[patch 16/39] hrtimer: Get rid of hrtimer softirq

[Thomas Gleixner]

hrtimer softirq is a leftover from the initial implementation and
serves only the purpose to handle the enqueueing of already expired
timers in the high resolution timer mode. We discussed whether we
change the return value and force all start sites to handle that the
timer is already expired, but that would be a Herculean task and I'm
not sure whether its a good idea to enforce that handling on
everyone.

A simpler solution is to enforce a timer interrupt instead of raising
and scheduling a softirq. Just use the existing infrastructure to do
so and remove all the softirq leftovers.

Signed-off-by: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
---
 include/linux/hrtimer.h    |    1 
 include/linux/interrupt.h  |    1 
 include/trace/events/irq.h |    1 
 kernel/softirq.c           |    2 
 kernel/time/hrtimer.c      |  163 +++++++++++----------------------------------
 kernel/time/timer.c        |    2 
 6 files changed, 44 insertions(+), 126 deletions(-)

Index: tip/include/linux/hrtimer.h
===================================================================
--- tip.orig/include/linux/hrtimer.h
+++ tip/include/linux/hrtimer.h
@@ -444,7 +444,6 @@ extern int schedule_hrtimeout(ktime_t *e
 
 /* Soft interrupt function to run the hrtimer queues: */
 extern void hrtimer_run_queues(void);
-extern void hrtimer_run_pending(void);
 
 /* Bootup initialization: */
 extern void __init hrtimers_init(void);
Index: tip/include/linux/interrupt.h
===================================================================
--- tip.orig/include/linux/interrupt.h
+++ tip/include/linux/interrupt.h
@@ -401,7 +401,6 @@ enum
 	BLOCK_IOPOLL_SOFTIRQ,
 	TASKLET_SOFTIRQ,
 	SCHED_SOFTIRQ,
-	HRTIMER_SOFTIRQ,
 	RCU_SOFTIRQ,    /* Preferable RCU should always be the last softirq */
 
 	NR_SOFTIRQS
Index: tip/include/trace/events/irq.h
===================================================================
--- tip.orig/include/trace/events/irq.h
+++ tip/include/trace/events/irq.h
@@ -20,7 +20,6 @@ struct softirq_action;
 			 softirq_name(BLOCK_IOPOLL),	\
 			 softirq_name(TASKLET),		\
 			 softirq_name(SCHED),		\
-			 softirq_name(HRTIMER),		\
 			 softirq_name(RCU))
 
 /**
Index: tip/kernel/softirq.c
===================================================================
--- tip.orig/kernel/softirq.c
+++ tip/kernel/softirq.c
@@ -59,7 +59,7 @@ DEFINE_PER_CPU(struct task_struct *, kso
 
 const char * const softirq_to_name[NR_SOFTIRQS] = {
 	"HI", "TIMER", "NET_TX", "NET_RX", "BLOCK", "BLOCK_IOPOLL",
-	"TASKLET", "SCHED", "HRTIMER", "RCU"
+	"TASKLET", "SCHED", "RCU"
 };
 
 /*
Index: tip/kernel/time/hrtimer.c
===================================================================
--- tip.orig/kernel/time/hrtimer.c
+++ tip/kernel/time/hrtimer.c
@@ -555,59 +555,48 @@ hrtimer_force_reprogram(struct hrtimer_c
 }
 
 /*
- * Shared reprogramming for clock_realtime and clock_monotonic
- *
  * When a timer is enqueued and expires earlier than the already enqueued
  * timers, we have to check, whether it expires earlier than the timer for
  * which the clock event device was armed.
  *
- * Note, that in case the state has HRTIMER_STATE_CALLBACK set, no reprogramming
- * and no expiry check happens. The timer gets enqueued into the rbtree. The
- * reprogramming and expiry check is done in the hrtimer_interrupt or in the
- * softirq.
- *
  * Called with interrupts disabled and base->cpu_base.lock held
  */
-static int hrtimer_reprogram(struct hrtimer *timer,
-			     struct hrtimer_clock_base *base)
+static void hrtimer_reprogram(struct hrtimer *timer,
+			      struct hrtimer_clock_base *base)
 {
 	struct hrtimer_cpu_base *cpu_base = this_cpu_ptr(&hrtimer_bases);
 	ktime_t expires = ktime_sub(hrtimer_get_expires(timer), base->offset);
-	int res;
 
 	WARN_ON_ONCE(hrtimer_get_expires_tv64(timer) < 0);
 
