2 * Deadline Scheduling Class (SCHED_DEADLINE)
4 * Earliest Deadline First (EDF) + Constant Bandwidth Server (CBS).
6 * Tasks that periodically executes their instances for less than their
7 * runtime won't miss any of their deadlines.
8 * Tasks that are not periodic or sporadic or that tries to execute more
9 * than their reserved bandwidth will be slowed down (and may potentially
10 * miss some of their deadlines), and won't affect any other task.
12 * Copyright (C) 2012 Dario Faggioli <raistlin@linux.it>,
13 * Juri Lelli <juri.lelli@gmail.com>,
14 * Michael Trimarchi <michael@amarulasolutions.com>,
15 * Fabio Checconi <fchecconi@gmail.com>
19 #include <linux/slab.h>
21 struct dl_bandwidth def_dl_bandwidth
;
23 static inline struct task_struct
*dl_task_of(struct sched_dl_entity
*dl_se
)
25 return container_of(dl_se
, struct task_struct
, dl
);
28 static inline struct rq
*rq_of_dl_rq(struct dl_rq
*dl_rq
)
30 return container_of(dl_rq
, struct rq
, dl
);
33 static inline struct dl_rq
*dl_rq_of_se(struct sched_dl_entity
*dl_se
)
35 struct task_struct
*p
= dl_task_of(dl_se
);
36 struct rq
*rq
= task_rq(p
);
41 static inline int on_dl_rq(struct sched_dl_entity
*dl_se
)
43 return !RB_EMPTY_NODE(&dl_se
->rb_node
);
46 static inline int is_leftmost(struct task_struct
*p
, struct dl_rq
*dl_rq
)
48 struct sched_dl_entity
*dl_se
= &p
->dl
;
50 return dl_rq
->rb_leftmost
== &dl_se
->rb_node
;
53 void init_dl_bandwidth(struct dl_bandwidth
*dl_b
, u64 period
, u64 runtime
)
55 raw_spin_lock_init(&dl_b
->dl_runtime_lock
);
56 dl_b
->dl_period
= period
;
57 dl_b
->dl_runtime
= runtime
;
60 void init_dl_bw(struct dl_bw
*dl_b
)
62 raw_spin_lock_init(&dl_b
->lock
);
63 raw_spin_lock(&def_dl_bandwidth
.dl_runtime_lock
);
64 if (global_rt_runtime() == RUNTIME_INF
)
67 dl_b
->bw
= to_ratio(global_rt_period(), global_rt_runtime());
68 raw_spin_unlock(&def_dl_bandwidth
.dl_runtime_lock
);
72 void init_dl_rq(struct dl_rq
*dl_rq
)
74 dl_rq
->rb_root
= RB_ROOT
;
77 /* zero means no -deadline tasks */
78 dl_rq
->earliest_dl
.curr
= dl_rq
->earliest_dl
.next
= 0;
80 dl_rq
->dl_nr_migratory
= 0;
81 dl_rq
->overloaded
= 0;
82 dl_rq
->pushable_dl_tasks_root
= RB_ROOT
;
84 init_dl_bw(&dl_rq
->dl_bw
);
90 static inline int dl_overloaded(struct rq
*rq
)
92 return atomic_read(&rq
->rd
->dlo_count
);
95 static inline void dl_set_overload(struct rq
*rq
)
100 cpumask_set_cpu(rq
->cpu
, rq
->rd
->dlo_mask
);
102 * Must be visible before the overload count is
103 * set (as in sched_rt.c).
105 * Matched by the barrier in pull_dl_task().
108 atomic_inc(&rq
->rd
->dlo_count
);
111 static inline void dl_clear_overload(struct rq
*rq
)
116 atomic_dec(&rq
->rd
->dlo_count
);
117 cpumask_clear_cpu(rq
->cpu
, rq
->rd
->dlo_mask
);
120 static void update_dl_migration(struct dl_rq
*dl_rq
)
122 if (dl_rq
->dl_nr_migratory
&& dl_rq
->dl_nr_running
> 1) {
123 if (!dl_rq
->overloaded
) {
124 dl_set_overload(rq_of_dl_rq(dl_rq
));
125 dl_rq
->overloaded
= 1;
127 } else if (dl_rq
->overloaded
) {
128 dl_clear_overload(rq_of_dl_rq(dl_rq
));
129 dl_rq
->overloaded
= 0;
133 static void inc_dl_migration(struct sched_dl_entity
*dl_se
, struct dl_rq
*dl_rq
)
135 struct task_struct
*p
= dl_task_of(dl_se
);
137 if (p
->nr_cpus_allowed
> 1)
138 dl_rq
->dl_nr_migratory
++;
140 update_dl_migration(dl_rq
);
143 static void dec_dl_migration(struct sched_dl_entity
*dl_se
, struct dl_rq
*dl_rq
)
145 struct task_struct
*p
= dl_task_of(dl_se
);
147 if (p
->nr_cpus_allowed
> 1)
148 dl_rq
->dl_nr_migratory
--;
150 update_dl_migration(dl_rq
);
154 * The list of pushable -deadline task is not a plist, like in
155 * sched_rt.c, it is an rb-tree with tasks ordered by deadline.
