Merge tag 'pm-4.19-rc1-2' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm

Pull more power management updates from Rafael Wysocki:
 "These fix the main idle loop and the menu cpuidle governor, clean up
  the latter, fix a mistake in the PCI bus type's support for system
  suspend and resume, fix the ondemand and conservative cpufreq
  governors, address a build issue in the system wakeup framework and
  make the ACPI C-states desciptions less confusing.

  Specifics:

   - Make the idle loop handle stopped scheduler tick correctly (Rafael
     Wysocki).

   - Prevent the menu cpuidle governor from letting CPUs spend too much
     time in shallow idle states when it is invoked with scheduler tick
     stopped and clean it up somewhat (Rafael Wysocki).

   - Avoid invoking the platform firmware to make the platform enter the
     ACPI S3 sleep state with suspended PCIe root ports which may
     confuse the firmware and cause it to crash (Rafael Wysocki).

   - Fix sysfs-related race in the ondemand and conservative cpufreq
     governors which may cause the system to crash if the governor
     module is removed during an update of CPU frequency limits (Henry
     Willard).

   - Select SRCU when building the system wakeup framework to avoid a
     build issue in it (zhangyi).

   - Make the descriptions of ACPI C-states vendor-neutral to avoid
     confusion (Prarit Bhargava)"

* tag 'pm-4.19-rc1-2' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
  cpuidle: menu: Handle stopped tick more aggressively
  sched: idle: Avoid retaining the tick when it has been stopped
  PCI / ACPI / PM: Resume all bridges on suspend-to-RAM
  cpuidle: menu: Update stale polling override comment
  cpufreq: governor: Avoid accessing invalid governor_data
  x86/ACPI/cstate: Make APCI C1 FFH MWAIT C-state description vendor-neutral
  cpuidle: menu: Fix white space
  PM / sleep: wakeup: Fix build error caused by missing SRCU support

1 files changed, 28 insertions, 17 deletions

diff --git a/drivers/cpuidle/governors/menu.c b/drivers/cpuidle/governors/menu.c
index 1aef60d160eb..110483f0e3fb 100644
--- a/drivers/cpuidle/governors/menu.c
+++ b/drivers/cpuidle/governors/menu.c
@@ -328,9 +328,8 @@ static int menu_select(struct cpuidle_driver *drv, struct cpuidle_device *dev,
 		unsigned int polling_threshold;
 
 		/*
-		 * We want to default to C1 (hlt), not to busy polling
-		 * unless the timer is happening really really soon, or
-		 * C1's exit latency exceeds the user configured limit.
+		 * Default to a physical idle state, not to busy polling, unless
+		 * a timer is going to trigger really really soon.
 		 */
 		polling_threshold = max_t(unsigned int, 20, s->target_residency);
 		if (data->next_timer_us > polling_threshold &&
@@ -349,14 +348,12 @@ static int menu_select(struct cpuidle_driver *drv, struct cpuidle_device *dev,
 		 * If the tick is already stopped, the cost of possible short
 		 * idle duration misprediction is much higher, because the CPU
 		 * may be stuck in a shallow idle state for a long time as a
-		 * result of it.  In that case say we might mispredict and try
-		 * to force the CPU into a state for which we would have stopped
-		 * the tick, unless a timer is going to expire really soon
-		 * anyway.
+		 * result of it.  In that case say we might mispredict and use
+		 * the known time till the closest timer event for the idle
+		 * state selection.
 		 */
 		if (data->predicted_us < TICK_USEC)
-			data->predicted_us = min_t(unsigned int, TICK_USEC,
-						   ktime_to_us(delta_next));
+			data->predicted_us = ktime_to_us(delta_next);
 	} else {
 		/*
 		 * Use the performance multiplier and the user-configurable
@@ -381,8 +378,22 @@ static int menu_select(struct cpuidle_driver *drv, struct cpuidle_device *dev,
 			continue;
 		if (idx == -1)
 			idx = i; /* first enabled state */
-		if (s->target_residency > data->predicted_us)
-			break;
+		if (s->target_residency > data->predicted_us) {
+			if (!tick_nohz_tick_stopped())
+				break;
+
+			/*
+			 * If the state selected so far is shallow and this
+			 * state's target residency matches the time till the
+			 * closest timer event, select this one to avoid getting
+			 * stuck in the shallow one for too long.
+			 */
+			if (drv->states[idx].target_residency < TICK_USEC &&
+			    s->target_residency <= ktime_to_us(delta_next))
+				idx = i;
+
+			goto out;
+		}
 		if (s->exit_latency > latency_req) {
 			/*
 			 * If we break out of the loop for latency reasons, use
@@ -403,14 +414,13 @@ static int menu_select(struct cpuidle_driver *drv, struct cpuidle_device *dev,
 	 * 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) {
+	if (((drv->states[idx].flags & CPUIDLE_FLAG_POLLING) ||
+	     expected_interval < TICK_USEC) && !tick_nohz_tick_stopped()) {
 		unsigned int delta_next_us = ktime_to_us(delta_next);
 
 		*stop_tick = false;
 
-		if (!tick_nohz_tick_stopped() && idx > 0 &&
-		    drv->states[idx].target_residency > delta_next_us) {
+		if (idx > 0 && drv->states[idx].target_residency > delta_next_us) {
 			/*
 			 * The tick is not going to be stopped and the target
 			 * residency of the state to be returned is not within
@@ -418,8 +428,8 @@ static int menu_select(struct cpuidle_driver *drv, struct cpuidle_device *dev,
 			 * tick, so try to correct that.
 			 */
 			for (i = idx - 1; i >= 0; i--) {
-			    if (drv->states[i].disabled ||
-			        dev->states_usage[i].disable)
+				if (drv->states[i].disabled ||
+				    dev->states_usage[i].disable)
 					continue;
 
 				idx = i;
@@ -429,6 +439,7 @@ static int menu_select(struct cpuidle_driver *drv, struct cpuidle_device *dev,
 		}
 	}
 
+out:
 	data->last_state_idx = idx;
 
 	return data->last_state_idx;