 	/*
-	 * When the callback is running, we do not reprogram the clock event
-	 * device. The timer callback is either running on a different CPU or
-	 * the callback is executed in the hrtimer_interrupt context. The
-	 * reprogramming is handled either by the softirq, which called the
-	 * callback or at the end of the hrtimer_interrupt.
+	 * If the timer is not on the current cpu, we cannot reprogram
+	 * the other cpus clock event device.
 	 */
-	if (hrtimer_callback_running(timer))
-		return 0;
+	if (base->cpu_base != cpu_base)
+		return;
+
+	/*
+	 * If the hrtimer interrupt is running, then it will
+	 * reevaluate the clock bases and reprogram the clock event
+	 * device. The callbacks are always executed in hard interrupt
+	 * context so we don't need an extra check for a running
+	 * callback.
+	 */
+	if (cpu_base->in_hrtirq)
+		return;
 
 	/*
 	 * CLOCK_REALTIME timer might be requested with an absolute
-	 * expiry time which is less than base->offset. Nothing wrong
-	 * about that, just avoid to call into the tick code, which
-	 * has now objections against negative expiry values.
+	 * expiry time which is less than base->offset. Set it to 0.
 	 */
 	if (expires.tv64 < 0)
-		return -ETIME;
+		expires.tv64 = 0;
 
 	if (expires.tv64 >= cpu_base->expires_next.tv64)
-		return 0;
-
-	/*
-	 * When the target cpu of the timer is currently executing
-	 * hrtimer_interrupt(), then we do not touch the clock event
-	 * device. hrtimer_interrupt() will reevaluate all clock bases
-	 * before reprogramming the device.
-	 */
-	if (cpu_base->in_hrtirq)
-		return 0;
+		return;
 
+	/* Update the pointer to the next expiring timer */
 	cpu_base->next_timer = timer;
 
 	/*
@@ -617,15 +606,14 @@ static int hrtimer_reprogram(struct hrti
 	 * to make progress.
 	 */
 	if (cpu_base->hang_detected)
-		return 0;
+		return;
 
 	/*
-	 * Clockevents returns -ETIME, when the event was in the past.
+	 * Program the timer hardware. We enforce the expiry for
+	 * events which are already in the past.
 	 */
-	res = tick_program_event(expires, 0);
-	if (!IS_ERR_VALUE(res))
-		cpu_base->expires_next = expires;
-	return res;
+	cpu_base->expires_next = expires;
+	tick_program_event(expires, 1);
 }
 
 /*
@@ -660,19 +648,11 @@ static void retrigger_next_event(void *a
  */
 static int hrtimer_switch_to_hres(void)
 {
-	int cpu = smp_processor_id();
-	struct hrtimer_cpu_base *base = &per_cpu(hrtimer_bases, cpu);
-	unsigned long flags;
-
-	if (base->hres_active)
-		return 1;
-
-	local_irq_save(flags);
+	struct hrtimer_cpu_base *base = this_cpu_ptr(&hrtimer_bases);
 
 	if (tick_init_highres()) {
-		local_irq_restore(flags);
 		printk(KERN_WARNING "Could not switch to high resolution "
-				    "mode on CPU %d\n", cpu);
+				    "mode on CPU %d\n", base->cpu);
 		return 0;
 	}
 	base->hres_active = 1;
@@ -681,7 +661,6 @@ static int hrtimer_switch_to_hres(void)
 	tick_setup_sched_timer();
 	/* "Retrigger" the interrupt to get things going */
 	retrigger_next_event(NULL);
-	local_irq_restore(flags);
 	return 1;
 }
 
@@ -976,26 +955,8 @@ int __hrtimer_start_range_ns(struct hrti
 		 * on dynticks target.
 		 */
 		wake_up_nohz_cpu(new_base->cpu_base->cpu);
-	} else if (new_base->cpu_base == this_cpu_ptr(&hrtimer_bases) &&
-			hrtimer_reprogram(timer, new_base)) {
-		/*
-		 * Only allow reprogramming if the new base is on this CPU.
-		 * (it might still be on another CPU if the timer was pending)
-		 *
-		 * XXX send_remote_softirq() ?
-		 */
-		if (wakeup) {
-			/*
-			 * We need to drop cpu_base->lock to avoid a
-			 * lock ordering issue vs. rq->lock.
-			 */
-			raw_spin_unlock(&new_base->cpu_base->lock);
-			raise_softirq_irqoff(HRTIMER_SOFTIRQ);
-			local_irq_restore(flags);
-			return ret;
-		} else {
-			__raise_softirq_irqoff(HRTIMER_SOFTIRQ);
-		}
+	} else {
+		hrtimer_reprogram(timer, new_base);
 	}
 
 	unlock_hrtimer_base(timer, &flags);
@@ -1346,7 +1307,7 @@ retry:
  * local version of hrtimer_peek_ahead_timers() called with interrupts
  * disabled.
  */
-static void __hrtimer_peek_ahead_timers(void)
+static inline void __hrtimer_peek_ahead_timers(void)
 {
 	struct tick_device *td;
 