157 static void enqueue_pushable_dl_task(struct rq
*rq
, struct task_struct
*p
)
159 struct dl_rq
*dl_rq
= &rq
->dl
;
160 struct rb_node
**link
= &dl_rq
->pushable_dl_tasks_root
.rb_node
;
161 struct rb_node
*parent
= NULL
;
162 struct task_struct
*entry
;
165 BUG_ON(!RB_EMPTY_NODE(&p
->pushable_dl_tasks
));
169 entry
= rb_entry(parent
, struct task_struct
,
171 if (dl_entity_preempt(&p
->dl
, &entry
->dl
))
172 link
= &parent
->rb_left
;
174 link
= &parent
->rb_right
;
180 dl_rq
->pushable_dl_tasks_leftmost
= &p
->pushable_dl_tasks
;
182 rb_link_node(&p
->pushable_dl_tasks
, parent
, link
);
183 rb_insert_color(&p
->pushable_dl_tasks
, &dl_rq
->pushable_dl_tasks_root
);
186 static void dequeue_pushable_dl_task(struct rq
*rq
, struct task_struct
*p
)
188 struct dl_rq
*dl_rq
= &rq
->dl
;
190 if (RB_EMPTY_NODE(&p
->pushable_dl_tasks
))
193 if (dl_rq
->pushable_dl_tasks_leftmost
== &p
->pushable_dl_tasks
) {
194 struct rb_node
*next_node
;
196 next_node
= rb_next(&p
->pushable_dl_tasks
);
197 dl_rq
->pushable_dl_tasks_leftmost
= next_node
;
200 rb_erase(&p
->pushable_dl_tasks
, &dl_rq
->pushable_dl_tasks_root
);
201 RB_CLEAR_NODE(&p
->pushable_dl_tasks
);
204 static inline int has_pushable_dl_tasks(struct rq
*rq
)
206 return !RB_EMPTY_ROOT(&rq
->dl
.pushable_dl_tasks_root
);
209 static int push_dl_task(struct rq
*rq
);
211 static inline bool need_pull_dl_task(struct rq
*rq
, struct task_struct
*prev
)
213 return dl_task(prev
);
216 static DEFINE_PER_CPU(struct callback_head
, dl_push_head
);
217 static DEFINE_PER_CPU(struct callback_head
, dl_pull_head
);
219 static void push_dl_tasks(struct rq
*);
220 static void pull_dl_task(struct rq
*);
222 static inline void queue_push_tasks(struct rq
*rq
)
224 if (!has_pushable_dl_tasks(rq
))
227 queue_balance_callback(rq
, &per_cpu(dl_push_head
, rq
->cpu
), push_dl_tasks
);
230 static inline void queue_pull_task(struct rq
*rq
)
232 queue_balance_callback(rq
, &per_cpu(dl_pull_head
, rq
->cpu
), pull_dl_task
);
235 static struct rq
*find_lock_later_rq(struct task_struct
*task
, struct rq
*rq
);
237 static struct rq
*dl_task_offline_migration(struct rq
*rq
, struct task_struct
*p
)
239 struct rq
*later_rq
= NULL
;
240 bool fallback
= false;
242 later_rq
= find_lock_later_rq(p
, rq
);
248 * If we cannot preempt any rq, fall back to pick any
252 cpu
= cpumask_any_and(cpu_active_mask
, tsk_cpus_allowed(p
));
253 if (cpu
>= nr_cpu_ids
) {
255 * Fail to find any suitable cpu.
256 * The task will never come back!
258 BUG_ON(dl_bandwidth_enabled());
261 * If admission control is disabled we
262 * try a little harder to let the task
265 cpu
= cpumask_any(cpu_active_mask
);
267 later_rq
= cpu_rq(cpu
);
268 double_lock_balance(rq
, later_rq
);
272 * By now the task is replenished and enqueued; migrate it.
274 deactivate_task(rq
, p
, 0);
275 set_task_cpu(p
, later_rq
->cpu
);
276 activate_task(later_rq
, p
, 0);
279 resched_curr(later_rq
);
281 double_unlock_balance(later_rq
, rq
);
289 void enqueue_pushable_dl_task(struct rq
*rq
, struct task_struct
*p
)
294 void dequeue_pushable_dl_task(struct rq
*rq
, struct task_struct
*p
)
299 void inc_dl_migration(struct sched_dl_entity
*dl_se
, struct dl_rq
*dl_rq
)
304 void dec_dl_migration(struct sched_dl_entity
*dl_se
, struct dl_rq
*dl_rq
)
308 static inline bool need_pull_dl_task(struct rq
*rq
, struct task_struct
*prev
)
313 static inline void pull_dl_task(struct rq
*rq
)
317 static inline void queue_push_tasks(struct rq
*rq
)
321 static inline void queue_pull_task(struct rq
*rq
)
324 #endif /* CONFIG_SMP */
326 static void enqueue_task_dl(struct rq
*rq
, struct task_struct
*p
, int flags
);
327 static void __dequeue_task_dl(struct rq
*rq
, struct task_struct
*p
, int flags
);
328 static void check_preempt_curr_dl(struct rq
*rq
, struct task_struct
*p
,
332 * We are being explicitly informed that a new instance is starting,
333 * and this means that:
334 * - the absolute deadline of the entity has to be placed at
335 * current time + relative deadline;
336 * - the runtime of the entity has to be set to the maximum value.
