tick-sched.c

Kernel code of linux kernel

/**
 * tick_nohz_get_sleep_length - return the length of the current sleep
 *
 * Called from power state control code with interrupts disabled
 */
ktime_t tick_nohz_get_sleep_length(void)
{
	struct tick_sched *ts = &__get_cpu_var(tick_cpu_sched);

	return ts->sleep_length;
}

/**
 * tick_nohz_restart_sched_tick - restart the idle tick from the idle task
 *
 * Restart the idle tick when the CPU is woken up from idle
 */
void tick_nohz_restart_sched_tick(void)
{
	int cpu = smp_processor_id();
	struct tick_sched *ts = &per_cpu(tick_cpu_sched, cpu);
	unsigned long ticks;
	ktime_t now;

	local_irq_disable();
	tick_nohz_stop_idle(cpu);

	if (!ts->inidle || !ts->tick_stopped) {
		ts->inidle = 0;
		local_irq_enable();
		return;
	}

	ts->inidle = 0;

	rcu_exit_nohz();

	/* Update jiffies first */
	select_nohz_load_balancer(0);
	now = ktime_get();
	tick_do_update_jiffies64(now);
	cpu_clear(cpu, nohz_cpu_mask);

	/*
	 * We stopped the tick in idle. Update process times would miss the
	 * time we slept as update_process_times does only a 1 tick
	 * accounting. Enforce that this is accounted to idle !
	 */
	ticks = jiffies - ts->idle_jiffies;
	/*
	 * We might be one off. Do not randomly account a huge number of ticks!
	 */
	if (ticks && ticks < LONG_MAX) {
		add_preempt_count(HARDIRQ_OFFSET);
		account_system_time(current, HARDIRQ_OFFSET,
				    jiffies_to_cputime(ticks));
		sub_preempt_count(HARDIRQ_OFFSET);
	}

	touch_softlockup_watchdog();
	/*
	 * Cancel the scheduled timer and restore the tick
	 */
	ts->tick_stopped  = 0;
	ts->idle_exittime = now;
	hrtimer_cancel(&ts->sched_timer);
	ts->sched_timer.expires = ts->idle_tick;

	while (1) {
		/* Forward the time to expire in the future */
		hrtimer_forward(&ts->sched_timer, now, tick_period);

		if (ts->nohz_mode == NOHZ_MODE_HIGHRES) {
			hrtimer_start(&ts->sched_timer,
				      ts->sched_timer.expires,
				      HRTIMER_MODE_ABS);
			/* Check, if the timer was already in the past */
			if (hrtimer_active(&ts->sched_timer))
				break;
		} else {
			if (!tick_program_event(ts->sched_timer.expires, 0))
				break;
		}
		/* Update jiffies and reread time */
		tick_do_update_jiffies64(now);
		now = ktime_get();
	}
	local_irq_enable();
}

static int tick_nohz_reprogram(struct tick_sched *ts, ktime_t now)
{
	hrtimer_forward(&ts->sched_timer, now, tick_period);
	return tick_program_event(ts->sched_timer.expires, 0);
}

/*
 * The nohz low res interrupt handler
 */
static void tick_nohz_handler(struct clock_event_device *dev)
{
	struct tick_sched *ts = &__get_cpu_var(tick_cpu_sched);
	struct pt_regs *regs = get_irq_regs();
	int cpu = smp_processor_id();
	ktime_t now = ktime_get();

	dev->next_event.tv64 = KTIME_MAX;

	/*
	 * Check if the do_timer duty was dropped. We don't care about
	 * concurrency: This happens only when the cpu in charge went
	 * into a long sleep. If two cpus happen to assign themself to
	 * this duty, then the jiffies update is still serialized by
	 * xtime_lock.
	 */
	if (unlikely(tick_do_timer_cpu == TICK_DO_TIMER_NONE))
		tick_do_timer_cpu = cpu;

	/* Check, if the jiffies need an update */
	if (tick_do_timer_cpu == cpu)
		tick_do_update_jiffies64(now);

	/*
	 * When we are idle and the tick is stopped, we have to touch
	 * the watchdog as we might not schedule for a really long
	 * time. This happens on complete idle SMP systems while
	 * waiting on the login prompt. We also increment the "start
	 * of idle" jiffy stamp so the idle accounting adjustment we
	 * do when we go busy again does not account too much ticks.
	 */
	if (ts->tick_stopped) {
		touch_softlockup_watchdog();
		ts->idle_jiffies++;
	}

	update_process_times(user_mode(regs));
	profile_tick(CPU_PROFILING);

	/* Do not restart, when we are in the idle loop */
	if (ts->tick_stopped)
		return;

	while (tick_nohz_reprogram(ts, now)) {
		now = ktime_get();
		tick_do_update_jiffies64(now);
	}
}

/**
 * tick_nohz_switch_to_nohz - switch to nohz mode
 */
static void tick_nohz_switch_to_nohz(void)
{
	struct tick_sched *ts = &__get_cpu_var(tick_cpu_sched);
	ktime_t next;

	if (!tick_nohz_enabled)
		return;

	local_irq_disable();
	if (tick_switch_to_oneshot(tick_nohz_handler)) {
		local_irq_enable();
		return;
	}

	ts->nohz_mode = NOHZ_MODE_LOWRES;

	/*
	 * Recycle the hrtimer in ts, so we can share the
	 * hrtimer_forward with the highres code.
	 */
	hrtimer_init(&ts->sched_timer, CLOCK_MONOTONIC, HRTIMER_MODE_ABS);
	/* Get the next period */
	next = tick_init_jiffy_update();

	for (;;) {
		ts->sched_timer.expires = next;
		if (!tick_program_event(next, 0))
			break;
		next = ktime_add(next, tick_period);
	}
	local_irq_enable();

	printk(KERN_INFO "Switched to NOHz mode on CPU #%d\n",
	       smp_processor_id());
}