@@ -1358,29 +1319,6 @@ static void __hrtimer_peek_ahead_timers(
 		hrtimer_interrupt(td->evtdev);
 }
 
-/**
- * hrtimer_peek_ahead_timers -- run soft-expired timers now
- *
- * hrtimer_peek_ahead_timers will peek at the timer queue of
- * the current cpu and check if there are any timers for which
- * the soft expires time has passed. If any such timers exist,
- * they are run immediately and then removed from the timer queue.
- *
- */
-void hrtimer_peek_ahead_timers(void)
-{
-	unsigned long flags;
-
-	local_irq_save(flags);
-	__hrtimer_peek_ahead_timers();
-	local_irq_restore(flags);
-}
-
-static void run_hrtimer_softirq(struct softirq_action *h)
-{
-	hrtimer_peek_ahead_timers();
-}
-
 #else /* CONFIG_HIGH_RES_TIMERS */
 
 static inline void __hrtimer_peek_ahead_timers(void) { }
@@ -1388,31 +1326,7 @@ static inline void __hrtimer_peek_ahead_
 #endif	/* !CONFIG_HIGH_RES_TIMERS */
 
 /*
- * Called from timer softirq every jiffy, expire hrtimers:
- *
- * For HRT its the fall back code to run the softirq in the timer
- * softirq context in case the hrtimer initialization failed or has
- * not been done yet.
- */
-void hrtimer_run_pending(void)
-{
-	if (hrtimer_hres_active())
-		return;
-
-	/*
-	 * This _is_ ugly: We have to check in the softirq context,
-	 * whether we can switch to highres and / or nohz mode. The
-	 * clocksource switch happens in the timer interrupt with
-	 * xtime_lock held. Notification from there only sets the
-	 * check bit in the tick_oneshot code, otherwise we might
-	 * deadlock vs. xtime_lock.
-	 */
-	if (tick_check_oneshot_change(!hrtimer_is_hres_enabled()))
-		hrtimer_switch_to_hres();
-}
-
-/*
- * Called from hardirq context every jiffy
+ * Called from run_local_timers in hardirq context every jiffy
  */
 void hrtimer_run_queues(void)
 {
@@ -1422,6 +1336,18 @@ void hrtimer_run_queues(void)
 	if (__hrtimer_hres_active(cpu_base))
 		return;
 
+	/*
+	 * This _is_ ugly: We have to check periodically, whether we
+	 * can switch to highres and / or nohz mode. The clocksource
+	 * switch happens with xtime_lock held. Notification from
+	 * there only sets the check bit in the tick_oneshot code,
+	 * otherwise we might deadlock vs. xtime_lock.
+	 */
+	if (tick_check_oneshot_change(!hrtimer_is_hres_enabled())) {
+		hrtimer_switch_to_hres();
+		return;
+	}
+
 	raw_spin_lock(&cpu_base->lock);
 	now = hrtimer_update_base(cpu_base);
 	__hrtimer_run_queues(cpu_base, now);
@@ -1692,9 +1618,6 @@ void __init hrtimers_init(void)
 	hrtimer_cpu_notify(&hrtimers_nb, (unsigned long)CPU_UP_PREPARE,
 			  (void *)(long)smp_processor_id());
 	register_cpu_notifier(&hrtimers_nb);
-#ifdef CONFIG_HIGH_RES_TIMERS
-	open_softirq(HRTIMER_SOFTIRQ, run_hrtimer_softirq);
-#endif
 }
 
 /**
Index: tip/kernel/time/timer.c
===================================================================
--- tip.orig/kernel/time/timer.c
+++ tip/kernel/time/timer.c
@@ -1409,8 +1409,6 @@ static void run_timer_softirq(struct sof
 {
 	struct tvec_base *base = __this_cpu_read(tvec_bases);
 
-	hrtimer_run_pending();
-
 	if (time_after_eq(jiffies, base->timer_jiffies))
 		__run_timers(base);
 }


--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/