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;
From:   "Rafael J. Wysocki" <rjw@rjwysocki.net>
To:     Peter Zijlstra <peterz@infradead.org>,
        Linux PM <linux-pm@vger.kernel.org>
Cc:     Frederic Weisbecker <fweisbec@gmail.com>,
        Thomas Gleixner <tglx@linutronix.de>,
        Paul McKenney <paulmck@linux.vnet.ibm.com>,
        Thomas Ilsche <thomas.ilsche@tu-dresden.de>,
        Doug Smythies <dsmythies@telus.net>,
        Rik van Riel <riel@surriel.com>,
        Aubrey Li <aubrey.li@linux.intel.com>,
        Mike Galbraith <mgalbraith@suse.de>,
        LKML <linux-kernel@vger.kernel.org>
Subject: [RFT][PATCH v7.1 5/8] cpuidle: Return nohz hint from cpuidle_select()
Date:   Wed, 21 Mar 2018 07:48:38 +0100
Message-ID: <2111949.c172CzMuf8@aspire.rjw.lan>
In-Reply-To: <1635957.yuHkCe9oyz@aspire.rjw.lan>
References: <2390019.oHdSGtR3EE@aspire.rjw.lan> <1635957.yuHkCe9oyz@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

From: Rafael J. Wysocki <rafael.j.wysocki@intel.com>

Add a new pointer argument to cpuidle_select() and to the ->select
cpuidle governor callback to allow a boolean value indicating
whether or not the tick should be stopped before entering the
selected state to be returned from there.

Make the ladder governor ignore that pointer (to preserve its
current behavior) and make the menu governor return 'false" through
it if:
 (1) the idle exit latency is constrained at 0, or
 (2) the selected state is a polling one, or
 (3) the expected idle period duration is within the tick period
     range.

In addition to that, the correction factor computations in the menu
governor need to take the possibility that the tick may not be
stopped into account to avoid artificially small correction factor
values.  To that end, add a mechanism to record tick wakeups, as
suggested by Peter Zijlstra, and use it to modify the menu_reflect()
behavior when tick wakeup occurs.  Namely, if the CPU is woken up by
the tick, the predicted idle duration is likely too short, so make
menu_reflect() try to compensate by bumping up the correction factor
with a (sufficiently large) constant value.

Since the value returned through the new argument pointer of
cpuidle_select() is not used by its caller yet, this change by
itself is not expected to alter the functionality of the code.

Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
---

Another variant of patch [5/8] to test.

This one doesn't run menu_update() at all on tick wakeups, but simply
bumps up the correction factor alone then.

It may give better results, but please test and compare.

I have created a git branch with this variant for easier testing:

 git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm.git \
 idle-loop-v7.1

---
 drivers/cpuidle/cpuidle.c          |   10 ++++++-
 drivers/cpuidle/governors/ladder.c |    3 +-
 drivers/cpuidle/governors/menu.c   |   49 +++++++++++++++++++++++++++++++++----
 include/linux/cpuidle.h            |    8 +++---
 include/linux/tick.h               |    2 +
 kernel/sched/idle.c                |    4 ++-
 kernel/time/tick-sched.c           |   20 +++++++++++++++
 7 files changed, 84 insertions(+), 12 deletions(-)

Index: linux-pm/include/linux/cpuidle.h
===================================================================
--- linux-pm.orig/include/linux/cpuidle.h
+++ linux-pm/include/linux/cpuidle.h
@@ -135,7 +135,8 @@ extern bool cpuidle_not_available(struct
 				  struct cpuidle_device *dev);
 
 extern int cpuidle_select(struct cpuidle_driver *drv,
-			  struct cpuidle_device *dev);
+			  struct cpuidle_device *dev,
+			  bool *stop_tick);
 extern int cpuidle_enter(struct cpuidle_driver *drv,
 			 struct cpuidle_device *dev, int index);
 extern void cpuidle_reflect(struct cpuidle_device *dev, int index);
@@ -167,7 +168,7 @@ static inline bool cpuidle_not_available
 					 struct cpuidle_device *dev)
 {return true; }
 static inline int cpuidle_select(struct cpuidle_driver *drv,
-				 struct cpuidle_device *dev)
+				 struct cpuidle_device *dev, bool *stop_tick)
 {return -ENODEV; }
 static inline int cpuidle_enter(struct cpuidle_driver *drv,
 				struct cpuidle_device *dev, int index)
@@ -250,7 +251,8 @@ struct cpuidle_governor {
 					struct cpuidle_device *dev);
 
 	int  (*select)		(struct cpuidle_driver *drv,
-					struct cpuidle_device *dev);
+					struct cpuidle_device *dev,
+					bool *stop_tick);
 	void (*reflect)		(struct cpuidle_device *dev, int index);
 };
 
