1 files changed, 137 insertions, 88 deletions

diff --git a/kernel/hrtimer.c b/kernel/hrtimer.c
index e5d98ce50f89..0086628b6e97 100644
--- a/kernel/hrtimer.c
+++ b/kernel/hrtimer.c
@@ -127,11 +127,11 @@ struct hrtimer_clock_base *lock_hrtimer_base(const struct hrtimer *timer,
 	for (;;) {
 		base = timer->base;
 		if (likely(base != NULL)) {
-			spin_lock_irqsave(&base->cpu_base->lock, *flags);
+			raw_spin_lock_irqsave(&base->cpu_base->lock, *flags);
 			if (likely(base == timer->base))
 				return base;
 			/* The timer has migrated to another CPU: */
-			spin_unlock_irqrestore(&base->cpu_base->lock, *flags);
+			raw_spin_unlock_irqrestore(&base->cpu_base->lock, *flags);
 		}
 		cpu_relax();
 	}
@@ -208,13 +208,13 @@ again:
 
 		/* See the comment in lock_timer_base() */
 		timer->base = NULL;
-		spin_unlock(&base->cpu_base->lock);
-		spin_lock(&new_base->cpu_base->lock);
+		raw_spin_unlock(&base->cpu_base->lock);
+		raw_spin_lock(&new_base->cpu_base->lock);
 
 		if (cpu != this_cpu && hrtimer_check_target(timer, new_base)) {
 			cpu = this_cpu;
-			spin_unlock(&new_base->cpu_base->lock);
-			spin_lock(&base->cpu_base->lock);
+			raw_spin_unlock(&new_base->cpu_base->lock);
+			raw_spin_lock(&base->cpu_base->lock);
 			timer->base = base;
 			goto again;
 		}
@@ -230,7 +230,7 @@ lock_hrtimer_base(const struct hrtimer *timer, unsigned long *flags)
 {
 	struct hrtimer_clock_base *base = timer->base;
 
-	spin_lock_irqsave(&base->cpu_base->lock, *flags);
+	raw_spin_lock_irqsave(&base->cpu_base->lock, *flags);
 
 	return base;
 }
@@ -509,13 +509,14 @@ static inline int hrtimer_hres_active(void)
  * next event
  * Called with interrupts disabled and base->lock held
  */
-static void hrtimer_force_reprogram(struct hrtimer_cpu_base *cpu_base)
+static void
+hrtimer_force_reprogram(struct hrtimer_cpu_base *cpu_base, int skip_equal)
 {
 	int i;
 	struct hrtimer_clock_base *base = cpu_base->clock_base;
-	ktime_t expires;
+	ktime_t expires, expires_next;
 
-	cpu_base->expires_next.tv64 = KTIME_MAX;
+	expires_next.tv64 = KTIME_MAX;
 
 	for (i = 0; i < HRTIMER_MAX_CLOCK_BASES; i++, base++) {
 		struct hrtimer *timer;
@@ -531,10 +532,15 @@ static void hrtimer_force_reprogram(struct hrtimer_cpu_base *cpu_base)
 		 */
 		if (expires.tv64 < 0)
 			expires.tv64 = 0;
-		if (expires.tv64 < cpu_base->expires_next.tv64)
-			cpu_base->expires_next = expires;
+		if (expires.tv64 < expires_next.tv64)
+			expires_next = expires;
 	}
 
+	if (skip_equal && expires_next.tv64 == cpu_base->expires_next.tv64)
+		return;
+
+	cpu_base->expires_next.tv64 = expires_next.tv64;
+
 	if (cpu_base->expires_next.tv64 != KTIME_MAX)
 		tick_program_event(cpu_base->expires_next, 1);
 }
@@ -551,7 +557,7 @@ static void hrtimer_force_reprogram(struct hrtimer_cpu_base *cpu_base)
 static int hrtimer_reprogram(struct hrtimer *timer,
 			     struct hrtimer_clock_base *base)
 {
-	ktime_t *expires_next = &__get_cpu_var(hrtimer_bases).expires_next;
+	struct hrtimer_cpu_base *cpu_base = &__get_cpu_var(hrtimer_bases);
 	ktime_t expires = ktime_sub(hrtimer_get_expires(timer), base->offset);
 	int res;
 
@@ -576,7 +582,16 @@ static int hrtimer_reprogram(struct hrtimer *timer,
 	if (expires.tv64 < 0)
 		return -ETIME;
 
-	if (expires.tv64 >= expires_next->tv64)
+	if (expires.tv64 >= cpu_base->expires_next.tv64)
+		return 0;
+
+	/*
+	 * If a hang was detected in the last timer interrupt then we
+	 * do not schedule a timer which is earlier than the expiry
+	 * which we enforced in the hang detection. We want the system
+	 * to make progress.
+	 */
+	if (cpu_base->hang_detected)
 		return 0;
 
