Release 4.9 kernel/sched/idle.c

/*
 * Generic entry point for the idle threads
 */
#include <linux/sched.h>
#include <linux/cpu.h>
#include <linux/cpuidle.h>
#include <linux/cpuhotplug.h>
#include <linux/tick.h>
#include <linux/mm.h>
#include <linux/stackprotector.h>
#include <linux/suspend.h>

#include <asm/tlb.h>

#include <trace/events/power.h>

#include "sched.h"

/* Linker adds these: start and end of __cpuidle functions */
extern char __cpuidle_text_start[], __cpuidle_text_end[];

/**
 * sched_idle_set_state - Record idle state for the current CPU.
 * @idle_state: State to record.
 */


void sched_idle_set_state(struct cpuidle_state *idle_state)
{
	idle_set_state(this_rq(), idle_state);
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
rafael j. wysocki
rafael j. wysocki
18
100.00%
1
100.00%
Total
18
100.00%
1
100.00%


static int __read_mostly cpu_idle_force_poll;



void cpu_idle_poll_ctrl(bool enable)
{
	if (enable) {
		cpu_idle_force_poll++;
	} else {
		cpu_idle_force_poll--;
		WARN_ON_ONCE(cpu_idle_force_poll < 0);
	}
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
nicolas pitre
nicolas pitre
30
100.00%
1
100.00%
Total
30
100.00%
1
100.00%

#ifdef CONFIG_GENERIC_IDLE_POLL_SETUP


static int __init cpu_idle_poll_setup(char *__unused)
{
	cpu_idle_force_poll = 1;
	return 1;
}
Contributors
Person
Tokens
Prop
Commits
CommitProp
nicolas pitre
nicolas pitre
18
100.00%
1
100.00%
Total
18
100.00%
1
100.00%

__setup("nohlt", cpu_idle_poll_setup);



static int __init cpu_idle_nopoll_setup(char *__unused)
{
	cpu_idle_force_poll = 0;
	return 1;
}
Contributors
Person
Tokens
Prop
Commits
CommitProp
nicolas pitre
nicolas pitre
18
100.00%
1
100.00%
Total
18
100.00%
1
100.00%

__setup("hlt", cpu_idle_nopoll_setup);
#endif



static noinline int __cpuidle cpu_idle_poll(void)
{
	rcu_idle_enter();
	trace_cpu_idle_rcuidle(0, smp_processor_id());
	local_irq_enable();
	stop_critical_timings();
	while (!tif_need_resched() &&
		(cpu_idle_force_poll || tick_check_broadcast_expired()))
		cpu_relax();
	start_critical_timings();
	trace_cpu_idle_rcuidle(PWR_EVENT_EXIT, smp_processor_id());
	rcu_idle_exit();
	return 1;
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
nicolas pitre
nicolas pitre
45
75.00%
1
25.00%
preeti u. murthy
preeti u. murthy
7
11.67%
1
25.00%
daniel bristot de oliveira
daniel bristot de oliveira
6
10.00%
1
25.00%
chris metcalf
chris metcalf
2
3.33%
1
25.00%
Total
60
100.00%
4
100.00%

/* Weak implementations for optional arch specific functions */


void __weak arch_cpu_idle_prepare(void) { }
Contributors
Person
Tokens
Prop
Commits
CommitProp
nicolas pitre
nicolas pitre
7
100.00%
1
100.00%
Total
7
100.00%
1
100.00%



void __weak arch_cpu_idle_enter(void) { }
Contributors
Person
Tokens
Prop
Commits
CommitProp
nicolas pitre
nicolas pitre
7
100.00%
1
100.00%
Total
7
100.00%
1
100.00%



void __weak arch_cpu_idle_exit(void) { }
Contributors
Person
Tokens
Prop
Commits
CommitProp
nicolas pitre
nicolas pitre
7
100.00%
1
100.00%
Total
7
100.00%
1
100.00%



void __weak arch_cpu_idle_dead(void) { }
Contributors
Person
Tokens
Prop
Commits
CommitProp
nicolas pitre
nicolas pitre
7
100.00%
1
100.00%
Total
7
100.00%
1
100.00%



void __weak arch_cpu_idle(void)
{
	cpu_idle_force_poll = 1;
	local_irq_enable();
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
nicolas pitre
nicolas pitre
15
100.00%
1
100.00%
Total
15
100.00%
1
100.00%

