hrtimers: Avoid touching inactive timer bases

Instead of iterating over all possible timer bases avoid it by marking
the active bases in the cpu base.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Peter Zijlstra <peterz@infradead.org>

6 files changed, 29 insertions, 19 deletions

diff --git a/include/linux/hrtimer.h b/include/linux/hrtimer.h
index cc5f5f51db10..771c95802edc 100644
--- a/include/linux/hrtimer.h
+++ b/include/linux/hrtimer.h
@@ -143,7 +143,8 @@ struct hrtimer_sleeper {
  */
 struct hrtimer_clock_base {
 	struct hrtimer_cpu_base	*cpu_base;
-	clockid_t		index;
+	int			index;
+	clockid_t		clockid;
 	struct timerqueue_head	active;
 	ktime_t			resolution;
 	ktime_t			(*get_time)(void);
@@ -162,7 +163,7 @@ enum  hrtimer_base_type {
  * struct hrtimer_cpu_base - the per cpu clock bases
  * @lock:		lock protecting the base and associated clock bases
  *			and timers
- * @clock_base:		array of clock bases for this cpu
+ * @active_bases:	Bitfield to mark bases with active timers
  * @expires_next:	absolute time of the next event which was scheduled
  *			via clock_set_next_event()
  * @hres_active:	State of high resolution mode
@@ -171,9 +172,11 @@ enum  hrtimer_base_type {
  * @nr_retries:		Total number of hrtimer interrupt retries
  * @nr_hangs:		Total number of hrtimer interrupt hangs
  * @max_hang_time:	Maximum time spent in hrtimer_interrupt
+ * @clock_base:		array of clock bases for this cpu
  */
 struct hrtimer_cpu_base {
 	raw_spinlock_t			lock;
+	unsigned long			active_bases;
 #ifdef CONFIG_HIGH_RES_TIMERS
 	ktime_t				expires_next;
 	int				hres_active;
diff --git a/include/linux/thread_info.h b/include/linux/thread_info.h
index 20fc303947d3..8d03f079688c 100644
--- a/include/linux/thread_info.h
+++ b/include/linux/thread_info.h
@@ -29,7 +29,7 @@ struct restart_block {
 		} futex;
 		/* For nanosleep */
 		struct {
-			clockid_t index;
+			clockid_t clockid;
 			struct timespec __user *rmtp;
 #ifdef CONFIG_COMPAT
 			struct compat_timespec __user *compat_rmtp;
diff --git a/kernel/hrtimer.c b/kernel/hrtimer.c
index 26dd32f9f6b2..1b08f6d67f12 100644
--- a/kernel/hrtimer.c
+++ b/kernel/hrtimer.c
@@ -64,17 +64,20 @@ DEFINE_PER_CPU(struct hrtimer_cpu_base, hrtimer_bases) =
 	.clock_base =
 	{
 		{
-			.index = CLOCK_REALTIME,
+			.index = HRTIMER_BASE_REALTIME,
+			.clockid = CLOCK_REALTIME,
 			.get_time = &ktime_get_real,
 			.resolution = KTIME_LOW_RES,
 		},
 		{
-			.index = CLOCK_MONOTONIC,
+			.index = HRTIMER_BASE_MONOTONIC,
+			.clockid = CLOCK_MONOTONIC,
 			.get_time = &ktime_get,
 			.resolution = KTIME_LOW_RES,
 		},
 		{
-			.index = CLOCK_BOOTTIME,
+			.index = HRTIMER_BASE_BOOTTIME,
+			.clockid = CLOCK_BOOTTIME,
 			.get_time = &ktime_get_boottime,
 			.resolution = KTIME_LOW_RES,
 		},
@@ -196,7 +199,7 @@ switch_hrtimer_base(struct hrtimer *timer, struct hrtimer_clock_base *base,
 	struct hrtimer_cpu_base *new_cpu_base;
 	int this_cpu = smp_processor_id();
 	int cpu = hrtimer_get_target(this_cpu, pinned);
-	int basenum = hrtimer_clockid_to_base(base->index);
+	int basenum = base->index;
 
 again:
 	new_cpu_base = &per_cpu(hrtimer_bases, cpu);
@@ -857,6 +860,7 @@ static int enqueue_hrtimer(struct hrtimer *timer,
 	debug_activate(timer);
 
 	timerqueue_add(&base->active, &timer->node);
+	base->cpu_base->active_bases |= 1 << base->index;
 
 	/*
 	 * HRTIMER_STATE_ENQUEUED is or'ed to the current state to preserve the
@@ -898,6 +902,8 @@ static void __remove_hrtimer(struct hrtimer *timer,
 #endif
 	}
 	timerqueue_del(&base->active, &timer->node);
+	if (!timerqueue_getnext(&base->active))
+		base->cpu_base->active_bases &= ~(1 << base->index);
 out:
 	timer->state = newstate;
 }
@@ -1235,7 +1241,6 @@ static void __run_hrtimer(struct hrtimer *timer, ktime_t *now)
 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, entry_time, delta;
 	int i, retries = 0;
 