338 * The capability of specifying such event is useful whenever a -deadline
339 * entity wants to (try to!) synchronize its behaviour with the scheduler's
340 * one, and to (try to!) reconcile itself with its own scheduling
343 static inline void setup_new_dl_entity(struct sched_dl_entity
*dl_se
,
344 struct sched_dl_entity
*pi_se
)
346 struct dl_rq
*dl_rq
= dl_rq_of_se(dl_se
);
347 struct rq
*rq
= rq_of_dl_rq(dl_rq
);
349 WARN_ON(!dl_se
->dl_new
|| dl_se
->dl_throttled
);
352 * We use the regular wall clock time to set deadlines in the
353 * future; in fact, we must consider execution overheads (time
354 * spent on hardirq context, etc.).
356 dl_se
->deadline
= rq_clock(rq
) + pi_se
->dl_deadline
;
357 dl_se
->runtime
= pi_se
->dl_runtime
;
362 * Pure Earliest Deadline First (EDF) scheduling does not deal with the
363 * possibility of a entity lasting more than what it declared, and thus
364 * exhausting its runtime.
366 * Here we are interested in making runtime overrun possible, but we do
367 * not want a entity which is misbehaving to affect the scheduling of all
369 * Therefore, a budgeting strategy called Constant Bandwidth Server (CBS)
370 * is used, in order to confine each entity within its own bandwidth.
372 * This function deals exactly with that, and ensures that when the runtime
373 * of a entity is replenished, its deadline is also postponed. That ensures
374 * the overrunning entity can't interfere with other entity in the system and
375 * can't make them miss their deadlines. Reasons why this kind of overruns
376 * could happen are, typically, a entity voluntarily trying to overcome its
377 * runtime, or it just underestimated it during sched_setattr().
379 static void replenish_dl_entity(struct sched_dl_entity
*dl_se
,
380 struct sched_dl_entity
*pi_se
)
382 struct dl_rq
*dl_rq
= dl_rq_of_se(dl_se
);
383 struct rq
*rq
= rq_of_dl_rq(dl_rq
);
385 BUG_ON(pi_se
->dl_runtime
<= 0);
388 * This could be the case for a !-dl task that is boosted.
389 * Just go with full inherited parameters.
391 if (dl_se
->dl_deadline
== 0) {
392 dl_se
->deadline
= rq_clock(rq
) + pi_se
->dl_deadline
;
393 dl_se
->runtime
= pi_se
->dl_runtime
;
397 * We keep moving the deadline away until we get some
398 * available runtime for the entity. This ensures correct
399 * handling of situations where the runtime overrun is
402 while (dl_se
->runtime
<= 0) {
403 dl_se
->deadline
+= pi_se
->dl_period
;
404 dl_se
->runtime
+= pi_se
->dl_runtime
;
408 * At this point, the deadline really should be "in
409 * the future" with respect to rq->clock. If it's
410 * not, we are, for some reason, lagging too much!
411 * Anyway, after having warn userspace abut that,
412 * we still try to keep the things running by
413 * resetting the deadline and the budget of the
416 if (dl_time_before(dl_se
->deadline
, rq_clock(rq
))) {
417 printk_deferred_once("sched: DL replenish lagged to much\n");
418 dl_se
->deadline
= rq_clock(rq
) + pi_se
->dl_deadline
;
419 dl_se
->runtime
= pi_se
->dl_runtime
;
422 if (dl_se
->dl_yielded
)
423 dl_se
->dl_yielded
= 0;
424 if (dl_se
->dl_throttled
)
425 dl_se
->dl_throttled
= 0;
429 * Here we check if --at time t-- an entity (which is probably being
430 * [re]activated or, in general, enqueued) can use its remaining runtime
431 * and its current deadline _without_ exceeding the bandwidth it is
432 * assigned (function returns true if it can't). We are in fact applying
433 * one of the CBS rules: when a task wakes up, if the residual runtime
434 * over residual deadline fits within the allocated bandwidth, then we
435 * can keep the current (absolute) deadline and residual budget without
436 * disrupting the schedulability of the system. Otherwise, we should
437 * refill the runtime and set the deadline a period in the future,
438 * because keeping the current (absolute) deadline of the task would
439 * result in breaking guarantees promised to other tasks (refer to
440 * Documentation/scheduler/sched-deadline.txt for more informations).