/**
 * default_idle_call - Default CPU idle routine.
 *
 * To use when the cpuidle framework cannot be used.
 */


void __cpuidle default_idle_call(void)
{
	if (current_clr_polling_and_test()) {
		local_irq_enable();
	} else {
		stop_critical_timings();
		arch_cpu_idle();
		start_critical_timings();
	}
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
rafael j. wysocki
rafael j. wysocki
19
63.33%
1
33.33%
lucas stach
lucas stach
10
33.33%
1
33.33%
chris metcalf
chris metcalf
1
3.33%
1
33.33%
Total
30
100.00%
3
100.00%



static int call_cpuidle(struct cpuidle_driver *drv, struct cpuidle_device *dev,
		      int next_state)
{
	/*
         * The idle task must be scheduled, it is pointless to go to idle, just
         * update no idle residency and return.
         */
	if (current_clr_polling_and_test()) {
		dev->last_residency = 0;
		local_irq_enable();
		return -EBUSY;
	}

	/*
         * Enter the idle state previously returned by the governor decision.
         * This function will block until an interrupt occurs and will take
         * care of re-enabling the local interrupts
         */
	return cpuidle_enter(drv, dev, next_state);
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
rafael j. wysocki
rafael j. wysocki
51
100.00%
2
100.00%
Total
51
100.00%
2
100.00%

/**
 * cpuidle_idle_call - the main idle function
 *
 * NOTE: no locks or semaphores should be used here
 *
 * On archs that support TIF_POLLING_NRFLAG, is called with polling
 * set, and it returns with polling set.  If it ever stops polling, it
 * must clear the polling bit.
 */


static void cpuidle_idle_call(void)
{
	struct cpuidle_device *dev = cpuidle_get_device();
	struct cpuidle_driver *drv = cpuidle_get_cpu_driver(dev);
	int next_state, entered_state;

	/*
         * Check if the idle task must be rescheduled. If it is the
         * case, exit the function after re-enabling the local irq.
         */
	if (need_resched()) {
		local_irq_enable();
		return;
	}

	/*
         * Tell the RCU framework we are entering an idle section,
         * so no more rcu read side critical sections and one more
         * step to the grace period
         */
	rcu_idle_enter();

	if (cpuidle_not_available(drv, dev)) {
		default_idle_call();
		goto exit_idle;
	}

	/*
         * Suspend-to-idle ("freeze") is a system state in which all user space
         * has been frozen, all I/O devices have been suspended and the only
         * activity happens here and in iterrupts (if any).  In that case bypass
         * the cpuidle governor and go stratight for the deepest idle state
         * available.  Possibly also suspend the local tick and the entire
         * timekeeping to prevent timer interrupts from kicking us out of idle
         * until a proper wakeup interrupt happens.
         */
	if (idle_should_freeze()) {
		entered_state = cpuidle_enter_freeze(drv, dev);
		if (entered_state > 0) {
			local_irq_enable();
			goto exit_idle;
		}

		next_state = cpuidle_find_deepest_state(drv, dev);
		call_cpuidle(drv, dev, next_state);
	} else {
		/*
                 * Ask the cpuidle framework to choose a convenient idle state.
                 */
		next_state = cpuidle_select(drv, dev);
		entered_state = call_cpuidle(drv, dev, next_state);
		/*
                 * Give the governor an opportunity to reflect on the outcome
                 */
		cpuidle_reflect(dev, entered_state);
	}

exit_idle:
	__current_set_polling();

	/*
         * It is up to the idle functions to reenable local interrupts
         */
	if (WARN_ON_ONCE(irqs_disabled()))
		local_irq_enable();

	rcu_idle_exit();
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
daniel lezcano
daniel lezcano
85
51.52%
5
33.33%
rafael j. wysocki
rafael j. wysocki
71
43.03%
5
33.33%
peter zijlstra
peter zijlstra
5
3.03%
2
13.33%
catalin marinas
catalin marinas
2
1.21%
1
6.67%
sudeep holla
sudeep holla
1
0.61%
1
6.67%
ingo molnar
ingo molnar
1
0.61%
1
6.67%
Total
165
100.00%
15
100.00%

/*
 * Generic idle loop implementation
 *
 * Called with polling cleared.
 */


static void cpu_idle_loop(void)
{
	int cpu = smp_processor_id();

	while (1) {
		/*
                 * If the arch has a polling bit, we maintain an invariant:
                 *
                 * Our polling bit is clear if we're not scheduled (i.e. if
                 * rq->curr != rq->idle).  This means that, if rq->idle has
                 * the polling bit set, then setting need_resched is
                 * guaranteed to cause the cpu to reschedule.
                 */