 	/*
@@ -584,7 +599,7 @@ static int hrtimer_reprogram(struct hrtimer *timer,
 	 */
 	res = tick_program_event(expires, 0);
 	if (!IS_ERR_VALUE(res))
-		*expires_next = expires;
+		cpu_base->expires_next = expires;
 	return res;
 }
 
@@ -613,12 +628,12 @@ static void retrigger_next_event(void *arg)
 	base = &__get_cpu_var(hrtimer_bases);
 
 	/* Adjust CLOCK_REALTIME offset */
-	spin_lock(&base->lock);
+	raw_spin_lock(&base->lock);
 	base->clock_base[CLOCK_REALTIME].offset =
 		timespec_to_ktime(realtime_offset);
 
-	hrtimer_force_reprogram(base);
-	spin_unlock(&base->lock);
+	hrtimer_force_reprogram(base, 0);
+	raw_spin_unlock(&base->lock);
 }
 
 /*
@@ -679,9 +694,9 @@ static inline int hrtimer_enqueue_reprogram(struct hrtimer *timer,
 {
 	if (base->cpu_base->hres_active && hrtimer_reprogram(timer, base)) {
 		if (wakeup) {
-			spin_unlock(&base->cpu_base->lock);
+			raw_spin_unlock(&base->cpu_base->lock);
 			raise_softirq_irqoff(HRTIMER_SOFTIRQ);
-			spin_lock(&base->cpu_base->lock);
+			raw_spin_lock(&base->cpu_base->lock);
 		} else
 			__raise_softirq_irqoff(HRTIMER_SOFTIRQ);
 
@@ -720,8 +735,6 @@ static int hrtimer_switch_to_hres(void)
 	/* "Retrigger" the interrupt to get things going */
 	retrigger_next_event(NULL);
 	local_irq_restore(flags);
-	printk(KERN_DEBUG "Switched to high resolution mode on CPU %d\n",
-	       smp_processor_id());
 	return 1;
 }
 
@@ -730,7 +743,8 @@ static int hrtimer_switch_to_hres(void)
 static inline int hrtimer_hres_active(void) { return 0; }
 static inline int hrtimer_is_hres_enabled(void) { return 0; }
 static inline int hrtimer_switch_to_hres(void) { return 0; }
-static inline void hrtimer_force_reprogram(struct hrtimer_cpu_base *base) { }
+static inline void
+hrtimer_force_reprogram(struct hrtimer_cpu_base *base, int skip_equal) { }
 static inline int hrtimer_enqueue_reprogram(struct hrtimer *timer,
 					    struct hrtimer_clock_base *base,
 					    int wakeup)
@@ -742,17 +756,33 @@ static inline void hrtimer_init_timer_hres(struct hrtimer *timer) { }
 
 #endif /* CONFIG_HIGH_RES_TIMERS */
 
-#ifdef CONFIG_TIMER_STATS
-void __timer_stats_hrtimer_set_start_info(struct hrtimer *timer, void *addr)
+static inline void timer_stats_hrtimer_set_start_info(struct hrtimer *timer)
 {
+#ifdef CONFIG_TIMER_STATS
 	if (timer->start_site)
 		return;
-
-	timer->start_site = addr;
+	timer->start_site = __builtin_return_address(0);
 	memcpy(timer->start_comm, current->comm, TASK_COMM_LEN);
 	timer->start_pid = current->pid;
+#endif
+}
+
+static inline void timer_stats_hrtimer_clear_start_info(struct hrtimer *timer)
+{
+#ifdef CONFIG_TIMER_STATS
+	timer->start_site = NULL;
+#endif
 }
+
+static inline void timer_stats_account_hrtimer(struct hrtimer *timer)
+{
+#ifdef CONFIG_TIMER_STATS
+	if (likely(!timer_stats_active))
+		return;
+	timer_stats_update_stats(timer, timer->start_pid, timer->start_site,
+				 timer->function, timer->start_comm, 0);
 #endif
+}
 
 /*
  * Counterpart to lock_hrtimer_base above:
@@ -760,7 +790,7 @@ void __timer_stats_hrtimer_set_start_info(struct hrtimer *timer, void *addr)
 static inline
 void unlock_hrtimer_base(const struct hrtimer *timer, unsigned long *flags)
 {
-	spin_unlock_irqrestore(&timer->base->cpu_base->lock, *flags);
+	raw_spin_unlock_irqrestore(&timer->base->cpu_base->lock, *flags);
 }
 