442 * This function returns true if:
444 * runtime / (deadline - t) > dl_runtime / dl_period ,
446 * IOW we can't recycle current parameters.
448 * Notice that the bandwidth check is done against the period. For
449 * task with deadline equal to period this is the same of using
450 * dl_deadline instead of dl_period in the equation above.
452 static bool dl_entity_overflow(struct sched_dl_entity
*dl_se
,
453 struct sched_dl_entity
*pi_se
, u64 t
)
458 * left and right are the two sides of the equation above,
459 * after a bit of shuffling to use multiplications instead
462 * Note that none of the time values involved in the two
463 * multiplications are absolute: dl_deadline and dl_runtime
464 * are the relative deadline and the maximum runtime of each
465 * instance, runtime is the runtime left for the last instance
466 * and (deadline - t), since t is rq->clock, is the time left
467 * to the (absolute) deadline. Even if overflowing the u64 type
468 * is very unlikely to occur in both cases, here we scale down
469 * as we want to avoid that risk at all. Scaling down by 10
470 * means that we reduce granularity to 1us. We are fine with it,
471 * since this is only a true/false check and, anyway, thinking
472 * of anything below microseconds resolution is actually fiction
473 * (but still we want to give the user that illusion >;).
475 left
= (pi_se
->dl_period
>> DL_SCALE
) * (dl_se
->runtime
>> DL_SCALE
);
476 right
= ((dl_se
->deadline
- t
) >> DL_SCALE
) *
477 (pi_se
->dl_runtime
>> DL_SCALE
);
479 return dl_time_before(right
, left
);
483 * When a -deadline entity is queued back on the runqueue, its runtime and
484 * deadline might need updating.
486 * The policy here is that we update the deadline of the entity only if:
487 * - the current deadline is in the past,
488 * - using the remaining runtime with the current deadline would make
489 * the entity exceed its bandwidth.
491 static void update_dl_entity(struct sched_dl_entity
*dl_se
,
492 struct sched_dl_entity
*pi_se
)
494 struct dl_rq
*dl_rq
= dl_rq_of_se(dl_se
);
495 struct rq
*rq
= rq_of_dl_rq(dl_rq
);
498 * The arrival of a new instance needs special treatment, i.e.,
499 * the actual scheduling parameters have to be "renewed".
502 setup_new_dl_entity(dl_se
, pi_se
);
506 if (dl_time_before(dl_se
->deadline
, rq_clock(rq
)) ||
507 dl_entity_overflow(dl_se
, pi_se
, rq_clock(rq
))) {
508 dl_se
->deadline
= rq_clock(rq
) + pi_se
->dl_deadline
;
509 dl_se
->runtime
= pi_se
->dl_runtime
;
514 * If the entity depleted all its runtime, and if we want it to sleep
515 * while waiting for some new execution time to become available, we
516 * set the bandwidth enforcement timer to the replenishment instant
517 * and try to activate it.
519 * Notice that it is important for the caller to know if the timer
520 * actually started or not (i.e., the replenishment instant is in
521 * the future or in the past).
523 static int start_dl_timer(struct task_struct
*p
)
525 struct sched_dl_entity
*dl_se
= &p
->dl
;
526 struct hrtimer
*timer
= &dl_se
->dl_timer
;
527 struct rq
*rq
= task_rq(p
);
531 lockdep_assert_held(&rq
->lock
);
534 * We want the timer to fire at the deadline, but considering
535 * that it is actually coming from rq->clock and not from
536 * hrtimer's time base reading.
538 act
= ns_to_ktime(dl_se
->deadline
);
539 now
= hrtimer_cb_get_time(timer
);
540 delta
= ktime_to_ns(now
) - rq_clock(rq
);
541 act
= ktime_add_ns(act
, delta
);
544 * If the expiry time already passed, e.g., because the value
545 * chosen as the deadline is too small, don't even try to
546 * start the timer in the past!
548 if (ktime_us_delta(act
, now
) < 0)
552 * !enqueued will guarantee another callback; even if one is already in
553 * progress. This ensures a balanced {get,put}_task_struct().
555 * The race against __run_timer() clearing the enqueued state is
556 * harmless because we're holding task_rq()->lock, therefore the timer
557 * expiring after we've done the check will wait on its task_rq_lock()
558 * and observe our state.
560 if (!hrtimer_is_queued(timer
)) {
562 hrtimer_start(timer
, act
, HRTIMER_MODE_ABS
);
569 * This is the bandwidth enforcement timer callback. If here, we know
570 * a task is not on its dl_rq, since the fact that the timer was running
571 * means the task is throttled and needs a runtime replenishment.
573 * However, what we actually do depends on the fact the task is active,
574 * (it is on its rq) or has been removed from there by a call to
575 * dequeue_task_dl(). In the former case we must issue the runtime
576 * replenishment and add the task back to the dl_rq; in the latter, we just
577 * do nothing but clearing dl_throttled, so that runtime and deadline
578 * updating (and the queueing back to dl_rq) will be done by the
579 * next call to enqueue_task_dl().
