1 files changed, 99 insertions, 43 deletions

diff --git a/kernel/time/tick-sched.c b/kernel/time/tick-sched.c
index 5eb04bb59802..c7b9b1e4af7d 100644
--- a/kernel/time/tick-sched.c
+++ b/kernel/time/tick-sched.c
@@ -53,48 +53,69 @@ static ktime_t last_jiffies_update;
  */
 static void tick_do_update_jiffies64(ktime_t now)
 {
-	unsigned long ticks = 0;
+	unsigned long ticks = 1;
 	ktime_t delta;
 
 	/*
-	 * Do a quick check without holding jiffies_lock:
-	 * The READ_ONCE() pairs with two updates done later in this function.
+	 * Do a quick check without holding jiffies_lock. The READ_ONCE()
+	 * pairs with the update done later in this function.
+	 *
+	 * This is also an intentional data race which is even safe on
+	 * 32bit in theory. If there is a concurrent update then the check
+	 * might give a random answer. It does not matter because if it
+	 * returns then the concurrent update is already taking care, if it
+	 * falls through then it will pointlessly contend on jiffies_lock.
+	 *
+	 * Though there is one nasty case on 32bit due to store tearing of
+	 * the 64bit value. If the first 32bit store makes the quick check
+	 * return on all other CPUs and the writing CPU context gets
+	 * delayed to complete the second store (scheduled out on virt)
+	 * then jiffies can become stale for up to ~2^32 nanoseconds
+	 * without noticing. After that point all CPUs will wait for
+	 * jiffies lock.
+	 *
+	 * OTOH, this is not any different than the situation with NOHZ=off
+	 * where one CPU is responsible for updating jiffies and
+	 * timekeeping. If that CPU goes out for lunch then all other CPUs
+	 * will operate on stale jiffies until it decides to come back.
 	 */
-	delta = ktime_sub(now, READ_ONCE(last_jiffies_update));
-	if (delta < tick_period)
+	if (ktime_before(now, READ_ONCE(tick_next_period)))
 		return;
 
 	/* Reevaluate with jiffies_lock held */
-	write_seqlock(&jiffies_lock);
-
-	delta = ktime_sub(now, last_jiffies_update);
-	if (delta >= tick_period) {
-
-		delta = ktime_sub(delta, tick_period);
-		/* Pairs with the lockless read in this function. */
-		WRITE_ONCE(last_jiffies_update,
-			   ktime_add(last_jiffies_update, tick_period));
+	raw_spin_lock(&jiffies_lock);
+	if (ktime_before(now, tick_next_period)) {
+		raw_spin_unlock(&jiffies_lock);
+		return;
+	}
 
-		/* Slow path for long timeouts */
-		if (unlikely(delta >= tick_period)) {
-			s64 incr = ktime_to_ns(tick_period);
+	write_seqcount_begin(&jiffies_seq);
 
-			ticks = ktime_divns(delta, incr);
+	delta = ktime_sub(now, tick_next_period);
+	if (unlikely(delta >= TICK_NSEC)) {
+		/* Slow path for long idle sleep times */
+		s64 incr = TICK_NSEC;
 
-			/* Pairs with the lockless read in this function. */
-			WRITE_ONCE(last_jiffies_update,
-				   ktime_add_ns(last_jiffies_update,
-						incr * ticks));
-		}
-		do_timer(++ticks);
+		ticks += ktime_divns(delta, incr);
 
-		/* Keep the tick_next_period variable up to date */
-		tick_next_period = ktime_add(last_jiffies_update, tick_period);
+		last_jiffies_update = ktime_add_ns(last_jiffies_update,
+						   incr * ticks);
 	} else {
-		write_sequnlock(&jiffies_lock);
-		return;
+		last_jiffies_update = ktime_add_ns(last_jiffies_update,
+						   TICK_NSEC);
 	}
-	write_sequnlock(&jiffies_lock);
+
+	do_timer(ticks);
+
+	/*
+	 * Keep the tick_next_period variable up to date.  WRITE_ONCE()
+	 * pairs with the READ_ONCE() in the lockless quick check above.
+	 */
+	WRITE_ONCE(tick_next_period,
+		   ktime_add_ns(last_jiffies_update, TICK_NSEC));
+
+	write_seqcount_end(&jiffies_seq);
+	raw_spin_unlock(&jiffies_lock);
 	update_wall_time();
 }
 