@@ -1257,12 +1262,15 @@ retry:
 	 */
 	cpu_base->expires_next.tv64 = KTIME_MAX;
 
-	base = cpu_base->clock_base;
-
 	for (i = 0; i < HRTIMER_MAX_CLOCK_BASES; i++) {
-		ktime_t basenow;
+		struct hrtimer_clock_base *base;
 		struct timerqueue_node *node;
+		ktime_t basenow;
+
+		if (!(cpu_base->active_bases & (1 << i)))
+			continue;
 
+		base = cpu_base->clock_base + i;
 		basenow = ktime_add(now, base->offset);
 
 		while ((node = timerqueue_getnext(&base->active))) {
@@ -1295,7 +1303,6 @@ retry:
 
 			__run_hrtimer(timer, &basenow);
 		}
-		base++;
 	}
 
 	/*
@@ -1526,7 +1533,7 @@ long __sched hrtimer_nanosleep_restart(struct restart_block *restart)
 	struct timespec __user  *rmtp;
 	int ret = 0;
 
-	hrtimer_init_on_stack(&t.timer, restart->nanosleep.index,
+	hrtimer_init_on_stack(&t.timer, restart->nanosleep.clockid,
 				HRTIMER_MODE_ABS);
 	hrtimer_set_expires_tv64(&t.timer, restart->nanosleep.expires);
 
@@ -1578,7 +1585,7 @@ long hrtimer_nanosleep(struct timespec *rqtp, struct timespec __user *rmtp,
 
 	restart = &current_thread_info()->restart_block;
 	restart->fn = hrtimer_nanosleep_restart;
-	restart->nanosleep.index = t.timer.base->index;
+	restart->nanosleep.clockid = t.timer.base->clockid;
 	restart->nanosleep.rmtp = rmtp;
 	restart->nanosleep.expires = hrtimer_get_expires_tv64(&t.timer);
 
diff --git a/kernel/posix-cpu-timers.c b/kernel/posix-cpu-timers.c
index 0791b13df7bf..58f405b581e7 100644
--- a/kernel/posix-cpu-timers.c
+++ b/kernel/posix-cpu-timers.c
@@ -1514,7 +1514,7 @@ static int posix_cpu_nsleep(const clockid_t which_clock, int flags,
 			return -EFAULT;
 
 		restart_block->fn = posix_cpu_nsleep_restart;
-		restart_block->nanosleep.index = which_clock;
+		restart_block->nanosleep.clockid = which_clock;
 		restart_block->nanosleep.rmtp = rmtp;
 		restart_block->nanosleep.expires = timespec_to_ns(rqtp);
 	}
@@ -1523,7 +1523,7 @@ static int posix_cpu_nsleep(const clockid_t which_clock, int flags,
 
 static long posix_cpu_nsleep_restart(struct restart_block *restart_block)
 {
-	clockid_t which_clock = restart_block->nanosleep.index;
+	clockid_t which_clock = restart_block->nanosleep.clockid;
 	struct timespec t;
 	struct itimerspec it;
 	int error;
diff --git a/kernel/posix-timers.c b/kernel/posix-timers.c
index e5498d7405c3..a1b5edf1bf92 100644
--- a/kernel/posix-timers.c
+++ b/kernel/posix-timers.c
@@ -1056,7 +1056,7 @@ SYSCALL_DEFINE4(clock_nanosleep, const clockid_t, which_clock, int, flags,
  */
 long clock_nanosleep_restart(struct restart_block *restart_block)
 {
-	clockid_t which_clock = restart_block->nanosleep.index;
+	clockid_t which_clock = restart_block->nanosleep.clockid;
 	struct k_clock *kc = clockid_to_kclock(which_clock);
 
 	if (WARN_ON_ONCE(!kc || !kc->nsleep_restart))
diff --git a/kernel/time/alarmtimer.c b/kernel/time/alarmtimer.c
index c6027fe9a4e6..2d966244ea60 100644
--- a/kernel/time/alarmtimer.c
+++ b/kernel/time/alarmtimer.c
@@ -494,7 +494,7 @@ static int update_rmtp(ktime_t exp, enum  alarmtimer_type type,
  */
 static long __sched alarm_timer_nsleep_restart(struct restart_block *restart)
 {
-	enum  alarmtimer_type type = restart->nanosleep.index;
+	enum  alarmtimer_type type = restart->nanosleep.clockid;
 	ktime_t exp;
 	struct timespec __user  *rmtp;
 	struct alarm alarm;
@@ -573,7 +573,7 @@ static int alarm_timer_nsleep(const clockid_t which_clock, int flags,
 
 	restart = &current_thread_info()->restart_block;
 	restart->fn = alarm_timer_nsleep_restart;
-	restart->nanosleep.index = type;
+	restart->nanosleep.clockid = type;
 	restart->nanosleep.expires = exp.tv64;
 	restart->nanosleep.rmtp = rmtp;
 	ret = -ERESTART_RESTARTBLOCK;