581 static enum hrtimer_restart
dl_task_timer(struct hrtimer
*timer
)
583 struct sched_dl_entity
*dl_se
= container_of(timer
,
584 struct sched_dl_entity
,
586 struct task_struct
*p
= dl_task_of(dl_se
);
590 rq
= task_rq_lock(p
, &flags
);
593 * The task might have changed its scheduling policy to something
594 * different than SCHED_DEADLINE (through switched_fromd_dl()).
597 __dl_clear_params(p
);
602 * This is possible if switched_from_dl() raced against a running
603 * callback that took the above !dl_task() path and we've since then
604 * switched back into SCHED_DEADLINE.
606 * There's nothing to do except drop our task reference.
612 * The task might have been boosted by someone else and might be in the
613 * boosting/deboosting path, its not throttled.
615 if (dl_se
->dl_boosted
)
619 * Spurious timer due to start_dl_timer() race; or we already received
620 * a replenishment from rt_mutex_setprio().
622 if (!dl_se
->dl_throttled
)
629 * If the throttle happened during sched-out; like:
636 * __dequeue_task_dl()
639 * We can be both throttled and !queued. Replenish the counter
640 * but do not enqueue -- wait for our wakeup to do that.
642 if (!task_on_rq_queued(p
)) {
643 replenish_dl_entity(dl_se
, dl_se
);
647 enqueue_task_dl(rq
, p
, ENQUEUE_REPLENISH
);
648 if (dl_task(rq
->curr
))
649 check_preempt_curr_dl(rq
, p
, 0);
655 * Perform balancing operations here; after the replenishments. We
656 * cannot drop rq->lock before this, otherwise the assertion in
657 * start_dl_timer() about not missing updates is not true.
659 * If we find that the rq the task was on is no longer available, we
660 * need to select a new rq.
662 * XXX figure out if select_task_rq_dl() deals with offline cpus.
664 if (unlikely(!rq
->online
))
665 rq
= dl_task_offline_migration(rq
, p
);
668 * Queueing this task back might have overloaded rq, check if we need
669 * to kick someone away.
671 if (has_pushable_dl_tasks(rq
)) {
673 * Nothing relies on rq->lock after this, so its safe to drop
676 lockdep_unpin_lock(&rq
->lock
);
678 lockdep_pin_lock(&rq
->lock
);
683 task_rq_unlock(rq
, p
, &flags
);
686 * This can free the task_struct, including this hrtimer, do not touch
687 * anything related to that after this.
691 return HRTIMER_NORESTART
;
694 void init_dl_task_timer(struct sched_dl_entity
*dl_se
)
696 struct hrtimer
*timer
= &dl_se
->dl_timer
;
698 hrtimer_init(timer
, CLOCK_MONOTONIC
, HRTIMER_MODE_REL
);
699 timer
->function
= dl_task_timer
;
703 int dl_runtime_exceeded(struct sched_dl_entity
*dl_se
)
705 return (dl_se
->runtime
<= 0);
708 extern bool sched_rt_bandwidth_account(struct rt_rq
*rt_rq
);
711 * Update the current task's runtime statistics (provided it is still
712 * a -deadline task and has not been removed from the dl_rq).
714 static void update_curr_dl(struct rq
*rq
)
716 struct task_struct
*curr
= rq
->curr
;
717 struct sched_dl_entity
*dl_se
= &curr
->dl
;
720 if (!dl_task(curr
) || !on_dl_rq(dl_se
))
724 * Consumed budget is computed considering the time as
725 * observed by schedulable tasks (excluding time spent
726 * in hardirq context, etc.). Deadlines are instead
727 * computed using hard walltime. This seems to be the more
728 * natural solution, but the full ramifications of this
729 * approach need further study.
731 delta_exec
= rq_clock_task(rq
) - curr
->se
.exec_start
;
732 if (unlikely((s64
)delta_exec
<= 0))
735 schedstat_set(curr
->se
.statistics
.exec_max
,
736 max(curr
->se
.statistics
.exec_max
, delta_exec
));
738 curr
->se
.sum_exec_runtime
+= delta_exec
;
739 account_group_exec_runtime(curr
, delta_exec
);
741 curr
->se
.exec_start
= rq_clock_task(rq
);
742 cpuacct_charge(curr
, delta_exec
);
744 sched_rt_avg_update(rq
, delta_exec
);
746 dl_se
->runtime
-= dl_se
->dl_yielded
? 0 : delta_exec
;
747 if (dl_runtime_exceeded(dl_se
)) {
748 dl_se
->dl_throttled
= 1;
749 __dequeue_task_dl(rq
, curr
, 0);
750 if (unlikely(dl_se
->dl_boosted
|| !start_dl_timer(curr
)))
751 enqueue_task_dl(rq
, curr
, ENQUEUE_REPLENISH
);
753 if (!is_leftmost(curr
, &rq
->dl
))
758 * Because -- for now -- we share the rt bandwidth, we need to
759 * account our runtime there too, otherwise actual rt tasks
760 * would be able to exceed the shared quota.
