From patchwork Fri Jan 13 13:40:56 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Vincent Guittot X-Patchwork-Id: 642880 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 A2641C54EBD for ; Fri, 13 Jan 2023 13:47:42 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S232777AbjAMNrk (ORCPT ); Fri, 13 Jan 2023 08:47:40 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:48910 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S241753AbjAMNrO (ORCPT ); Fri, 13 Jan 2023 08:47:14 -0500 Received: from mail-wr1-x429.google.com (mail-wr1-x429.google.com [IPv6:2a00:1450:4864:20::429]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 122F15D405 for ; Fri, 13 Jan 2023 05:41:01 -0800 (PST) Received: by mail-wr1-x429.google.com with SMTP id r2so21103536wrv.7 for ; Fri, 13 Jan 2023 05:41:00 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linaro.org; s=google; h=content-transfer-encoding:mime-version:message-id:date:subject:cc :to:from:from:to:cc:subject:date:message-id:reply-to; bh=to8C6Bev+T1XukXaLfCpj5N27WyqBL3BbOApapHPhM4=; b=aalcNYKlp8WuG/reKf77olUau63Gzba5biFTUHqYq4rfPg8JbvsOuj1uewF3tZuGaq 57N3Y+um6bIbfyBeLpxBZH7d22BtoAPBNHp5dEkD2OgEeIu1tPjTOUvHeddYmWJH1vcJ RyuPn0pMKO4di9gDltpytCmsr4w4CKin/Abjsj2PO+Y/q6Lw6sf7HdVRpCKYEovK4rk3 8jVn7GQjQoGajj+fmNLKlB+1fSSl7E1VFAgaABqYT33TDqUZeR4Zy4Pi0NZ/KCw6vaQf RK9U3q5vPmAHbXQMuyEVf+4YFSnwmQrGRoFJA7Gv+XgmrxUAjuAauHl2W29KHmYQeQM1 Aagg== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=content-transfer-encoding:mime-version:message-id:date:subject:cc :to:from:x-gm-message-state:from:to:cc:subject:date:message-id :reply-to; bh=to8C6Bev+T1XukXaLfCpj5N27WyqBL3BbOApapHPhM4=; b=F5S/G4T5ejARKWULWaLjdm1HVxYOUJ6pZXLthUmpBWKgwaSlwQrmfYMRt3/hjb3P8p wTF0yrbmKWh3L7DlIlS5RPuqP9optp+tcjLLqtBHiuSCvw/EXbrt5Ezx6+d8+tqvJmZh u/V0GVIoYtTFZRhR3GchKtKwLYUVwt3P0q65YjQWzlojoRFeA6G58xgDFQFRFL0s2hdp AhN8nczxbIPqMyL+3NooMryVZJPUnPekQ5WGjFpfLzfbzazo052SodbfIp3GZABfbyNX qjVR4yCXP+5wPudTR8jKg/Ik/kTBlVEJMheMj79qiaRdQzQaPyBf1IuNoU+PPweq5QpQ VFpA== X-Gm-Message-State: AFqh2ko74Ktwa2/2ekLKuRVGMJDyDr/dyD5HHzBeCgDzyDFzQLbIe4rw Vvz/fCLvtHsv47qtIVr669+ACg== X-Google-Smtp-Source: AMrXdXsZfvkZpxlEymlnMv1aal6S7GbjoACMXVtjWz+lBia188KCa1/Teg/KrcFO/k3PLvBeTvghfA== X-Received: by 2002:adf:a497:0:b0:2b2:35:e0a4 with SMTP id g23-20020adfa497000000b002b20035e0a4mr21248053wrb.15.1673617259540; Fri, 13 Jan 2023 05:40:59 -0800 (PST) Received: from vingu-book.. ([2a01:e0a:f:6020:1563:65bf:c344:661e]) by smtp.gmail.com with ESMTPSA id q4-20020adfdfc4000000b002bc6c180738sm14586348wrn.90.2023.01.13.05.40.58 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Fri, 13 Jan 2023 05:40:58 -0800 (PST) From: Vincent Guittot To: mingo@kernel.org, peterz@infradead.org, dietmar.eggemann@arm.com, qyousef@layalina.io, rafael@kernel.org, viresh.kumar@linaro.org, vschneid@redhat.com, linux-pm@vger.kernel.org, linux-kernel@vger.kernel.org Cc: lukasz.luba@arm.com, wvw@google.com, xuewen.yan94@gmail.com, han.lin@mediatek.com, Jonathan.JMChen@mediatek.com, Vincent Guittot Subject: [PATCH v3] sched/fair: unlink misfit task from cpu overutilized Date: Fri, 13 Jan 2023 14:40:56 +0100 Message-Id: <20230113134056.257691-1-vincent.guittot@linaro.org> X-Mailer: git-send-email 2.34.1 MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: linux-pm@vger.kernel.org By taking into account uclamp_min, the 1:1 relation between task misfit and cpu overutilized is no more true as a task with a small util_avg of may not fit a high capacity cpu because of uclamp_min constraint. Add a new state in util_fits_cpu() to reflect the case that task would fit a CPU except for the uclamp_min hint which is a performance requirement. Use -1 to reflect that a CPU doesn't fit only because of uclamp_min so we can use this new value to take additional action to select the best CPU that doesn't match uclamp_min hint. Signed-off-by: Vincent Guittot --- Change since v2: - fix a condition in feec() - add comments