Index: linux-pm/kernel/sched/idle.c
===================================================================
--- linux-pm.orig/kernel/sched/idle.c
+++ linux-pm/kernel/sched/idle.c
@@ -188,13 +188,15 @@ static void cpuidle_idle_call(void)
 		next_state = cpuidle_find_deepest_state(drv, dev);
 		call_cpuidle(drv, dev, next_state);
 	} else {
+		bool stop_tick = true;
+
 		tick_nohz_idle_stop_tick();
 		rcu_idle_enter();
 
 		/*
 		 * Ask the cpuidle framework to choose a convenient idle state.
 		 */
-		next_state = cpuidle_select(drv, dev);
+		next_state = cpuidle_select(drv, dev, &stop_tick);
 		entered_state = call_cpuidle(drv, dev, next_state);
 		/*
 		 * Give the governor an opportunity to reflect on the outcome
Index: linux-pm/drivers/cpuidle/cpuidle.c
===================================================================
--- linux-pm.orig/drivers/cpuidle/cpuidle.c
+++ linux-pm/drivers/cpuidle/cpuidle.c
@@ -272,12 +272,18 @@ int cpuidle_enter_state(struct cpuidle_d
  *
  * @drv: the cpuidle driver
  * @dev: the cpuidle device
+ * @stop_tick: indication on whether or not to stop the tick
  *
  * Returns the index of the idle state.  The return value must not be negative.
+ *
+ * The memory location pointed to by @stop_tick is expected to be written the
+ * 'false' boolean value if the scheduler tick should not be stopped before
+ * entering the returned state.
  */
-int cpuidle_select(struct cpuidle_driver *drv, struct cpuidle_device *dev)
+int cpuidle_select(struct cpuidle_driver *drv, struct cpuidle_device *dev,
+		   bool *stop_tick)
 {
-	return cpuidle_curr_governor->select(drv, dev);
+	return cpuidle_curr_governor->select(drv, dev, stop_tick);
 }
 
 /**
Index: linux-pm/drivers/cpuidle/governors/ladder.c
===================================================================
--- linux-pm.orig/drivers/cpuidle/governors/ladder.c
+++ linux-pm/drivers/cpuidle/governors/ladder.c
@@ -63,9 +63,10 @@ static inline void ladder_do_selection(s
  * ladder_select_state - selects the next state to enter
  * @drv: cpuidle driver
  * @dev: the CPU
+ * @dummy: not used
  */
 static int ladder_select_state(struct cpuidle_driver *drv,
-				struct cpuidle_device *dev)
+			       struct cpuidle_device *dev, bool *dummy)
 {
 	struct ladder_device *ldev = this_cpu_ptr(&ladder_devices);
 	struct device *device = get_cpu_device(dev->cpu);
Index: linux-pm/drivers/cpuidle/governors/menu.c
===================================================================
--- linux-pm.orig/drivers/cpuidle/governors/menu.c
+++ linux-pm/drivers/cpuidle/governors/menu.c
@@ -279,8 +279,10 @@ again:
  * menu_select - selects the next idle state to enter
  * @drv: cpuidle driver containing state data
  * @dev: the CPU
+ * @stop_tick: indication on whether or not to stop the tick
  */
-static int menu_select(struct cpuidle_driver *drv, struct cpuidle_device *dev)
+static int menu_select(struct cpuidle_driver *drv, struct cpuidle_device *dev,
+		       bool *stop_tick)
 {
 	struct menu_device *data = this_cpu_ptr(&menu_devices);
 	struct device *device = get_cpu_device(dev->cpu);
@@ -303,8 +305,10 @@ static int menu_select(struct cpuidle_dr
 		latency_req = resume_latency;
 
 	/* Special case when user has set very strict latency requirement */
-	if (unlikely(latency_req == 0))
+	if (unlikely(latency_req == 0)) {
+		*stop_tick = false;
 		return 0;
+	}
 
 	/* determine the expected residency time, round up */
 	data->next_timer_us = ktime_to_us(tick_nohz_get_sleep_length());
@@ -354,6 +358,7 @@ static int menu_select(struct cpuidle_dr
 	if (latency_req > interactivity_req)
 		latency_req = interactivity_req;
 