#else

static inline void tick_nohz_switch_to_nohz(void) { }

#endif /* NO_HZ */

/*
 * High resolution timer specific code
 */
#ifdef CONFIG_HIGH_RES_TIMERS
/*
 * We rearm the timer until we get disabled by the idle code.
 * Called with interrupts disabled and timer->base->cpu_base->lock held.
 */
static enum hrtimer_restart tick_sched_timer(struct hrtimer *timer)
{
	struct tick_sched *ts =
		container_of(timer, struct tick_sched, sched_timer);
	struct pt_regs *regs = get_irq_regs();
	ktime_t now = ktime_get();
	int cpu = smp_processor_id();

#ifdef CONFIG_NO_HZ
	/*
	 * Check if the do_timer duty was dropped. We don't care about
	 * concurrency: This happens only when the cpu in charge went
	 * into a long sleep. If two cpus happen to assign themself to
	 * this duty, then the jiffies update is still serialized by
	 * xtime_lock.
	 */
	if (unlikely(tick_do_timer_cpu == TICK_DO_TIMER_NONE))
		tick_do_timer_cpu = cpu;
#endif

	/* Check, if the jiffies need an update */
	if (tick_do_timer_cpu == cpu)
		tick_do_update_jiffies64(now);

	/*
	 * Do not call, when we are not in irq context and have
	 * no valid regs pointer
	 */
	if (regs) {
		/*
		 * When we are idle and the tick is stopped, we have to touch
		 * the watchdog as we might not schedule for a really long
		 * time. This happens on complete idle SMP systems while
		 * waiting on the login prompt. We also increment the "start of
		 * idle" jiffy stamp so the idle accounting adjustment we do
		 * when we go busy again does not account too much ticks.
		 */
		if (ts->tick_stopped) {
			touch_softlockup_watchdog();
			ts->idle_jiffies++;
		}
		update_process_times(user_mode(regs));
		profile_tick(CPU_PROFILING);
	}

	/* Do not restart, when we are in the idle loop */
	if (ts->tick_stopped)
		return HRTIMER_NORESTART;

	hrtimer_forward(timer, now, tick_period);

	return HRTIMER_RESTART;
}

/**
 * tick_setup_sched_timer - setup the tick emulation timer
 */
void tick_setup_sched_timer(void)
{
	struct tick_sched *ts = &__get_cpu_var(tick_cpu_sched);
	ktime_t now = ktime_get();
	u64 offset;

	/*
	 * Emulate tick processing via per-CPU hrtimers:
	 */
	hrtimer_init(&ts->sched_timer, CLOCK_MONOTONIC, HRTIMER_MODE_ABS);
	ts->sched_timer.function = tick_sched_timer;
	ts->sched_timer.cb_mode = HRTIMER_CB_IRQSAFE_PERCPU;

	/* Get the next period (per cpu) */
	ts->sched_timer.expires = tick_init_jiffy_update();
	offset = ktime_to_ns(tick_period) >> 1;
	do_div(offset, num_possible_cpus());
	offset *= smp_processor_id();
	ts->sched_timer.expires = ktime_add_ns(ts->sched_timer.expires, offset);

	for (;;) {
		hrtimer_forward(&ts->sched_timer, now, tick_period);
		hrtimer_start(&ts->sched_timer, ts->sched_timer.expires,
			      HRTIMER_MODE_ABS);
		/* Check, if the timer was already in the past */
		if (hrtimer_active(&ts->sched_timer))
			break;
		now = ktime_get();
	}

#ifdef CONFIG_NO_HZ
	if (tick_nohz_enabled)
		ts->nohz_mode = NOHZ_MODE_HIGHRES;
#endif
}
#endif /* HIGH_RES_TIMERS */

#if defined CONFIG_NO_HZ || defined CONFIG_HIGH_RES_TIMERS
void tick_cancel_sched_timer(int cpu)
{
	struct tick_sched *ts = &per_cpu(tick_cpu_sched, cpu);

# ifdef CONFIG_HIGH_RES_TIMERS
	if (ts->sched_timer.base)
		hrtimer_cancel(&ts->sched_timer);
# endif

	ts->nohz_mode = NOHZ_MODE_INACTIVE;
}
#endif

/**
 * Async notification about clocksource changes
 */
void tick_clock_notify(void)
{
	int cpu;

	for_each_possible_cpu(cpu)
		set_bit(0, &per_cpu(tick_cpu_sched, cpu).check_clocks);
}

/*
 * Async notification about clock event changes
 */
void tick_oneshot_notify(void)
{
	struct tick_sched *ts = &__get_cpu_var(tick_cpu_sched);

	set_bit(0, &ts->check_clocks);
}

/**
 * Check, if a change happened, which makes oneshot possible.
 *
 * Called cyclic from the hrtimer softirq (driven by the timer
 * softirq) allow_nohz signals, that we can switch into low-res nohz
 * mode, because high resolution timers are disabled (either compile
 * or runtime).
 */
int tick_check_oneshot_change(int allow_nohz)
{
	struct tick_sched *ts = &__get_cpu_var(tick_cpu_sched);

	if (!test_and_clear_bit(0, &ts->check_clocks))
		return 0;

	if (ts->nohz_mode != NOHZ_MODE_INACTIVE)
		return 0;

	if (!timekeeping_valid_for_hres() || !tick_is_oneshot_available())
		return 0;

	if (!allow_nohz)
		return 1;

	tick_nohz_switch_to_nohz();
	return 0;
}