kernel/sched/fair.c | 108 ++++++++++++++++++++++++++++++++++---------- 1 file changed, 83 insertions(+), 25 deletions(-) diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index e9d906a9bba9..29adb9e27b3d 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -4525,8 +4525,7 @@ static inline int util_fits_cpu(unsigned long util, * 2. The system is being saturated when we're operating near * max capacity, it doesn't make sense to block overutilized. */ - uclamp_max_fits = (capacity_orig == SCHED_CAPACITY_SCALE) && (uclamp_max == SCHED_CAPACITY_SCALE); - uclamp_max_fits = !uclamp_max_fits && (uclamp_max <= capacity_orig); + uclamp_max_fits = (uclamp_max <= capacity_orig) || (capacity_orig == SCHED_CAPACITY_SCALE); fits = fits || uclamp_max_fits; /* @@ -4561,8 +4560,8 @@ static inline int util_fits_cpu(unsigned long util, * handle the case uclamp_min > uclamp_max. */ uclamp_min = min(uclamp_min, uclamp_max); - if (util < uclamp_min && capacity_orig != SCHED_CAPACITY_SCALE) - fits = fits && (uclamp_min <= capacity_orig_thermal); + if (fits && (util < uclamp_min) && (uclamp_min > capacity_orig_thermal)) + return -1; return fits; } @@ -4572,7 +4571,11 @@ static inline int task_fits_cpu(struct task_struct *p, int cpu) unsigned long uclamp_min = uclamp_eff_value(p, UCLAMP_MIN); unsigned long uclamp_max = uclamp_eff_value(p, UCLAMP_MAX); unsigned long util = task_util_est(p); - return util_fits_cpu(util, uclamp_min, uclamp_max, cpu); + /* + * Return true only if the cpu fully fits the task requirements, which + * include the utilization but also the performance. + */ + return (util_fits_cpu(util, uclamp_min, uclamp_max, cpu) > 0); } static inline void update_misfit_status(struct task_struct *p, struct rq *rq) @@ -6132,6 +6135,7 @@ static inline bool cpu_overutilized(int cpu) unsigned long rq_util_min = uclamp_rq_get(cpu_rq(cpu), UCLAMP_MIN); unsigned long rq_util_max = uclamp_rq_get(cpu_rq(cpu), UCLAMP_MAX); + /* Return true only if the utlization doesn't fit its capacity */ return !util_fits_cpu(cpu_util_cfs(cpu), rq_util_min, rq_util_max, cpu); } @@ -6925,6 +6929,7 @@ static int select_idle_capacity(struct task_struct *p, struct sched_domain *sd, int target) { unsigned long task_util, util_min, util_max, best_cap = 0; + int fits, best_fits = 0; int cpu, best_cpu = -1; struct cpumask *cpus; @@ -6940,12 +6945,28 @@ select_idle_capacity(struct task_struct *p, struct sched_domain *sd, int target) if (!available_idle_cpu(cpu) && !sched_idle_cpu(cpu)) continue; - if (util_fits_cpu(task_util, util_min, util_max, cpu)) + + fits = util_fits_cpu(task_util, util_min, util_max, cpu); + + /* This CPU fits with all capacity and performance requirements */ + if (fits > 0) return cpu; + /* + * Only the min performance (i.e. uclamp_min) doesn't fit. Look + * for the CPU with highest performance capacity. + */ + else if (fits < 0) + cpu_cap = capacity_orig_of(cpu) - thermal_load_avg(cpu_rq(cpu)); - if (cpu_cap > best_cap) { + /* + * First, select cpu which fits better (-1 being better than 0). + * Then, select the one with largest capacity at same level. + */ + if ((fits < best_fits) || + ((fits == best_fits) && (cpu_cap > best_cap))) { best_cap = cpu_cap; best_cpu = cpu; + best_fits = fits; } } @@ -6958,7 +6979,11 @@ static inline bool asym_fits_cpu(unsigned long util, int cpu) { if (sched_asym_cpucap_active()) - return util_fits_cpu(util, util_min, util_max, cpu); + /* + * Return true only if the cpu fully fits the task requirements + * which include the utilization but also the performance. + */ + return (util_fits_cpu(util, util_min, util_max, cpu) > 0); return true; } @@ -7325,6 +7350,9 @@ static int find_energy_efficient_cpu(struct task_struct *p, int prev_cpu) unsigned long p_util_max = uclamp_is_used() ? uclamp_eff_value(p, UCLAMP_MAX) : 1024; struct root_domain *rd = this_rq()->rd; int cpu, best_energy_cpu, target = -1; + int prev_fits = -1, best_fits = -1; + unsigned long best_thermal_cap = 0; + unsigned long prev_thermal_cap = 0; struct sched_domain *sd; struct perf_domain *pd; struct energy_env eenv; @@ -7360,6 +7388,7 @@ static int find_energy_efficient_cpu(struct task_struct *p, int prev_cpu) unsigned long prev_spare_cap = 0; int max_spare_cap_cpu = -1; unsigned long base_energy; + int fits, max_fits = -1; cpumask_and(cpus, perf_domain_span(pd), cpu_online_mask); @@ -7412,7 +7441,9 @@ static int find_energy_efficient_cpu(struct task_struct *p, int prev_cpu) util_max = max(rq_util_max, p_util_max); } } - if (!util_fits_cpu(util, util_min, util_max, cpu)) + + fits = util_fits_cpu(util, util_min, util_max, cpu); + if (!fits) continue; lsub_positive(&cpu_cap, util); @@ -7420,7 +7451,9 @@ static int find_energy_efficient_cpu(struct task_struct *p, int prev_cpu) if (cpu == prev_cpu) { /* Always use prev_cpu as a candidate. */ prev_spare_cap = cpu_cap; - } else if (cpu_cap > max_spare_cap) { + prev_fits = fits; + } else if ((fits > max_fits) || + ((fits == max_fits) && (cpu_cap > max_spare_cap))) { /* * Find the CPU with the maximum spare capacity * among the remaining CPUs in the performance @@ -7428,6 +7461,7 @@ static int find_energy_efficient_cpu(struct task_struct *p, int prev_cpu) */ max_spare_cap = cpu_cap; max_spare_cap_cpu = cpu; + max_fits = fits; } } @@ -7446,26 +7480,50 @@ static int find_energy_efficient_cpu(struct task_struct *p, int prev_cpu) if (prev_delta < base_energy) goto unlock; prev_delta -= base_energy; + prev_thermal_cap = cpu_thermal_cap; best_delta = min(best_delta, prev_delta); } /* Evaluate the energy impact of using max_spare_cap_cpu. */ if (max_spare_cap_cpu >= 0 && max_spare_cap > prev_spare_cap) { + /* Current best energy cpu fits better */ + if (max_fits < best_fits) + continue; + + /* + * Both don't fit performance (i.e. uclamp_min) but + * best energy cpu has better performance. + */ + if ((max_fits < 0) && + (cpu_thermal_cap <= best_thermal_cap)) + continue; + cur_delta = compute_energy(&eenv, pd, cpus, p, max_spare_cap_cpu); /* CPU utilization has changed */ if (cur_delta < base_energy) goto unlock; cur_delta -= base_energy; - if (cur_delta < best_delta) { - best_delta = cur_delta; - best_energy_cpu = max_spare_cap_cpu; - } + + /* + * Both fit for the task but best energy cpu has lower + * energy impact. + */ + if ((max_fits > 0) && (best_fits > 0) && + (cur_delta >= best_delta)) + continue; + + best_delta = cur_delta; + best_energy_cpu = max_spare_cap_cpu; + best_fits = max_fits; + best_thermal_cap = cpu_thermal_cap; } } rcu_read_unlock(); - if (best_delta < prev_delta) + if ((best_fits > prev_fits) || + ((best_fits > 0) && (best_delta < prev_delta)) || + ((best_fits < 0) && (best_thermal_cap > prev_thermal_cap))) target = best_energy_cpu; return target; @@ -10259,24 +10317,23 @@ static struct sched_group *find_busiest_group(struct lb_env *env) */ update_sd_lb_stats(env, &sds); - if (sched_energy_enabled()) { - struct root_domain *rd = env->dst_rq->rd; - - if (rcu_dereference(rd->pd) && !READ_ONCE(rd->overutilized)) - goto out_balanced; - } - - local = &sds.local_stat; - busiest = &sds.busiest_stat; - /* There is no busy sibling group to pull tasks from */ if (!sds.busiest) goto out_balanced; + busiest = &sds.busiest_stat; + /* Misfit tasks should be dealt with regardless of the avg load */ if (busiest->group_type == group_misfit_task) goto force_balance; + if (sched_energy_enabled()) { + struct root_domain *rd = env->dst_rq->rd; + + if (rcu_dereference(rd->pd) && !READ_ONCE(rd->overutilized)) + goto out_balanced; + } + /* ASYM feature bypasses nice load balance check */ if (busiest->group_type == group_asym_packing) goto force_balance; @@ -10289,6 +10346,7 @@ static struct sched_group *find_busiest_group(struct lb_env *env) if (busiest->group_type == group_imbalanced) goto force_balance; + local = &sds.local_stat; /* * If the local group is busier than the selected busiest group * don't try and pull any tasks.