@@ -105,12 +126,25 @@ static ktime_t tick_init_jiffy_update(void)
 {
 	ktime_t period;
 
-	write_seqlock(&jiffies_lock);
+	raw_spin_lock(&jiffies_lock);
+	write_seqcount_begin(&jiffies_seq);
 	/* Did we start the jiffies update yet ? */
-	if (last_jiffies_update == 0)
+	if (last_jiffies_update == 0) {
+		u32 rem;
+
+		/*
+		 * Ensure that the tick is aligned to a multiple of
+		 * TICK_NSEC.
+		 */
+		div_u64_rem(tick_next_period, TICK_NSEC, &rem);
+		if (rem)
+			tick_next_period += TICK_NSEC - rem;
+
 		last_jiffies_update = tick_next_period;
+	}
 	period = last_jiffies_update;
-	write_sequnlock(&jiffies_lock);
+	write_seqcount_end(&jiffies_seq);
+	raw_spin_unlock(&jiffies_lock);
 	return period;
 }
 
@@ -202,6 +236,11 @@ static bool check_tick_dependency(atomic_t *dep)
 		return true;
 	}
 
+	if (val & TICK_DEP_MASK_RCU) {
+		trace_tick_stop(0, TICK_DEP_MASK_RCU);
+		return true;
+	}
+
 	return false;
 }
 
@@ -328,6 +367,7 @@ void tick_nohz_dep_set_cpu(int cpu, enum tick_dep_bits bit)
 		preempt_enable();
 	}
 }
+EXPORT_SYMBOL_GPL(tick_nohz_dep_set_cpu);
 
 void tick_nohz_dep_clear_cpu(int cpu, enum tick_dep_bits bit)
 {
@@ -335,6 +375,7 @@ void tick_nohz_dep_clear_cpu(int cpu, enum tick_dep_bits bit)
 
 	atomic_andnot(BIT(bit), &ts->tick_dep_mask);
 }
+EXPORT_SYMBOL_GPL(tick_nohz_dep_clear_cpu);
 
 /*
  * Set a per-task tick dependency. Posix CPU timers need this in order to elapse
@@ -402,7 +443,7 @@ void __init tick_nohz_full_setup(cpumask_var_t cpumask)
 	tick_nohz_full_running = true;
 }
 
-static int tick_nohz_cpu_down(unsigned int cpu)
+bool tick_nohz_cpu_hotpluggable(unsigned int cpu)
 {
 	/*
 	 * The tick_do_timer_cpu CPU handles housekeeping duty (unbound
@@ -410,8 +451,13 @@ static int tick_nohz_cpu_down(unsigned int cpu)
 	 * CPUs. It must remain online when nohz full is enabled.
 	 */
 	if (tick_nohz_full_running && tick_do_timer_cpu == cpu)
-		return -EBUSY;
-	return 0;
+		return false;
+	return true;
+}
+
+static int tick_nohz_cpu_down(unsigned int cpu)
+{
+	return tick_nohz_cpu_hotpluggable(cpu) ? 0 : -EBUSY;
 }
 
 void __init tick_nohz_init(void)
@@ -636,7 +682,7 @@ static void tick_nohz_restart(struct tick_sched *ts, ktime_t now)
 	hrtimer_set_expires(&ts->sched_timer, ts->last_tick);
 
 	/* Forward the time to expire in the future */
-	hrtimer_forward(&ts->sched_timer, now, tick_period);
+	hrtimer_forward(&ts->sched_timer, now, TICK_NSEC);
 
 	if (ts->nohz_mode == NOHZ_MODE_HIGHRES) {
 		hrtimer_start_expires(&ts->sched_timer,
@@ -665,10 +711,10 @@ static ktime_t tick_nohz_next_event(struct tick_sched *ts, int cpu)
 