762 * Account to the root rt group for now.
764 * The solution we're working towards is having the RT groups scheduled
765 * using deadline servers -- however there's a few nasties to figure
766 * out before that can happen.
768 if (rt_bandwidth_enabled()) {
769 struct rt_rq
*rt_rq
= &rq
->rt
;
771 raw_spin_lock(&rt_rq
->rt_runtime_lock
);
773 * We'll let actual RT tasks worry about the overflow here, we
774 * have our own CBS to keep us inline; only account when RT
775 * bandwidth is relevant.
777 if (sched_rt_bandwidth_account(rt_rq
))
778 rt_rq
->rt_time
+= delta_exec
;
779 raw_spin_unlock(&rt_rq
->rt_runtime_lock
);
785 static struct task_struct
*pick_next_earliest_dl_task(struct rq
*rq
, int cpu
);
787 static inline u64
next_deadline(struct rq
*rq
)
789 struct task_struct
*next
= pick_next_earliest_dl_task(rq
, rq
->cpu
);
791 if (next
&& dl_prio(next
->prio
))
792 return next
->dl
.deadline
;
797 static void inc_dl_deadline(struct dl_rq
*dl_rq
, u64 deadline
)
799 struct rq
*rq
= rq_of_dl_rq(dl_rq
);
801 if (dl_rq
->earliest_dl
.curr
== 0 ||
802 dl_time_before(deadline
, dl_rq
->earliest_dl
.curr
)) {
804 * If the dl_rq had no -deadline tasks, or if the new task
805 * has shorter deadline than the current one on dl_rq, we
806 * know that the previous earliest becomes our next earliest,
807 * as the new task becomes the earliest itself.
809 dl_rq
->earliest_dl
.next
= dl_rq
->earliest_dl
.curr
;
810 dl_rq
->earliest_dl
.curr
= deadline
;
811 cpudl_set(&rq
->rd
->cpudl
, rq
->cpu
, deadline
, 1);
812 } else if (dl_rq
->earliest_dl
.next
== 0 ||
813 dl_time_before(deadline
, dl_rq
->earliest_dl
.next
)) {
815 * On the other hand, if the new -deadline task has a
816 * a later deadline than the earliest one on dl_rq, but
817 * it is earlier than the next (if any), we must
818 * recompute the next-earliest.
820 dl_rq
->earliest_dl
.next
= next_deadline(rq
);
824 static void dec_dl_deadline(struct dl_rq
*dl_rq
, u64 deadline
)
826 struct rq
*rq
= rq_of_dl_rq(dl_rq
);
829 * Since we may have removed our earliest (and/or next earliest)
830 * task we must recompute them.
832 if (!dl_rq
->dl_nr_running
) {
833 dl_rq
->earliest_dl
.curr
= 0;
834 dl_rq
->earliest_dl
.next
= 0;
835 cpudl_set(&rq
->rd
->cpudl
, rq
->cpu
, 0, 0);
837 struct rb_node
*leftmost
= dl_rq
->rb_leftmost
;
838 struct sched_dl_entity
*entry
;
840 entry
= rb_entry(leftmost
, struct sched_dl_entity
, rb_node
);
841 dl_rq
->earliest_dl
.curr
= entry
->deadline
;
842 dl_rq
->earliest_dl
.next
= next_deadline(rq
);
843 cpudl_set(&rq
->rd
->cpudl
, rq
->cpu
, entry
->deadline
, 1);
849 static inline void inc_dl_deadline(struct dl_rq
*dl_rq
, u64 deadline
) {}
850 static inline void dec_dl_deadline(struct dl_rq
*dl_rq
, u64 deadline
) {}
852 #endif /* CONFIG_SMP */
855 void inc_dl_tasks(struct sched_dl_entity
*dl_se
, struct dl_rq
*dl_rq
)
857 int prio
= dl_task_of(dl_se
)->prio
;
858 u64 deadline
= dl_se
->deadline
;
860 WARN_ON(!dl_prio(prio
));
861 dl_rq
->dl_nr_running
++;
862 add_nr_running(rq_of_dl_rq(dl_rq
), 1);
864 inc_dl_deadline(dl_rq
, deadline
);
865 inc_dl_migration(dl_se
, dl_rq
);
869 void dec_dl_tasks(struct sched_dl_entity
*dl_se
, struct dl_rq
*dl_rq
)
871 int prio
= dl_task_of(dl_se
)->prio
;
873 WARN_ON(!dl_prio(prio
));
874 WARN_ON(!dl_rq
->dl_nr_running
);
875 dl_rq
->dl_nr_running
--;
876 sub_nr_running(rq_of_dl_rq(dl_rq
), 1);
878 dec_dl_deadline(dl_rq
, dl_se
->deadline
);
879 dec_dl_migration(dl_se
, dl_rq
);
882 static void __enqueue_dl_entity(struct sched_dl_entity
*dl_se
)
884 struct dl_rq
*dl_rq
= dl_rq_of_se(dl_se
);
885 struct rb_node
**link
= &dl_rq
->rb_root
.rb_node
;
886 struct rb_node
*parent
= NULL
;
887 struct sched_dl_entity
*entry
;
890 BUG_ON(!RB_EMPTY_NODE(&dl_se
->rb_node
));
