static void iseries_dedicated_idle(void)
{
set_thread_flag(TIF_POLLING_NRFLAG);
while (1) {
tick_nohz_idle_enter();
rcu_idle_enter();
if (!need_resched()) {
while (!need_resched()) {
ppc64_runlatch_off();
HMT_low();
if (hvlpevent_is_pending()) {
HMT_medium();
ppc64_runlatch_on();
process_iSeries_events();
}
}
HMT_medium();
}
ppc64_runlatch_on();
rcu_idle_exit();
tick_nohz_idle_exit();
schedule_preempt_disabled();
}
}
static void iseries_shared_idle(void)
{
while (1) {
tick_nohz_idle_enter();
rcu_idle_enter();
while (!need_resched() && !hvlpevent_is_pending()) {
local_irq_disable();
ppc64_runlatch_off();
/* Recheck with irqs off */
if (!need_resched() && !hvlpevent_is_pending())
yield_shared_processor();
HMT_medium();
local_irq_enable();
}
ppc64_runlatch_on();
rcu_idle_exit();
tick_nohz_idle_exit();
if (hvlpevent_is_pending())
process_iSeries_events();
schedule_preempt_disabled();
}
}
void cpu_idle(void)
{
set_thread_flag(TIF_POLLING_NRFLAG);
/* endless idle loop with no priority at all */
while (1) {
tick_nohz_idle_enter();
rcu_idle_enter();
while (!need_resched()) {
check_pgt_cache();
rmb();
clear_thread_flag(TIF_POLLING_NRFLAG);
local_irq_disable();
/* Don't trace irqs off for idle */
stop_critical_timings();
if (!need_resched() && powersave != NULL)
powersave();
start_critical_timings();
local_irq_enable();
set_thread_flag(TIF_POLLING_NRFLAG);
}
rcu_idle_exit();
tick_nohz_idle_exit();
preempt_enable_no_resched();
schedule();
preempt_disable();
}
}
/*
* Generic idle loop implementation
*
* Called with polling cleared.
*/
static void do_idle(void)
{
/*
* 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();
tick_nohz_idle_enter();
while (!need_resched()) {
check_pgt_cache();
rmb();
if (cpu_is_offline(smp_processor_id())) {
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();
}
/*
* The body of the idle task.
*/
void cpu_idle(void)
{
if (ppc_md.idle_loop)
ppc_md.idle_loop(); /* doesn't return */
set_thread_flag(TIF_POLLING_NRFLAG);
while (1) {
tick_nohz_idle_enter();
rcu_idle_enter();
while (!need_resched() && !cpu_should_die()) {
ppc64_runlatch_off();
if (ppc_md.power_save) {
clear_thread_flag(TIF_POLLING_NRFLAG);
/*
* smp_mb is so clearing of TIF_POLLING_NRFLAG
* is ordered w.r.t. need_resched() test.
*/
smp_mb();
local_irq_disable();
/* Don't trace irqs off for idle */
stop_critical_timings();
/* check again after disabling irqs */
if (!need_resched() && !cpu_should_die())
ppc_md.power_save();
start_critical_timings();
/* Some power_save functions return with
* interrupts enabled, some don't.
*/
if (irqs_disabled())
local_irq_enable();
set_thread_flag(TIF_POLLING_NRFLAG);
} else {
/*
* Go into low thread priority and possibly
* low power mode.
*/
HMT_low();
HMT_very_low();
}
}
HMT_medium();
ppc64_runlatch_on();
rcu_idle_exit();
tick_nohz_idle_exit();
if (cpu_should_die()) {
sched_preempt_enable_no_resched();
cpu_die();
}
schedule_preempt_disabled();
}
}
/*
* Generic idle loop implementation
*/
static void cpu_idle_loop(void)
{
while (1) {
tick_nohz_idle_enter();
while (!need_resched()) {
check_pgt_cache();
rmb();
if (cpu_is_offline(smp_processor_id()))
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 {
if (!current_clr_polling_and_test()) {
stop_critical_timings();
rcu_idle_enter();
arch_cpu_idle();
WARN_ON_ONCE(irqs_disabled());
rcu_idle_exit();
start_critical_timings();
} else {
local_irq_enable();
}
__current_set_polling();
}
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();
schedule_preempt_disabled();
}
}
/*
* Generic idle loop implementation
*
* Called with polling cleared.
