Received: by 10.213.65.68 with SMTP id h4csp567054imn; Tue, 20 Mar 2018 09:47:52 -0700 (PDT) X-Google-Smtp-Source: AG47ELvhmq3WG0h8UGOW/HDQJ8pXsmhMYIpIofDqVP42wE1IU2Hhc+d8luObVaRuzSCQaDsLeLlu X-Received: by 2002:a17:902:76c5:: with SMTP id j5-v6mr8845732plt.166.1521564472531; Tue, 20 Mar 2018 09:47:52 -0700 (PDT) ARC-Seal: i=1; a=rsa-sha256; t=1521564472; cv=none; d=google.com; s=arc-20160816; b=g6dNU9uA7BvNVBaAYfCuieHlJcgD1QE115EwPpX2ulv4vuMXcmSreG4dIgnWAEF2tX rJTcPad8/zYGANdTNtf69lzZ259pLIdnl1EdArmm3usI7Y/jA9c1t+oHg7EizfH9wUUT 2W//aogcnShLT9m+3NCLNBJyqjmNdteK1Mig7ThCwkOZ/96v9ogOLGDFBo9kQTUvUWeC n6/RUu5gIvAqThAn6HPBkKZLQKYkTwURmrN85TrOZ3XlFQ5AdJRNpBG4C6FbOx40qoB/ Cj0m5SZ1c+FNH3tYSNegfghRM4rkMu5L32yq0IsCKbNGAgeSXncIoTR805PbyPCAHMx+ UfCA== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=arc-20160816; h=list-id:precedence:sender:content-transfer-encoding:mime-version :references:in-reply-to:message-id:date:subject:cc:to:from :arc-authentication-results; bh=XMDtrnX12WpmpbupQMHqbpi8pjHx7nLJSSCGLZShTuM=; b=eDUPXwb1RW2u3qiLFoSgyiSCGPngn5idKabFphFXiNdMujadIlLawLhlwv2ANyyUvP eGBnUV+RJCwYhsCBMbnjdQk8T0GdUDKLTb9g14fhwqPlKUiurX4o/g1XfzUy1H1EWQQD pg4dJDs4lyxCWxkFttg5hkQ1GDvbHLzk5ZKibzbmtp5UX46SAvft0zP1QZ5dNca1smJL ZasYXMDbUOF2eJOxG0JMap2K5Ft7R08ffgEtNl0rxi36XQlFhN0CCCO+EzjI7PoMVjlw rdOoSRD6yLTdEKIcq1iGGUJI9afPnUwGSWF88oEWwQhQxmWUJNRUyrml9jFrUEZthI65 e48Q== ARC-Authentication-Results: i=1; mx.google.com; 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 Return-Path: Received: from vger.kernel.org (vger.kernel.org. [209.132.180.67]) by mx.google.com with ESMTP id e7si1394706pgc.571.2018.03.20.09.47.37; Tue, 20 Mar 2018 09:47:52 -0700 (PDT) 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; 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 Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752096AbeCTQp7 (ORCPT + 99 others); Tue, 20 Mar 2018 12:45:59 -0400 Received: from cloudserver094114.home.pl ([79.96.170.134]:47074 "EHLO cloudserver094114.home.pl" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751531AbeCTQnn (ORCPT ); Tue, 20 Mar 2018 12:43:43 -0400 Received: from 79.184.254.228.ipv4.supernova.orange.pl (79.184.254.228) (HELO aspire.rjw.lan) by serwer1319399.home.pl (79.96.170.134) with SMTP (IdeaSmtpServer 0.83) id 28165cbc651544f1; Tue, 20 Mar 2018 17:43:41 +0100 From: "Rafael J. Wysocki" To: Peter Zijlstra , Linux PM Cc: Frederic Weisbecker , Thomas Gleixner , Paul McKenney , Thomas Ilsche , Doug Smythies , Rik van Riel , Aubrey Li , Mike Galbraith , LKML Subject: [RFT][PATCH v7 7/8] cpuidle: menu: Refine idle state selection for running tick Date: Tue, 20 Mar 2018 16:46:56 +0100 Message-ID: <1581877.Tdb107SmOT@aspire.rjw.lan> In-Reply-To: <2390019.oHdSGtR3EE@aspire.rjw.lan> References: <2390019.oHdSGtR3EE@aspire.rjw.lan> MIME-Version: 1.0 Content-Transfer-Encoding: 7Bit Content-Type: text/plain; charset="us-ascii" Sender: linux-kernel-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org From: Rafael J. Wysocki Subject: [PATCH] cpuidle: menu: Refine idle state selection for running tick If the tick isn't stopped, the target residency of the state selected by the menu governor may be greater than the actual time to the next tick and that means lost energy. To avoid that, make tick_nohz_get_sleep_length() return the current time to the next event (before stopping the tick) in addition to the estimated one via an extra pointer argument and make menu_select() use that value to refine the state selection when necessary. Signed-off-by: Rafael J. Wysocki --- v5 -> v7: * Rebase on top of the new [5-6/8]. * Rename the new argument of tick_nohz_get_sleep_length() to "delta_next" (as requested by Peter). --- drivers/cpuidle/governors/menu.c | 22 ++++++++++++++++++++-- include/linux/tick.h | 7 ++++--- kernel/time/tick-sched.c | 7 +++++-- 3 files changed, 29 insertions(+), 7 deletions(-) Index: linux-pm/include/linux/tick.h =================================================================== --- linux-pm.orig/include/linux/tick.h +++ linux-pm/include/linux/tick.h @@ -121,7 +121,7 @@ extern void tick_nohz_idle_enter(void); extern void tick_nohz_idle_exit(void); extern void tick_nohz_irq_exit(void); extern bool tick_nohz_idle_got_tick(void); -extern ktime_t tick_nohz_get_sleep_length(void); +extern ktime_t tick_nohz_get_sleep_length(ktime_t *delta_next); extern unsigned long tick_nohz_get_idle_calls(void); extern unsigned long tick_nohz_get_idle_calls_cpu(int cpu); extern u64 get_cpu_idle_time_us(int cpu, u64 *last_update_time); @@ -144,9 +144,10 @@ static inline void tick_nohz_idle_enter( static inline void tick_nohz_idle_exit(void) { } static inline bool tick_nohz_idle_got_tick(void) { return false; } -static inline ktime_t tick_nohz_get_sleep_length(void) +static inline ktime_t tick_nohz_get_sleep_length(ktime_t *delta_next) { - return NSEC_PER_SEC / HZ; + *delta_next = NSEC_PER_SEC / HZ; + return *delta_next; } static inline u64 get_cpu_idle_time_us(int cpu, u64 *unused) { return -1; } static inline u64 get_cpu_iowait_time_us(int cpu, u64 *unused) { return -1; } Index: linux-pm/kernel/time/tick-sched.c =================================================================== --- linux-pm.orig/kernel/time/tick-sched.c +++ linux-pm/kernel/time/tick-sched.c @@ -1045,10 +1045,11 @@ bool tick_nohz_idle_got_tick(void) /** * tick_nohz_get_sleep_length - return the expected length of the current sleep + * @delta_next: duration until the next event if the tick cannot be stopped * * Called from power state control code with interrupts disabled */ -ktime_t tick_nohz_get_sleep_length(void) +ktime_t tick_nohz_get_sleep_length(ktime_t *delta_next) { struct clock_event_device *dev = __this_cpu_read(tick_cpu_device.evtdev); struct tick_sched *ts = this_cpu_ptr(&tick_cpu_sched); @@ -1061,6 +1062,8 @@ ktime_t tick_nohz_get_sleep_length(void) WARN_ON_ONCE(!ts->inidle); + *delta_next = ktime_sub(dev->next_event, now); + if (can_stop_idle_tick(cpu, ts)) { ktime_t next_event = tick_nohz_next_event(ts, cpu); @@ -1068,7 +1071,7 @@ ktime_t tick_nohz_get_sleep_length(void) return ktime_sub(next_event, now); } - return ktime_sub(dev->next_event, now); + return *delta_next; } /** Index: linux-pm/drivers/cpuidle/governors/menu.c =================================================================== --- linux-pm.orig/drivers/cpuidle/governors/menu.c +++ linux-pm/drivers/cpuidle/governors/menu.c @@ -295,6 +295,7 @@ static int menu_select(struct cpuidle_dr unsigned int expected_interval; unsigned long nr_iowaiters, cpu_load; int resume_latency = dev_pm_qos_raw_read_value(device); + ktime_t delta_next; if (data->needs_update) { menu_update(drv, dev); @@ -312,7 +313,7 @@ static int menu_select(struct cpuidle_dr } /* determine the expected residency time, round up */ - data->next_timer_us = ktime_to_us(tick_nohz_get_sleep_length()); + data->next_timer_us = ktime_to_us(tick_nohz_get_sleep_length(&delta_next)); get_iowait_load(&nr_iowaiters, &cpu_load); data->bucket = which_bucket(data->next_timer_us, nr_iowaiters); @@ -396,9 +397,26 @@ static int menu_select(struct cpuidle_dr * expected idle duration is shorter than the tick period length. */ if ((drv->states[idx].flags & CPUIDLE_FLAG_POLLING) || - expected_interval < TICK_USEC) + expected_interval < TICK_USEC) { *stop_tick = false; + if (!tick_nohz_tick_stopped()) { + unsigned int delta_next_us = ktime_to_us(delta_next); + + /* + * Because the tick is not going to be stopped, make + * sure that the target residency of the state to be + * returned is within the time to the next timer event + * including the tick. + */ + while (idx > 0 && + (drv->states[idx].target_residency > delta_next_us || + drv->states[idx].disabled || + dev->states_usage[idx].disable)) + idx--; + } + } + data->last_state_idx = idx; return data->last_state_idx;