 	/* Read jiffies and the time when jiffies were updated last */
 	do {
-		seq = read_seqbegin(&jiffies_lock);
+		seq = read_seqcount_begin(&jiffies_seq);
 		basemono = last_jiffies_update;
 		basejiff = jiffies;
-	} while (read_seqretry(&jiffies_lock, seq));
+	} while (read_seqcount_retry(&jiffies_seq, seq));
 	ts->last_jiffies = basejiff;
 	ts->timer_expires_base = basemono;
 
@@ -1198,7 +1244,7 @@ static void tick_nohz_handler(struct clock_event_device *dev)
 	if (unlikely(ts->tick_stopped))
 		return;
 
-	hrtimer_forward(&ts->sched_timer, now, tick_period);
+	hrtimer_forward(&ts->sched_timer, now, TICK_NSEC);
 	tick_program_event(hrtimer_get_expires(&ts->sched_timer), 1);
 }
 
@@ -1235,7 +1281,7 @@ static void tick_nohz_switch_to_nohz(void)
 	next = tick_init_jiffy_update();
 
 	hrtimer_set_expires(&ts->sched_timer, next);
-	hrtimer_forward_now(&ts->sched_timer, tick_period);
+	hrtimer_forward_now(&ts->sched_timer, TICK_NSEC);
 	tick_program_event(hrtimer_get_expires(&ts->sched_timer), 1);
 	tick_nohz_activate(ts, NOHZ_MODE_LOWRES);
 }
@@ -1301,7 +1347,7 @@ static enum hrtimer_restart tick_sched_timer(struct hrtimer *timer)
 	if (unlikely(ts->tick_stopped))
 		return HRTIMER_NORESTART;
 
-	hrtimer_forward(timer, now, tick_period);
+	hrtimer_forward(timer, now, TICK_NSEC);
 
 	return HRTIMER_RESTART;
 }
@@ -1335,13 +1381,13 @@ void tick_setup_sched_timer(void)
 
 	/* Offset the tick to avert jiffies_lock contention. */
 	if (sched_skew_tick) {
-		u64 offset = ktime_to_ns(tick_period) >> 1;
+		u64 offset = TICK_NSEC >> 1;
 		do_div(offset, num_possible_cpus());
 		offset *= smp_processor_id();
 		hrtimer_add_expires_ns(&ts->sched_timer, offset);
 	}
 
-	hrtimer_forward(&ts->sched_timer, now, tick_period);
+	hrtimer_forward(&ts->sched_timer, now, TICK_NSEC);
 	hrtimer_start_expires(&ts->sched_timer, HRTIMER_MODE_ABS_PINNED_HARD);
 	tick_nohz_activate(ts, NOHZ_MODE_HIGHRES);
 }
@@ -1351,13 +1397,23 @@ void tick_setup_sched_timer(void)
 void tick_cancel_sched_timer(int cpu)
 {
 	struct tick_sched *ts = &per_cpu(tick_cpu_sched, cpu);
+	ktime_t idle_sleeptime, iowait_sleeptime;
+	unsigned long idle_calls, idle_sleeps;
 
 # ifdef CONFIG_HIGH_RES_TIMERS
 	if (ts->sched_timer.base)
 		hrtimer_cancel(&ts->sched_timer);
 # endif
 
+	idle_sleeptime = ts->idle_sleeptime;
+	iowait_sleeptime = ts->iowait_sleeptime;
+	idle_calls = ts->idle_calls;
+	idle_sleeps = ts->idle_sleeps;
 	memset(ts, 0, sizeof(*ts));
+	ts->idle_sleeptime = idle_sleeptime;
+	ts->iowait_sleeptime = iowait_sleeptime;
+	ts->idle_calls = idle_calls;
+	ts->idle_sleeps = idle_sleeps;
 }
 #endif