =20 /** @@ -306,7 +329,7 @@ static inline bool sugov_cpu_is_busy(struct sugov_cpu *= sg_cpu) { return false; } */ static inline void ignore_dl_rate_limit(struct sugov_cpu *sg_cpu) { - if (cpu_bw_dl(cpu_rq(sg_cpu->cpu)) > sg_cpu->bw_dl) + if (cpu_bw_dl(cpu_rq(sg_cpu->cpu)) > sg_cpu->bw_min) sg_cpu->sg_policy->limits_changed =3D true; } =20 @@ -407,8 +430,8 @@ static void sugov_update_single_perf(struct update_util= _data *hook, u64 time, sugov_cpu_is_busy(sg_cpu) && sg_cpu->util < prev_util) sg_cpu->util =3D prev_util; =20 - cpufreq_driver_adjust_perf(sg_cpu->cpu, map_util_perf(sg_cpu->bw_dl), - map_util_perf(sg_cpu->util), max_cap); + cpufreq_driver_adjust_perf(sg_cpu->cpu, sg_cpu->bw_min, + sg_cpu->util, max_cap); =20 sg_cpu->sg_policy->last_freq_update_time =3D time; } diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index 922905194c0c..d4f7b2f49c44 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -7628,7 +7628,7 @@ static inline void eenv_pd_busy_time(struct energy_en= v *eenv, for_each_cpu(cpu, pd_cpus) { unsigned long util =3D cpu_util(cpu, p, -1, 0); =20 - busy_time +=3D effective_cpu_util(cpu, util, ENERGY_UTIL, NULL); + busy_time +=3D effective_cpu_util(cpu, util, NULL, NULL); } =20 eenv->pd_busy_time =3D min(eenv->pd_cap, busy_time); @@ -7651,7 +7651,7 @@ eenv_pd_max_util(struct energy_env *eenv, struct cpum= ask *pd_cpus, for_each_cpu(cpu, pd_cpus) { struct task_struct *tsk =3D (cpu =3D=3D dst_cpu) ? p : NULL; unsigned long util =3D cpu_util(cpu, p, dst_cpu, 1); - unsigned long eff_util; + unsigned long eff_util, min, max; =20 /* * Performance domain frequency: utilization clamping @@ -7660,7 +7660,23 @@ eenv_pd_max_util(struct energy_env *eenv, struct cpu= mask *pd_cpus, * NOTE: in case RT tasks are running, by default the * FREQUENCY_UTIL's utilization can be max OPP. */ - eff_util =3D effective_cpu_util(cpu, util, FREQUENCY_UTIL, tsk); + eff_util =3D effective_cpu_util(cpu, util, &min, &max); + + /* Task's uclamp can modify min and max value */ + if (tsk && uclamp_is_used()) { + min =3D max(min, uclamp_eff_value(p, UCLAMP_MIN)); + + /* + * If there is no active max uclamp constraint, + * directly use task's one otherwise keep max + */ + if (uclamp_rq_is_idle(cpu_rq(cpu))) + max =3D uclamp_eff_value(p, UCLAMP_MAX); + else + max =3D max(max, uclamp_eff_value(p, UCLAMP_MAX)); + } + + eff_util =3D sugov_effective_cpu_perf(cpu, eff_util, min, max); max_util =3D max(max_util, eff_util); } =20 diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h index 65cad0e5729e..3873b4de7cfa 100644 --- a/kernel/sched/sched.h +++ b/kernel/sched/sched.h @@ -2962,24 +2962,14 @@ static inline void cpufreq_update_util(struct rq *r= q, unsigned int flags) {} #endif =20 #ifdef CONFIG_SMP -/** - * enum cpu_util_type - CPU utilization type - * @FREQUENCY_UTIL: Utilization used to select frequency - * @ENERGY_UTIL: Utilization used during energy calculation - * - * The utilization signals of all scheduling classes (CFS/RT/DL) and IRQ t= ime - * need to be aggregated differently depending on the usage made of them. = This - * enum is used within effective_cpu_util() to differentiate the types of - * utilization expected by the callers, and adjust the aggregation accordi= ngly. - */ -enum cpu_util_type { - FREQUENCY_UTIL, - ENERGY_UTIL, -}; - unsigned long effective_cpu_util(int cpu, unsigned long util_cfs, - enum cpu_util_type type, - struct task_struct *p); + unsigned long *min, + unsigned long *max); + +unsigned long sugov_effective_cpu_perf(int cpu, unsigned long actual, + unsigned long min, + unsigned long max); + =20 /* * Verify the fitness of task @p to run on @cpu taking into account the --=20 2.34.1 From nobody Tue Dec 16 20:04:54 2025 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 96542CDB483 for ; Fri, 13 Oct 2023 15:15:09 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S232321AbjJMPPI (ORCPT ); Fri, 13 Oct 2023 11:15:08 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:39890 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S232281AbjJMPPF (ORCPT ); Fri, 13 Oct 2023 11:15:05 -0400 Received: from mail-wr1-x433.google.com (mail-wr1-x433.google.com [IPv6:2a00:1450:4864:20::433]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 0BB3D95 for ; Fri, 13 Oct 2023 08:15:04 -0700 (PDT) Received: by mail-wr1-x433.google.com with SMTP id ffacd0b85a97d-32d849cc152so2100954f8f.1 for ; Fri, 13 Oct 2023 08:15:03 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linaro.org; s=google; t=1697210102; x=1697814902; darn=vger.kernel.org; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:from:to:cc:subject:date :message-id:reply-to; bh=wyxGVZwTo+k49X7yswwdJnlfM7vyBM0WpBEkY1/lOJA=; b=G1NwDz0qQqbyqZ7aJ0VbOkS3c8yrJA91/nVu6A8vKUbr0LrDh4pGOegvalbSwQaJKp Et6nGuvHYdC7cjJChjV8BEaheCsmZbmztaUBGF+bpbtZFswpKsh9vJQ1V+e/4d4qlREU 2IWKBsNDZfOhYei911Xi3FkYcce1Jg5oS4OeIu7vEFf13v8p1Wj6Ne9bQRpNpwmsd+OO BYsHD2IGvII+K7PqiTPROvsS3iOvbyo0l9i6XMhlqA59Fu0ugygYMT+s+prlT3G4VtJw uOcgL+oDyIxIsFW0HDfUHZ6s85G4hpyfor7jlJBgp+UkA7K5af8EXJcAhATJycRBneZ7 yPOQ== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1697210102; x=1697814902; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=wyxGVZwTo+k49X7yswwdJnlfM7vyBM0WpBEkY1/lOJA=; b=uGdURPquFCI3anYshGj24N4stycqaeOk1hbnV2bXl2hPOQhZZri9QvfNMueSXOOTgZ Eor06kHDmFx8SLkhwxH1r9w23NKrIB332FTv1ueCDAeRfWUf+N1G1zB5/Ew4UYyYUVX3 GmD6F/I4rJcXtbTwjXfjjnSbdjhRTTe2cJitdiiKjemsVoRWCcjkr0/ESityC+Y/iao2 TzZGSpyAY/0vVcoFrrFrA95TrFMJLMLjp8JD5myW7dustYWdSHeksJ8mAV+4MJTZ1Gsw kIUr8++WX5xbRkinH+iMnCtuBvED+ymgKSByuEmpfBLQnwPU93i2CDJDtqYKBwsdTwxZ 5ATA== X-Gm-Message-State: AOJu0YyST1/MlUws/FRkmqxbPKzrDWDTqqyWkqgRtUD2gVn1iMGjGprY 2pt0GkgH+uTMYgyZZq140OIu1Q== X-Google-Smtp-Source: AGHT+IEtyHxKSYFD0gN9kXyRQUa5Ao8tAG9HXvKt/zy5aeSFsJOaTvPl5MiF6PRi8HLwud6CCtfeFQ== X-Received: by 2002:a05:6000:14e:b0:320:a19:7f87 with SMTP id r14-20020a056000014e00b003200a197f87mr23475295wrx.18.1697210102347; Fri, 13 Oct 2023 08:15:02 -0700 (PDT) Received: from vingu-book.. ([2a01:e0a:f:6020:38a9:4647:7a2a:5b13]) by smtp.gmail.com with ESMTPSA id d9-20020a5d6dc9000000b00327b5ca093dsm20814975wrz.117.2023.10.13.08.15.01 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Fri, 13 Oct 2023 08:15:01 -0700 (PDT) From: Vincent Guittot To: mingo@redhat.com, peterz@infradead.org, juri.lelli@redhat.com, dietmar.eggemann@arm.com, rostedt@goodmis.org, bsegall@google.com, mgorman@suse.de, bristot@redhat.com, vschneid@redhat.com, rafael@kernel.org, viresh.kumar@linaro.org, qyousef@layalina.io, linux-kernel@vger.kernel.org, linux-pm@vger.kernel.org