*/
static void cpu_idle_loop(void)
{
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();
tick_nohz_idle_enter();
while (!need_resched()) {
check_pgt_cache();
rmb();
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() ||
__get_cpu_var(idle_force_poll))
cpu_idle_poll();
else
cpuidle_idle_call();
arch_cpu_idle_exit();
}
tick_nohz_idle_exit();
__current_clr_polling();
/*
* We promise to reschedule if need_resched is set while
* polling is set. That means that clearing polling
* needs to be visible before rescheduling.
*/
smp_mb__after_atomic();
schedule_preempt_disabled();
if (cpu_is_offline(smp_processor_id()))
arch_cpu_idle_dead();
}
}
static void __cpuinit xen_play_dead(void) /* used only with HOTPLUG_CPU */
{
play_dead_common();
HYPERVISOR_vcpu_op(VCPUOP_down, smp_processor_id(), NULL);
cpu_bringup();
/*
* commit 4b0c0f294 (tick: Cleanup NOHZ per cpu data on cpu down)
* clears certain data that the cpu_idle loop (which called us
* and that we return from) expects. The only way to get that
* data back is to call:
*/
tick_nohz_idle_enter();
}
void cpu_idle(void)
{
if (ppc_md.idle_loop)
ppc_md.idle_loop();
set_thread_flag(TIF_POLLING_NRFLAG);
while (1) {
tick_nohz_idle_enter();
rcu_idle_enter();
while (!need_resched() && !cpu_should_die()) {
ppc64_runlatch_off();
if (ppc_md.power_save) {
clear_thread_flag(TIF_POLLING_NRFLAG);
smp_mb();
local_irq_disable();
stop_critical_timings();
if (!need_resched() && !cpu_should_die())
ppc_md.power_save();
start_critical_timings();
if (irqs_disabled())
local_irq_enable();
set_thread_flag(TIF_POLLING_NRFLAG);
} else {
HMT_low();
HMT_very_low();
}
}
HMT_medium();
ppc64_runlatch_on();
rcu_idle_exit();
tick_nohz_idle_exit();
if (cpu_should_die()) {
sched_preempt_enable_no_resched();
cpu_die();
}
schedule_preempt_disabled();
}
}
static void xen_pv_play_dead(void) /* used only with HOTPLUG_CPU */
{
play_dead_common();
HYPERVISOR_vcpu_op(VCPUOP_down, xen_vcpu_nr(smp_processor_id()), NULL);
cpu_bringup();
/*
* commit 4b0c0f294 (tick: Cleanup NOHZ per cpu data on cpu down)
* clears certain data that the cpu_idle loop (which called us
* and that we return from) expects. The only way to get that
* data back is to call:
*/
tick_nohz_idle_enter();
tick_nohz_idle_stop_tick_protected();
cpuhp_online_idle(CPUHP_AP_ONLINE_IDLE);
}
void cpu_idle(void)
{
while (1) {
tick_nohz_idle_enter();
local_irq_disable();
while (!need_resched()) {
idle_sleep();
/* interrupts wake us up, but aren't serviced */
local_irq_enable(); /* service interrupt */
local_irq_disable();
}
local_irq_enable();
tick_nohz_idle_exit();
schedule();
}
}
/*
* Generic idle loop implementation
*/
static void cpu_idle_loop(void)
{
while (1) {
tick_nohz_idle_enter();
while (!need_resched()) {
check_pgt_cache();
rmb();
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() ||
__get_cpu_var(idle_force_poll)) {
cpu_idle_poll();
} else {
if (!current_clr_polling_and_test()) {
stop_critical_timings();
rcu_idle_enter();
arch_cpu_idle();
WARN_ON_ONCE(irqs_disabled());
rcu_idle_exit();
start_critical_timings();
} else {
local_irq_enable();
}
__current_set_polling();
}
arch_cpu_idle_exit();
}
tick_nohz_idle_exit();
schedule_preempt_disabled();
if (cpu_is_offline(smp_processor_id()))
arch_cpu_idle_dead();
}
}
/*
* The idle thread, has rather strange semantics for calling pm_idle,
* but this is what x86 does and we need to do the same, so that
* things like cpuidle get called in the same way. The only difference
* is that we always respect 'hlt_counter' to prevent low power idle.
