linux-kernel - [PATCH v11 7/8] cpuidle: Pre-store next timer/tick before selecting an idle state

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-Id: <20190226145415.19411-8-ulf.hansson@linaro.org>
Date:   Tue, 26 Feb 2019 15:54:14 +0100
From:   Ulf Hansson <ulf.hansson@...aro.org>
To:     "Rafael J . Wysocki" <rjw@...ysocki.net>, linux-pm@...r.kernel.org
Cc:     Frederic Weisbecker <fweisbec@...il.com>,
        Thomas Gleixner <tglx@...utronix.de>,
        Sudeep Holla <sudeep.holla@....com>,
        Lorenzo Pieralisi <Lorenzo.Pieralisi@....com>,
        Mark Rutland <mark.rutland@....com>,
        Daniel Lezcano <daniel.lezcano@...aro.org>,
        "Raju P . L . S . S . S . N" <rplsssn@...eaurora.org>,
        Stephen Boyd <sboyd@...nel.org>,
        Tony Lindgren <tony@...mide.com>,
        Kevin Hilman <khilman@...nel.org>,
        Lina Iyer <ilina@...eaurora.org>,
        Ulf Hansson <ulf.hansson@...aro.org>,
        Viresh Kumar <viresh.kumar@...aro.org>,
        Vincent Guittot <vincent.guittot@...aro.org>,
        Geert Uytterhoeven <geert+renesas@...der.be>,
        linux-arm-kernel@...ts.infradead.org,
        linux-arm-msm@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: [PATCH v11 7/8] cpuidle: Pre-store next timer/tick before selecting an idle state

A common piece of data used by cpuidle governors, is the information about
when the next timer/tick is going to fire. Rather than having each governor
calling tick_nohz_get_next_timer|hrtimer() separately, let's consolidate
the code by calling these functions before invoking the ->select() callback
of the governor - and store the output data in the struct cpuidle_device.

Besides the consolidation benefit, the purpose of this change is also to
make the information about the next timer/tick, available outside the
cpuidle framework. Following changes that implements a new genpd governor,
makes use of this.

Signed-off-by: Ulf Hansson <ulf.hansson@...aro.org>
---

Changes in v11:
	- New patch.

---
 drivers/cpuidle/cpuidle.c        | 2 ++
 drivers/cpuidle/governors/menu.c | 6 ++----
 drivers/cpuidle/governors/teo.c  | 6 ++----
 include/linux/cpuidle.h          | 2 ++
 4 files changed, 8 insertions(+), 8 deletions(-)

diff --git a/drivers/cpuidle/cpuidle.c b/drivers/cpuidle/cpuidle.c
index 7f108309e871..3b148253036b 100644
--- a/drivers/cpuidle/cpuidle.c
+++ b/drivers/cpuidle/cpuidle.c
@@ -312,6 +312,8 @@ int cpuidle_enter_state(struct cpuidle_device *dev, struct cpuidle_driver *drv,
 int cpuidle_select(struct cpuidle_driver *drv, struct cpuidle_device *dev,
 		   bool *stop_tick)
 {
+	dev->next_timer = tick_nohz_get_next_timer();
+	dev->next_hrtimer = tick_nohz_get_next_hrtimer();
 	return cpuidle_curr_governor->select(drv, dev, stop_tick);
 }
 
diff --git a/drivers/cpuidle/governors/menu.c b/drivers/cpuidle/governors/menu.c
index 95e9122d6047..cdbe434e783d 100644
--- a/drivers/cpuidle/governors/menu.c
+++ b/drivers/cpuidle/governors/menu.c
@@ -287,8 +287,6 @@ static int menu_select(struct cpuidle_driver *drv, struct cpuidle_device *dev,
 	unsigned long nr_iowaiters;
 	ktime_t delta_next;
 	ktime_t now = ktime_get();
-	ktime_t next_hrtimer = tick_nohz_get_next_hrtimer();
-	ktime_t next_timer = tick_nohz_get_next_timer();
 
 	if (data->needs_update) {
 		menu_update(drv, dev);
@@ -298,14 +296,14 @@ static int menu_select(struct cpuidle_driver *drv, struct cpuidle_device *dev,
 	/*
 	 * Compute the duration before the next timer, whatever the origin
 	 */
-	delta_next = ktime_sub(next_timer, now);
+	delta_next = ktime_sub(dev->next_timer, now);
 	data->next_timer_us = ktime_to_us(delta_next);
 
 	/*
 	 * Compute the duration before next hrtimer which is the tick
 	 * or an earliest hrtimer
 	 */
-	delta_next = ktime_sub(next_hrtimer, now);
+	delta_next = ktime_sub(dev->next_hrtimer, now);
 
 	nr_iowaiters = nr_iowait_cpu(dev->cpu);
 	data->bucket = which_bucket(data->next_timer_us, nr_iowaiters);
diff --git a/drivers/cpuidle/governors/teo.c b/drivers/cpuidle/governors/teo.c
index bef1e95c597e..7af9851d9d40 100644
--- a/drivers/cpuidle/governors/teo.c
+++ b/drivers/cpuidle/governors/teo.c
@@ -245,8 +245,6 @@ static int teo_select(struct cpuidle_driver *drv, struct cpuidle_device *dev,
 	int max_early_idx, idx, i;
 	ktime_t delta_tick;
 	ktime_t now = ktime_get();
-	ktime_t next_hrtimer = tick_nohz_get_next_hrtimer();
-	ktime_t next_timer = tick_nohz_get_next_timer();
 
 	if (cpu_data->last_state >= 0) {
 		teo_update(drv, dev);
@@ -255,8 +253,8 @@ static int teo_select(struct cpuidle_driver *drv, struct cpuidle_device *dev,
 
 	cpu_data->time_span_ns = local_clock();
 
-	cpu_data->sleep_length_ns = ktime_sub(next_timer, now);
-	delta_tick = ktime_sub(next_hrtimer, now);
+	cpu_data->sleep_length_ns = ktime_sub(dev->next_timer, now);
+	delta_tick = ktime_sub(dev->next_hrtimer, now);
 	duration_us = ktime_to_us(cpu_data->sleep_length_ns);
 
 	count = 0;
diff --git a/include/linux/cpuidle.h b/include/linux/cpuidle.h
index 3b39472324a3..81ec924206f0 100644
--- a/include/linux/cpuidle.h
+++ b/include/linux/cpuidle.h
@@ -83,6 +83,8 @@ struct cpuidle_device {
 	unsigned int		use_deepest_state:1;
 	unsigned int		poll_time_limit:1;
 	unsigned int		cpu;
+	ktime_t			next_timer;
+	ktime_t			next_hrtimer;
 
 	int			last_residency;
 	struct cpuidle_state_usage	states_usage[CPUIDLE_STATE_MAX];
-- 
2.17.1