Cc: lukasz.luba@arm.com, Vincent Guittot Subject: [PATCH 2/2] sched/schedutil: rework iowait boost Date: Fri, 13 Oct 2023 17:14:50 +0200 Message-Id: <20231013151450.257891-3-vincent.guittot@linaro.org> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231013151450.257891-1-vincent.guittot@linaro.org> References: <20231013151450.257891-1-vincent.guittot@linaro.org> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" Use the max value that has already been computed inside sugov_get_util() to cap the iowait boost and remove dependency with uclamp_rq_util_with() which is not used anymore. Signed-off-by: Vincent Guittot --- kernel/sched/cpufreq_schedutil.c | 29 ++++++++------- kernel/sched/sched.h | 60 -------------------------------- 2 files changed, 14 insertions(+), 75 deletions(-) diff --git a/kernel/sched/cpufreq_schedutil.c b/kernel/sched/cpufreq_schedu= til.c index 8cb323522b90..820612867769 100644 --- a/kernel/sched/cpufreq_schedutil.c +++ b/kernel/sched/cpufreq_schedutil.c @@ -177,11 +177,12 @@ unsigned long sugov_effective_cpu_perf(int cpu, unsig= ned long actual, return max(target, max); } =20 -static void sugov_get_util(struct sugov_cpu *sg_cpu) +static void sugov_get_util(struct sugov_cpu *sg_cpu, unsigned long boost) { unsigned long min, max, util =3D cpu_util_cfs_boost(sg_cpu->cpu); =20 util =3D effective_cpu_util(sg_cpu->cpu, util, &min, &max); + util =3D max(util, boost); sg_cpu->bw_min =3D map_util_perf(min); sg_cpu->util =3D sugov_effective_cpu_perf(sg_cpu->cpu, util, min, max); } @@ -274,18 +275,16 @@ static void sugov_iowait_boost(struct sugov_cpu *sg_c= pu, u64 time, * This mechanism is designed to boost high frequently IO waiting tasks, w= hile * being more conservative on tasks which does sporadic IO operations. */ -static void sugov_iowait_apply(struct sugov_cpu *sg_cpu, u64 time, +static unsigned long sugov_iowait_apply(struct sugov_cpu *sg_cpu, u64 time, unsigned long max_cap) { - unsigned long boost; - /* No boost currently required */ if (!sg_cpu->iowait_boost) - return; + return 0; =20 /* Reset boost if the CPU appears to have been idle enough */ if (sugov_iowait_reset(sg_cpu, time, false)) - return; + return 0; =20 if (!sg_cpu->iowait_boost_pending) { /* @@ -294,7 +293,7 @@ static void sugov_iowait_apply(struct sugov_cpu *sg_cpu= , u64 time, sg_cpu->iowait_boost >>=3D 1; if (sg_cpu->iowait_boost < IOWAIT_BOOST_MIN) { sg_cpu->iowait_boost =3D 0; - return; + return 0; } } =20 @@ -304,10 +303,7 @@ static void sugov_iowait_apply(struct sugov_cpu *sg_cp= u, u64 time, * sg_cpu->util is already in capacity scale; convert iowait_boost * into the same scale so we can compare. */ - boost =3D (sg_cpu->iowait_boost * max_cap) >> SCHED_CAPACITY_SHIFT; - boost =3D uclamp_rq_util_with(cpu_rq(sg_cpu->cpu), boost, NULL); - if (sg_cpu->util < boost) - sg_cpu->util =3D boost; + return (sg_cpu->iowait_boost * max_cap) >> SCHED_CAPACITY_SHIFT; } =20 #ifdef CONFIG_NO_HZ_COMMON @@ -337,6 +333,8 @@ static inline bool sugov_update_single_common(struct su= gov_cpu *sg_cpu, u64 time, unsigned long max_cap, unsigned int flags) { + unsigned long boost; + sugov_iowait_boost(sg_cpu, time, flags); sg_cpu->last_update =3D time; =20 @@ -345,8 +343,8 @@ static inline bool