+	expected_interval = data->predicted_us;
 	/*
 	 * Find the idle state with the lowest power while satisfying
 	 * our constraints.
@@ -369,15 +374,30 @@ static int menu_select(struct cpuidle_dr
 			idx = i; /* first enabled state */
 		if (s->target_residency > data->predicted_us)
 			break;
-		if (s->exit_latency > latency_req)
+		if (s->exit_latency > latency_req) {
+			/*
+			 * If we break out of the loop for latency reasons, use
+			 * the target residency of the selected state as the
+			 * expected idle duration so that the tick is retained
+			 * as long as that target residency is low enough.
+			 */
+			expected_interval = drv->states[idx].target_residency;
 			break;
-
+		}
 		idx = i;
 	}
 
 	if (idx == -1)
 		idx = 0; /* No states enabled. Must use 0. */
 
+	/*
+	 * Don't stop the tick if the selected state is a polling one or if the
+	 * expected idle duration is shorter than the tick period length.
+	 */
+	if ((drv->states[idx].flags & CPUIDLE_FLAG_POLLING) ||
+	    expected_interval < TICK_USEC)
+		*stop_tick = false;
+
 	data->last_state_idx = idx;
 
 	return data->last_state_idx;
@@ -396,7 +416,26 @@ static void menu_reflect(struct cpuidle_
 	struct menu_device *data = this_cpu_ptr(&menu_devices);
 
 	data->last_state_idx = index;
-	data->needs_update = 1;
+	if (tick_nohz_idle_got_tick()) {
+		unsigned int new_factor = data->correction_factor[data->bucket];
+
+		/*
+		 * Only update the correction factor, don't update the repeating
+		 * pattern data to avoid polluting it with the tick period
+		 * length which is an artificial addition.
+		 */
+		new_factor -= new_factor / DECAY;
+		/*
+		 * If the CPU was woken up by the tick, the predicted idle time
+		 * was likely too short.  Try to compensate by bumping up the
+		 * correction factor.  Use 0.75 * RESOLUTION (which is easy
+		 * enough to get) that should work on the average.
+		 */
+		new_factor += RESOLUTION / 2 + RESOLUTION / 4;
+		data->correction_factor[data->bucket] = new_factor;
+	} else {
+		data->needs_update = 1;
+	}
 }
 
 /**
Index: linux-pm/kernel/time/tick-sched.c
===================================================================
--- linux-pm.orig/kernel/time/tick-sched.c
+++ linux-pm/kernel/time/tick-sched.c
@@ -991,6 +991,20 @@ void tick_nohz_irq_exit(void)
 }
 
 /**
+ * tick_nohz_idle_got_tick - Check whether or not the tick handler has run
+ */
+bool tick_nohz_idle_got_tick(void)
+{
+	struct tick_sched *ts = this_cpu_ptr(&tick_cpu_sched);
+
+	if (ts->inidle > 1) {
+		ts->inidle = 1;
+		return true;
+	}
+	return false;
+}
+
+/**
  * tick_nohz_get_sleep_length - return the length of the current sleep
  *
  * Called from power state control code with interrupts disabled
@@ -1101,6 +1115,9 @@ static void tick_nohz_handler(struct clo
 	struct pt_regs *regs = get_irq_regs();
 	ktime_t now = ktime_get();
 
+	if (ts->inidle)
+		ts->inidle = 2;
+
 	dev->next_event = KTIME_MAX;
 
 	tick_sched_do_timer(now);
@@ -1198,6 +1215,9 @@ static enum hrtimer_restart tick_sched_t
 	struct pt_regs *regs = get_irq_regs();
 	ktime_t now = ktime_get();
 
+	if (ts->inidle)
+		ts->inidle = 2;
+
 	tick_sched_do_timer(now);
 
 	/*
Index: linux-pm/include/linux/tick.h
===================================================================
--- linux-pm.orig/include/linux/tick.h
+++ linux-pm/include/linux/tick.h
@@ -119,6 +119,7 @@ extern void tick_nohz_idle_restart_tick(
 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 unsigned long tick_nohz_get_idle_calls(void);
 extern unsigned long tick_nohz_get_idle_calls_cpu(int cpu);
@@ -139,6 +140,7 @@ static inline void tick_nohz_idle_stop_t
 static inline void tick_nohz_idle_restart_tick(void) { }
 static inline void tick_nohz_idle_enter(void) { }
 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)
 {