 /**
@@ -873,19 +903,29 @@ static void __remove_hrtimer(struct hrtimer *timer,
 			     struct hrtimer_clock_base *base,
 			     unsigned long newstate, int reprogram)
 {
-	if (timer->state & HRTIMER_STATE_ENQUEUED) {
-		/*
-		 * Remove the timer from the rbtree and replace the
-		 * first entry pointer if necessary.
-		 */
-		if (base->first == &timer->node) {
-			base->first = rb_next(&timer->node);
-			/* Reprogram the clock event device. if enabled */
-			if (reprogram && hrtimer_hres_active())
-				hrtimer_force_reprogram(base->cpu_base);
+	if (!(timer->state & HRTIMER_STATE_ENQUEUED))
+		goto out;
+
+	/*
+	 * Remove the timer from the rbtree and replace the first
+	 * entry pointer if necessary.
+	 */
+	if (base->first == &timer->node) {
+		base->first = rb_next(&timer->node);
+#ifdef CONFIG_HIGH_RES_TIMERS
+		/* Reprogram the clock event device. if enabled */
+		if (reprogram && hrtimer_hres_active()) {
+			ktime_t expires;
+
+			expires = ktime_sub(hrtimer_get_expires(timer),
+					    base->offset);
+			if (base->cpu_base->expires_next.tv64 == expires.tv64)
+				hrtimer_force_reprogram(base->cpu_base, 1);
 		}
-		rb_erase(&timer->node, &base->active);
+#endif
 	}
+	rb_erase(&timer->node, &base->active);
+out:
 	timer->state = newstate;
 }
 
@@ -1083,7 +1123,7 @@ ktime_t hrtimer_get_next_event(void)
 	unsigned long flags;
 	int i;
 
-	spin_lock_irqsave(&cpu_base->lock, flags);
+	raw_spin_lock_irqsave(&cpu_base->lock, flags);
 
 	if (!hrtimer_hres_active()) {
 		for (i = 0; i < HRTIMER_MAX_CLOCK_BASES; i++, base++) {
@@ -1100,7 +1140,7 @@ ktime_t hrtimer_get_next_event(void)
 		}
 	}
 
-	spin_unlock_irqrestore(&cpu_base->lock, flags);
+	raw_spin_unlock_irqrestore(&cpu_base->lock, flags);
 
 	if (mindelta.tv64 < 0)
 		mindelta.tv64 = 0;
@@ -1182,11 +1222,11 @@ static void __run_hrtimer(struct hrtimer *timer, ktime_t *now)
 	 * they get migrated to another cpu, therefore its safe to unlock
 	 * the timer base.
 	 */
-	spin_unlock(&cpu_base->lock);
+	raw_spin_unlock(&cpu_base->lock);
 	trace_hrtimer_expire_entry(timer, now);
 	restart = fn(timer);
 	trace_hrtimer_expire_exit(timer);
-	spin_lock(&cpu_base->lock);
+	raw_spin_lock(&cpu_base->lock);
 
 	/*
 	 * Note: We clear the CALLBACK bit after enqueue_hrtimer and
@@ -1202,29 +1242,6 @@ static void __run_hrtimer(struct hrtimer *timer, ktime_t *now)
 
 #ifdef CONFIG_HIGH_RES_TIMERS
 
-static int force_clock_reprogram;
-
-/*
- * After 5 iteration's attempts, we consider that hrtimer_interrupt()
- * is hanging, which could happen with something that slows the interrupt
- * such as the tracing. Then we force the clock reprogramming for each future
- * hrtimer interrupts to avoid infinite loops and use the min_delta_ns
- * threshold that we will overwrite.
- * The next tick event will be scheduled to 3 times we currently spend on
- * hrtimer_interrupt(). This gives a good compromise, the cpus will spend
- * 1/4 of their time to process the hrtimer interrupts. This is enough to
- * let it running without serious starvation.
- */
-
-static inline void
-hrtimer_interrupt_hanging(struct clock_event_device *dev,
-			ktime_t try_time)
-{
-	force_clock_reprogram = 1;
-	dev->min_delta_ns = (unsigned long)try_time.tv64 * 3;
-	printk(KERN_WARNING "hrtimer: interrupt too slow, "
-		"forcing clock min delta to %lu ns\n", dev->min_delta_ns);
-}
 /*
  * High resolution timer interrupt
  * Called with interrupts disabled
@@ -1233,24 +1250,18 @@ void hrtimer_interrupt(struct clock_event_device *dev)
 {
 	struct hrtimer_cpu_base *cpu_base = &__get_cpu_var(hrtimer_bases);
 	struct hrtimer_clock_base *base;
-	ktime_t expires_next, now;
-	int nr_retries = 0;
-	int i;
+	ktime_t expires_next, now, entry_time, delta;
+	int i, retries = 0;
 
 	BUG_ON(!cpu_base->hres_active);
 	cpu_base->nr_events++;
 	dev->next_event.tv64 = KTIME_MAX;
 