sugov_update_single_common(struct su= gov_cpu *sg_cpu, if (!sugov_should_update_freq(sg_cpu->sg_policy, time)) return false; =20 - sugov_get_util(sg_cpu); - sugov_iowait_apply(sg_cpu, time, max_cap); + boost =3D sugov_iowait_apply(sg_cpu, time, max_cap); + sugov_get_util(sg_cpu, boost); =20 return true; } @@ -447,9 +445,10 @@ static unsigned int sugov_next_freq_shared(struct sugo= v_cpu *sg_cpu, u64 time) =20 for_each_cpu(j, policy->cpus) { struct sugov_cpu *j_sg_cpu =3D &per_cpu(sugov_cpu, j); + unsigned long boost; =20 - sugov_get_util(j_sg_cpu); - sugov_iowait_apply(j_sg_cpu, time, max_cap); + boost =3D sugov_iowait_apply(j_sg_cpu, time, max_cap); + sugov_get_util(j_sg_cpu, boost); =20 util =3D max(j_sg_cpu->util, util); } diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h index 3873b4de7cfa..b181edaf4d41 100644 --- a/kernel/sched/sched.h +++ b/kernel/sched/sched.h @@ -3026,59 +3026,6 @@ static inline bool uclamp_rq_is_idle(struct rq *rq) return rq->uclamp_flags & UCLAMP_FLAG_IDLE; } =20 -/** - * uclamp_rq_util_with - clamp @util with @rq and @p effective uclamp valu= es. - * @rq: The rq to clamp against. Must not be NULL. - * @util: The util value to clamp. - * @p: The task to clamp against. Can be NULL if you want to clamp - * against @rq only. - * - * Clamps the passed @util to the max(@rq, @p) effective uclamp values. - * - * If sched_uclamp_used static key is disabled, then just return the util - * without any clamping since uclamp aggregation at the rq level in the fa= st - * path is disabled, rendering this operation a NOP. - * - * Use uclamp_eff_value() if you don't care about uclamp values at rq leve= l. It - * will return the correct effective uclamp value of the task even if the - * static key is disabled. - */ -static __always_inline -unsigned long uclamp_rq_util_with(struct rq *rq, unsigned long util, - struct task_struct *p) -{ - unsigned long min_util =3D 0; - unsigned long max_util =3D 0; - - if (!static_branch_likely(&sched_uclamp_used)) - return util; - - if (p) { - min_util =3D uclamp_eff_value(p, UCLAMP_MIN); - max_util =3D uclamp_eff_value(p, UCLAMP_MAX); - - /* - * Ignore last runnable task's max clamp, as this task will - * reset it. Similarly, no need to read the rq's min clamp. - */ - if (uclamp_rq_is_idle(rq)) - goto out; - } - - min_util =3D max_t(unsigned long, min_util, uclamp_rq_get(rq, UCLAMP_MIN)= ); - max_util =3D max_t(unsigned long, max_util, uclamp_rq_get(rq, UCLAMP_MAX)= ); -out: - /* - * Since CPU's {min,max}_util clamps are MAX aggregated considering - * RUNNABLE tasks with _different_ clamps, we can end up with an - * inversion. Fix it now when the clamps are applied. - */ - if (unlikely(min_util >=3D max_util)) - return min_util; - - return clamp(util, min_util, max_util); -} - /* Is the rq being capped/throttled by uclamp_max? */ static inline bool uclamp_rq_is_capped(struct rq *rq) { @@ -3116,13 +3063,6 @@ static inline unsigned long uclamp_eff_value(struct = task_struct *p, return SCHED_CAPACITY_SCALE; } =20 -static inline -unsigned long uclamp_rq_util_with(struct rq *rq, unsigned long util, - struct task_struct *p) -{ - return util; -} - static inline bool uclamp_rq_is_capped(struct rq *rq) { return false; } =20 static inline bool uclamp_is_used(void) --=20 2.34.1