		__current_set_polling();
		quiet_vmstat();
		tick_nohz_idle_enter();

		while (!need_resched()) {
			check_pgt_cache();
			rmb();

			if (cpu_is_offline(cpu)) {
				cpuhp_report_idle_dead();
				arch_cpu_idle_dead();
			}

			local_irq_disable();
			arch_cpu_idle_enter();

			/*
                         * In poll mode we reenable interrupts and spin.
                         *
                         * Also if we detected in the wakeup from idle
                         * path that the tick broadcast device expired
                         * for us, we don't want to go deep idle as we
                         * know that the IPI is going to arrive right
                         * away
                         */
			if (cpu_idle_force_poll || tick_check_broadcast_expired())
				cpu_idle_poll();
			else
				cpuidle_idle_call();

			arch_cpu_idle_exit();
		}

		/*
                 * Since we fell out of the loop above, we know
                 * TIF_NEED_RESCHED must be set, propagate it into
                 * PREEMPT_NEED_RESCHED.
                 *
                 * This is required because for polling idle loops we will
                 * not have had an IPI to fold the state for us.
                 */
		preempt_set_need_resched();
		tick_nohz_idle_exit();
		__current_clr_polling();

		/*
                 * We promise to call sched_ttwu_pending and reschedule
                 * if need_resched is set while polling is set.  That
                 * means that clearing polling needs to be visible
                 * before doing these things.
                 */
		smp_mb__after_atomic();

		sched_ttwu_pending();
		schedule_preempt_disabled();
	}
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
nicolas pitre
nicolas pitre
71
68.93%
1
12.50%
andy lutomirski
andy lutomirski
10
9.71%
1
12.50%
peter zijlstra
peter zijlstra
7
6.80%
2
25.00%
gaurav jindal
gaurav jindal
7
6.80%
1
12.50%
paul e. mckenney
paul e. mckenney
4
3.88%
1
12.50%
christoph lameter
christoph lameter
3
2.91%
1
12.50%
thomas gleixner
thomas gleixner
1
0.97%
1
12.50%
Total
103
100.00%
8
100.00%



bool cpu_in_idle(unsigned long pc)
{
	return pc >= (unsigned long)__cpuidle_text_start &&
		pc < (unsigned long)__cpuidle_text_end;
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
chris metcalf
chris metcalf
26
100.00%
1
100.00%
Total
26
100.00%
1
100.00%



void cpu_startup_entry(enum cpuhp_state state)
{
	/*
         * This #ifdef needs to die, but it's too late in the cycle to
         * make this generic (arm and sh have never invoked the canary
         * init for the non boot cpus!). Will be fixed in 3.11
         */
#ifdef CONFIG_X86
	/*
         * If we're the non-boot CPU, nothing set the stack canary up
         * for us. The boot CPU already has it initialized but no harm
         * in doing it again. This is a good place for updating it, as
         * we wont ever return from this function (so the invalid
         * canaries already on the stack wont ever trigger).
         */
	boot_init_stack_canary();
#endif
	arch_cpu_idle_prepare();
	cpuhp_online_idle(state);
	cpu_idle_loop();
}
Contributors
Person
Tokens
Prop
Commits
CommitProp
nicolas pitre
nicolas pitre
25
83.33%
1
50.00%
thomas gleixner
thomas gleixner
5
16.67%
1
50.00%
Total
30
100.00%
2
100.00%

Overall Contributors
Person
Tokens
Prop
Commits
CommitProp
nicolas pitre
nicolas pitre
300
45.18%
1
3.33%
rafael j. wysocki
rafael j. wysocki
164
24.70%
7
23.33%
daniel lezcano
daniel lezcano
85
12.80%
5
16.67%
chris metcalf
chris metcalf
38
5.72%
1
3.33%
peter zijlstra
peter zijlstra
15
2.26%
4
13.33%
andy lutomirski
andy lutomirski
12
1.81%
1
3.33%
lucas stach
lucas stach
10
1.51%
1
3.33%
thomas gleixner
thomas gleixner
9
1.36%
2
6.67%
gaurav jindal
gaurav jindal
7
1.05%
1
3.33%
preeti u. murthy
preeti u. murthy
7
1.05%
1
3.33%
daniel bristot de oliveira
daniel bristot de oliveira
6
0.90%
1
3.33%
paul e. mckenney
paul e. mckenney
4
0.60%
1
3.33%
christoph lameter
christoph lameter
3
0.45%
1
3.33%
catalin marinas
catalin marinas
2
0.30%
1
3.33%
sudeep holla
sudeep holla
1
0.15%
1
3.33%
ingo molnar
ingo molnar
1
0.15%
1
3.33%
Total
664
100.00%
30
100.00%
  
Directory: kernel/sched