- retry:
-	/* 5 retries is enough to notice a hang */
-	if (!(++nr_retries % 5))
-		hrtimer_interrupt_hanging(dev, ktime_sub(ktime_get(), now));
-
-	now = ktime_get();
-
+	entry_time = now = ktime_get();
+retry:
 	expires_next.tv64 = KTIME_MAX;
 
-	spin_lock(&cpu_base->lock);
+	raw_spin_lock(&cpu_base->lock);
 	/*
 	 * We set expires_next to KTIME_MAX here with cpu_base->lock
 	 * held to prevent that a timer is enqueued in our queue via
@@ -1306,13 +1317,51 @@ void hrtimer_interrupt(struct clock_event_device *dev)
 	 * against it.
 	 */
 	cpu_base->expires_next = expires_next;
-	spin_unlock(&cpu_base->lock);
+	raw_spin_unlock(&cpu_base->lock);
 
 	/* Reprogramming necessary ? */
-	if (expires_next.tv64 != KTIME_MAX) {
-		if (tick_program_event(expires_next, force_clock_reprogram))
-			goto retry;
+	if (expires_next.tv64 == KTIME_MAX ||
+	    !tick_program_event(expires_next, 0)) {
+		cpu_base->hang_detected = 0;
+		return;
 	}
+
+	/*
+	 * The next timer was already expired due to:
+	 * - tracing
+	 * - long lasting callbacks
+	 * - being scheduled away when running in a VM
+	 *
+	 * We need to prevent that we loop forever in the hrtimer
+	 * interrupt routine. We give it 3 attempts to avoid
+	 * overreacting on some spurious event.
+	 */
+	now = ktime_get();
+	cpu_base->nr_retries++;
+	if (++retries < 3)
+		goto retry;
+	/*
+	 * Give the system a chance to do something else than looping
+	 * here. We stored the entry time, so we know exactly how long
+	 * we spent here. We schedule the next event this amount of
+	 * time away.
+	 */
+	cpu_base->nr_hangs++;
+	cpu_base->hang_detected = 1;
+	delta = ktime_sub(now, entry_time);
+	if (delta.tv64 > cpu_base->max_hang_time.tv64)
+		cpu_base->max_hang_time = delta;
+	/*
+	 * Limit it to a sensible value as we enforce a longer
+	 * delay. Give the CPU at least 100ms to catch up.
+	 */
+	if (delta.tv64 > 100 * NSEC_PER_MSEC)
+		expires_next = ktime_add_ns(now, 100 * NSEC_PER_MSEC);
+	else
+		expires_next = ktime_add(now, delta);
+	tick_program_event(expires_next, 1);
+	printk_once(KERN_WARNING "hrtimer: interrupt took %llu ns\n",
+		    ktime_to_ns(delta));
 }
 
 /*
@@ -1408,7 +1457,7 @@ void hrtimer_run_queues(void)
 			gettime = 0;
 		}
 
-		spin_lock(&cpu_base->lock);
+		raw_spin_lock(&cpu_base->lock);
 
 		while ((node = base->first)) {
 			struct hrtimer *timer;
@@ -1420,7 +1469,7 @@ void hrtimer_run_queues(void)
 
 			__run_hrtimer(timer, &base->softirq_time);
 		}
-		spin_unlock(&cpu_base->lock);
+		raw_spin_unlock(&cpu_base->lock);
 	}
 }
 
@@ -1576,7 +1625,7 @@ static void __cpuinit init_hrtimers_cpu(int cpu)
 	struct hrtimer_cpu_base *cpu_base = &per_cpu(hrtimer_bases, cpu);
 	int i;
 
-	spin_lock_init(&cpu_base->lock);
+	raw_spin_lock_init(&cpu_base->lock);
 
 	for (i = 0; i < HRTIMER_MAX_CLOCK_BASES; i++)
 		cpu_base->clock_base[i].cpu_base = cpu_base;
@@ -1634,16 +1683,16 @@ static void migrate_hrtimers(int scpu)
 	 * The caller is globally serialized and nobody else
 	 * takes two locks at once, deadlock is not possible.
 	 */
-	spin_lock(&new_base->lock);
-	spin_lock_nested(&old_base->lock, SINGLE_DEPTH_NESTING);
+	raw_spin_lock(&new_base->lock);
+	raw_spin_lock_nested(&old_base->lock, SINGLE_DEPTH_NESTING);
 
 	for (i = 0; i < HRTIMER_MAX_CLOCK_BASES; i++) {
 		migrate_hrtimer_list(&old_base->clock_base[i],
 				     &new_base->clock_base[i]);
 	}
 
-	spin_unlock(&old_base->lock);
-	spin_unlock(&new_base->lock);
+	raw_spin_unlock(&old_base->lock);
+	raw_spin_unlock(&new_base->lock);
 
 	/* Check, if we got expired work to do */
 	__hrtimer_peek_ahead_timers();