*/
void cpu_idle(void)
{
local_fiq_enable();
/* endless idle loop with no priority at all */
while (1) {
idle_notifier_call_chain(IDLE_START);
tick_nohz_idle_enter();
rcu_idle_enter();
while (!need_resched()) {
/*
* We need to disable interrupts here
* to ensure we don't miss a wakeup call.
*/
local_irq_disable();
#ifdef CONFIG_PL310_ERRATA_769419
wmb();
#endif
if (hlt_counter) {
local_irq_enable();
cpu_relax();
} else if (!need_resched()) {
stop_critical_timings();
if (cpuidle_idle_call())
pm_idle();
start_critical_timings();
/*
* pm_idle functions must always
* return with IRQs enabled.
*/
WARN_ON(irqs_disabled());
} else
local_irq_enable();
}
rcu_idle_exit();
tick_nohz_idle_exit();
idle_notifier_call_chain(IDLE_END);
schedule_preempt_disabled();
#ifdef CONFIG_HOTPLUG_CPU
if (cpu_is_offline(smp_processor_id()))
cpu_die();
#endif
}
}
/*
* The idle thread. There's no useful work to be
* done, so just try to conserve power and have a
* low exit latency (ie sit in a loop waiting for
* somebody to say that they'd like to reschedule)
*/
void cpu_idle(void)
{
int cpu = smp_processor_id();
current_thread_info()->status |= TS_POLLING;
if (no_idle_nap) {
while (1) {
while (!need_resched())
cpu_relax();
schedule();
}
}
/* endless idle loop with no priority at all */
while (1) {
tick_nohz_idle_enter();
rcu_idle_enter();
while (!need_resched()) {
if (cpu_is_offline(cpu))
BUG(); /* no HOTPLUG_CPU */
local_irq_disable();
__get_cpu_var(irq_stat).idle_timestamp = jiffies;
current_thread_info()->status &= ~TS_POLLING;
/*
* TS_POLLING-cleared state must be visible before we
* test NEED_RESCHED:
*/
smp_mb();
if (!need_resched())
_cpu_idle();
else
local_irq_enable();
current_thread_info()->status |= TS_POLLING;
}
rcu_idle_exit();
tick_nohz_idle_exit();
schedule_preempt_disabled();
}
}
/*
* The idle thread. There's no useful work to be
* done, so just try to conserve power and have a
* low exit latency (ie sit in a loop waiting for
* somebody to say that they'd like to reschedule)
*/
void cpu_idle(void)
{
int cpu = smp_processor_id();
/*
* If we're the non-boot CPU, nothing set the stack canary up
* for us. CPU0 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();
current_thread_info()->status |= TS_POLLING;
/* endless idle loop with no priority at all */
while (1) {
tick_nohz_idle_enter();
rcu_idle_enter();
while (!need_resched()) {
check_pgt_cache();
rmb();
if (cpu_is_offline(cpu))
play_dead();
local_touch_nmi();
local_irq_disable();
/* Don't trace irqs off for idle */
stop_critical_timings();
if (cpuidle_idle_call())
pm_idle();
start_critical_timings();
}
rcu_idle_exit();
tick_nohz_idle_exit();
preempt_enable_no_resched();
schedule();
preempt_disable();
}
}
/*
* The idle thread. We try to conserve power, while trying to keep
* overall latency low. The architecture specific idle is passed
* a value to indicate the level of "idleness" of the system.