894 entry
= rb_entry(parent
, struct sched_dl_entity
, rb_node
);
895 if (dl_time_before(dl_se
->deadline
, entry
->deadline
))
896 link
= &parent
->rb_left
;
898 link
= &parent
->rb_right
;
904 dl_rq
->rb_leftmost
= &dl_se
->rb_node
;
906 rb_link_node(&dl_se
->rb_node
, parent
, link
);
907 rb_insert_color(&dl_se
->rb_node
, &dl_rq
->rb_root
);
909 inc_dl_tasks(dl_se
, dl_rq
);
912 static void __dequeue_dl_entity(struct sched_dl_entity
*dl_se
)
914 struct dl_rq
*dl_rq
= dl_rq_of_se(dl_se
);
916 if (RB_EMPTY_NODE(&dl_se
->rb_node
))
919 if (dl_rq
->rb_leftmost
== &dl_se
->rb_node
) {
920 struct rb_node
*next_node
;
922 next_node
= rb_next(&dl_se
->rb_node
);
923 dl_rq
->rb_leftmost
= next_node
;
926 rb_erase(&dl_se
->rb_node
, &dl_rq
->rb_root
);
927 RB_CLEAR_NODE(&dl_se
->rb_node
);
929 dec_dl_tasks(dl_se
, dl_rq
);
933 enqueue_dl_entity(struct sched_dl_entity
*dl_se
,
934 struct sched_dl_entity
*pi_se
, int flags
)
936 BUG_ON(on_dl_rq(dl_se
));
939 * If this is a wakeup or a new instance, the scheduling
940 * parameters of the task might need updating. Otherwise,
941 * we want a replenishment of its runtime.
943 if (dl_se
->dl_new
|| flags
& ENQUEUE_WAKEUP
)
944 update_dl_entity(dl_se
, pi_se
);
945 else if (flags
& ENQUEUE_REPLENISH
)
946 replenish_dl_entity(dl_se
, pi_se
);
948 __enqueue_dl_entity(dl_se
);
951 static void dequeue_dl_entity(struct sched_dl_entity
*dl_se
)
953 __dequeue_dl_entity(dl_se
);
956 static void enqueue_task_dl(struct rq
*rq
, struct task_struct
*p
, int flags
)
958 struct task_struct
*pi_task
= rt_mutex_get_top_task(p
);
959 struct sched_dl_entity
*pi_se
= &p
->dl
;
962 * Use the scheduling parameters of the top pi-waiter
963 * task if we have one and its (absolute) deadline is
964 * smaller than our one... OTW we keep our runtime and
967 if (pi_task
&& p
->dl
.dl_boosted
&& dl_prio(pi_task
->normal_prio
)) {
968 pi_se
= &pi_task
->dl
;
969 } else if (!dl_prio(p
->normal_prio
)) {
971 * Special case in which we have a !SCHED_DEADLINE task
972 * that is going to be deboosted, but exceedes its
973 * runtime while doing so. No point in replenishing
974 * it, as it's going to return back to its original
975 * scheduling class after this.
977 BUG_ON(!p
->dl
.dl_boosted
|| flags
!= ENQUEUE_REPLENISH
);
982 * If p is throttled, we do nothing. In fact, if it exhausted
983 * its budget it needs a replenishment and, since it now is on
984 * its rq, the bandwidth timer callback (which clearly has not
985 * run yet) will take care of this.
987 if (p
->dl
.dl_throttled
&& !(flags
& ENQUEUE_REPLENISH
))
990 enqueue_dl_entity(&p
->dl
, pi_se
, flags
);
992 if (!task_current(rq
, p
) && p
->nr_cpus_allowed
> 1)
993 enqueue_pushable_dl_task(rq
, p
);
996 static void __dequeue_task_dl(struct rq
*rq
, struct task_struct
*p
, int flags
)
998 dequeue_dl_entity(&p
->dl
);
999 dequeue_pushable_dl_task(rq
, p
);
1002 static void dequeue_task_dl(struct rq
*rq
, struct task_struct
*p
, int flags
)
1005 __dequeue_task_dl(rq
, p
, flags
);
1009 * Yield task semantic for -deadline tasks is:
1011 * get off from the CPU until our next instance, with
1012 * a new runtime. This is of little use now, since we
1013 * don't have a bandwidth reclaiming mechanism. Anyway,
1014 * bandwidth reclaiming is planned for the future, and
1015 * yield_task_dl will indicate that some spare budget
1016 * is available for other task instances to use it.
1018 static void yield_task_dl(struct rq
*rq
)
1020 struct task_struct
*p
= rq
->curr
;
1023 * We make the task go to sleep until its current deadline by
1024 * forcing its runtime to zero. This way, update_curr_dl() stops
1025 * it and the bandwidth timer will wake it up and will give it
1026 * new scheduling parameters (thanks to dl_yielded=1).