*/
void cpu_idle(void)
{
/* endless idle loop with no priority at all */
while (1) {
void (*idle)(void) = pm_idle;
#ifdef CONFIG_HOTPLUG_CPU
if (cpu_is_offline(smp_processor_id()))
cpu_die();
#endif
if (!idle)
idle = default_idle;
tick_nohz_idle_enter();
rcu_idle_enter();
while (!need_resched())
idle();
rcu_idle_exit();
tick_nohz_idle_exit();
schedule_preempt_disabled();
}
}
void cpu_idle(void)
{
set_thread_flag(TIF_POLLING_NRFLAG);
/* endless idle loop with no priority at all */
while (1) {
void (*idle)(void) = pm_idle;
if (!idle)
idle = default_idle;
tick_nohz_idle_enter();
rcu_idle_enter();
while (!need_resched())
idle();
rcu_idle_exit();
tick_nohz_idle_exit();
schedule_preempt_disabled();
check_pgt_cache();
}
}
/*
* The idle thread. There's no useful work to be done, so just try to conserve
* power and have a low exit latency (ie sit in a loop waiting for somebody to
* say that they'd like to reschedule)
*/
void __noreturn cpu_idle(void)
{
int cpu;
/* CPU is going idle. */
cpu = smp_processor_id();
/* endless idle loop with no priority at all */
while (1) {
tick_nohz_idle_enter();
rcu_idle_enter();
while (!need_resched() && cpu_online(cpu)) {
#ifdef CONFIG_MIPS_MT_SMTC
extern void smtc_idle_loop_hook(void);
smtc_idle_loop_hook();
#endif
if (cpu_wait) {
/* Don't trace irqs off for idle */
stop_critical_timings();
(*cpu_wait)();
start_critical_timings();
}
}
#ifdef CONFIG_HOTPLUG_CPU
if (!cpu_online(cpu) && !cpu_isset(cpu, cpu_callin_map) &&
(system_state == SYSTEM_RUNNING ||
system_state == SYSTEM_BOOTING))
play_dead();
#endif
rcu_idle_exit();
tick_nohz_idle_exit();
schedule_preempt_disabled();
}
}
/*
* The idle thread. There's no useful work to be done, so just try to conserve
* power and have a low exit latency (ie sit in a loop waiting for somebody to
* say that they'd like to reschedule)
*/
void cpu_idle(void)
{
unsigned int cpu = smp_processor_id();
set_thread_flag(TIF_POLLING_NRFLAG);
/* endless idle loop with no priority at all */
while (1) {
tick_nohz_idle_enter();
rcu_idle_enter();
while (!need_resched()) {
check_pgt_cache();
rmb();
if (cpu_is_offline(cpu))
play_dead();
local_irq_disable();
/* Don't trace irqs off for idle */
stop_critical_timings();
if (cpuidle_idle_call())
sh_idle();
/*
* Sanity check to ensure that sh_idle() returns
* with IRQs enabled
*/
WARN_ON(irqs_disabled());
start_critical_timings();
}
rcu_idle_exit();
tick_nohz_idle_exit();
schedule_preempt_disabled();
}
}
static int clamp_thread(void *arg)
{
int cpunr = (unsigned long)arg;
DEFINE_TIMER(wakeup_timer, noop_timer, 0, 0);
static const struct sched_param param = {
.sched_priority = MAX_USER_RT_PRIO/2,
};
unsigned int count = 0;
unsigned int target_ratio;
set_bit(cpunr, cpu_clamping_mask);
set_freezable();
init_timer_on_stack(&wakeup_timer);
sched_setscheduler(current, SCHED_FIFO, ¶m);
while (true == clamping && !kthread_should_stop() &&
cpu_online(cpunr)) {
int sleeptime;
unsigned long target_jiffies;
unsigned int guard;
unsigned int compensation = 0;
int interval; /* jiffies to sleep for each attempt */
unsigned int duration_jiffies = msecs_to_jiffies(duration);
unsigned int window_size_now;
try_to_freeze();
/*
* make sure user selected ratio does not take effect until
* the next round. adjust target_ratio if user has changed
* target such that we can converge quickly.
*/
target_ratio = set_target_ratio;
guard = 1 + target_ratio/20;
window_size_now = window_size;
count++;
/*
* systems may have different ability to enter package level
* c-states, thus we need to compensate the injected idle ratio
* to achieve the actual target reported by the HW.
*/
compensation = get_compensation(target_ratio);
interval = duration_jiffies*100/(target_ratio+compensation);
/* align idle time */
target_jiffies = roundup(jiffies, interval);
sleeptime = target_jiffies - jiffies;
if (sleeptime <= 0)
sleeptime = 1;
schedule_timeout_interruptible(sleeptime);
/*
* only elected controlling cpu can collect stats and update
* control parameters.
*/
if (cpunr == control_cpu && !(count%window_size_now)) {
should_skip =
powerclamp_adjust_controls(target_ratio,
guard, window_size_now);
smp_mb();
}
if (should_skip)
continue;
target_jiffies = jiffies + duration_jiffies;
mod_timer(&wakeup_timer, target_jiffies);
if (unlikely(local_softirq_pending()))
continue;
/*
* stop tick sched during idle time, interrupts are still
* allowed. thus jiffies are updated properly.
*/
preempt_disable();
tick_nohz_idle_enter();
/* mwait until target jiffies is reached */
while (time_before(jiffies, target_jiffies)) {
unsigned long ecx = 1;
unsigned long eax = target_mwait;
/*
* REVISIT: may call enter_idle() to notify drivers who
* can save power during cpu idle. same for exit_idle()
*/
local_touch_nmi();
stop_critical_timings();
mwait_idle_with_hints(eax, ecx);
start_critical_timings();
atomic_inc(&idle_wakeup_counter);
}
tick_nohz_idle_exit();
preempt_enable();
}
del_timer_sync(&wakeup_timer);
clear_bit(cpunr, cpu_clamping_mask);
return 0;
}
/*
* 1 HZ polling while clamping is active, useful for userspace
* to monitor actual idle ratio.
*/
static void poll_pkg_cstate(struct work_struct *dummy);
static DECLARE_DELAYED_WORK(poll_pkg_cstate_work, poll_pkg_cstate);
static void poll_pkg_cstate(struct work_struct *dummy)
{
static u64 msr_last;
static u64 tsc_last;
static unsigned long jiffies_last;
u64 msr_now;
unsigned long jiffies_now;
u64 tsc_now;
u64 val64;
msr_now = pkg_state_counter();
rdtscll(tsc_now);
jiffies_now = jiffies;
/* calculate pkg cstate vs tsc ratio */
if (!msr_last || !tsc_last)
pkg_cstate_ratio_cur = 1;
else {
if (tsc_now - tsc_last) {
val64 = 100 * (msr_now - msr_last);
do_div(val64, (tsc_now - tsc_last));
pkg_cstate_ratio_cur = val64;
}
}
/* update record */
msr_last = msr_now;
jiffies_last = jiffies_now;
tsc_last = tsc_now;
if (true == clamping)
schedule_delayed_work(&poll_pkg_cstate_work, HZ);
}
static int start_power_clamp(void)
{
unsigned long cpu;
struct task_struct *thread;
/* check if pkg cstate counter is completely 0, abort in this case */
if (!has_pkg_state_counter()) {
pr_err("pkg cstate counter not functional, abort\n");
return -EINVAL;
}
set_target_ratio = clamp(set_target_ratio, 0U, MAX_TARGET_RATIO - 1);
/* prevent cpu hotplug */
get_online_cpus();
/* prefer BSP */
control_cpu = 0;
if (!cpu_online(control_cpu))
control_cpu = smp_processor_id();
clamping = true;
schedule_delayed_work(&poll_pkg_cstate_work, 0);
/* start one thread per online cpu */
for_each_online_cpu(cpu) {
struct task_struct **p =
per_cpu_ptr(powerclamp_thread, cpu);
thread = kthread_create_on_node(clamp_thread,
(void *) cpu,
cpu_to_node(cpu),
"kidle_inject/%ld", cpu);
/* bind to cpu here */
if (likely(!IS_ERR(thread))) {
kthread_bind(thread, cpu);
wake_up_process(thread);
*p = thread;
}
}
put_online_cpus();
return 0;
}
/*
* Generic idle loop implementation
*
* Called with polling cleared.
*/
static void cpu_idle_loop(void)
{
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();
tick_nohz_idle_enter();
while (!need_resched()) {
check_pgt_cache();
rmb();
if (cpu_is_offline(smp_processor_id()))
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();
/*
* We need to test and propagate the TIF_NEED_RESCHED
* bit here because we might not have send the
* reschedule IPI to idle tasks.
*/
if (tif_need_resched())
set_preempt_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();
sched_ttwu_pending();
schedule_preempt_disabled();
}
}
/*
* Generic idle loop implementation
*
* Called with polling cleared.
*/
static void cpu_idle_loop(void)
{
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(smp_processor_id())) {
rcu_cpu_notify(NULL, CPU_DYING_IDLE,
(void *)(long)smp_processor_id());
smp_mb(); /* all activity before dead. */
this_cpu_write(cpu_dead_idle, true);
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();
}
}