1028 if (p
->dl
.runtime
> 0) {
1029 rq
->curr
->dl
.dl_yielded
= 1;
1032 update_rq_clock(rq
);
1035 * Tell update_rq_clock() that we've just updated,
1036 * so we don't do microscopic update in schedule()
1037 * and double the fastpath cost.
1039 rq_clock_skip_update(rq
, true);
1044 static int find_later_rq(struct task_struct
*task
);
1047 select_task_rq_dl(struct task_struct
*p
, int cpu
, int sd_flag
, int flags
)
1049 struct task_struct
*curr
;
1052 if (sd_flag
!= SD_BALANCE_WAKE
)
1058 curr
= READ_ONCE(rq
->curr
); /* unlocked access */
1061 * If we are dealing with a -deadline task, we must
1062 * decide where to wake it up.
1063 * If it has a later deadline and the current task
1064 * on this rq can't move (provided the waking task
1065 * can!) we prefer to send it somewhere else. On the
1066 * other hand, if it has a shorter deadline, we
1067 * try to make it stay here, it might be important.
1069 if (unlikely(dl_task(curr
)) &&
1070 (curr
->nr_cpus_allowed
< 2 ||
1071 !dl_entity_preempt(&p
->dl
, &curr
->dl
)) &&
1072 (p
->nr_cpus_allowed
> 1)) {
1073 int target
= find_later_rq(p
);
1076 (dl_time_before(p
->dl
.deadline
,
1077 cpu_rq(target
)->dl
.earliest_dl
.curr
) ||
1078 (cpu_rq(target
)->dl
.dl_nr_running
== 0)))
1087 static void check_preempt_equal_dl(struct rq
*rq
, struct task_struct
*p
)
1090 * Current can't be migrated, useless to reschedule,
1091 * let's hope p can move out.
1093 if (rq
->curr
->nr_cpus_allowed
== 1 ||
1094 cpudl_find(&rq
->rd
->cpudl
, rq
->curr
, NULL
) == -1)
1098 * p is migratable, so let's not schedule it and
1099 * see if it is pushed or pulled somewhere else.
1101 if (p
->nr_cpus_allowed
!= 1 &&
1102 cpudl_find(&rq
->rd
->cpudl
, p
, NULL
) != -1)
1108 #endif /* CONFIG_SMP */
1111 * Only called when both the current and waking task are -deadline
1114 static void check_preempt_curr_dl(struct rq
*rq
, struct task_struct
*p
,
1117 if (dl_entity_preempt(&p
->dl
, &rq
->curr
->dl
)) {
1124 * In the unlikely case current and p have the same deadline
1125 * let us try to decide what's the best thing to do...
1127 if ((p
->dl
.deadline
== rq
->curr
->dl
.deadline
) &&
1128 !test_tsk_need_resched(rq
->curr
))
1129 check_preempt_equal_dl(rq
, p
);
1130 #endif /* CONFIG_SMP */
1133 #ifdef CONFIG_SCHED_HRTICK
1134 static void start_hrtick_dl(struct rq
*rq
, struct task_struct
*p
)
1136 hrtick_start(rq
, p
->dl
.runtime
);
1138 #else /* !CONFIG_SCHED_HRTICK */
1139 static void start_hrtick_dl(struct rq
*rq
, struct task_struct
*p
)
1144 static struct sched_dl_entity
*pick_next_dl_entity(struct rq
*rq
,
1145 struct dl_rq
*dl_rq
)
1147 struct rb_node
*left
= dl_rq
->rb_leftmost
;
1152 return rb_entry(left
, struct sched_dl_entity
, rb_node
);
1155 struct task_struct
*pick_next_task_dl(struct rq
*rq
, struct task_struct
*prev
)
1157 struct sched_dl_entity
*dl_se
;
1158 struct task_struct
*p
;
1159 struct dl_rq
*dl_rq
;
1163 if (need_pull_dl_task(rq
, prev
)) {
1165 * This is OK, because current is on_cpu, which avoids it being
1166 * picked for load-balance and preemption/IRQs are still
1167 * disabled avoiding further scheduler activity on it and we're
1168 * being very careful to re-start the picking loop.
1170 lockdep_unpin_lock(&rq
->lock
);
1172 lockdep_pin_lock(&rq
->lock
);
1174 * pull_rt_task() can drop (and re-acquire) rq->lock; this
1175 * means a stop task can slip in, in which case we need to
1176 * re-start task selection.
1178 if (rq
->stop
&& task_on_rq_queued(rq
->stop
))
1183 * When prev is DL, we may throttle it in put_prev_task().
1184 * So, we update time before we check for dl_nr_running.
1186 if (prev
->sched_class
== &dl_sched_class
)
1189 if (unlikely(!dl_rq
->dl_nr_running
))
1192 put_prev_task(rq
, prev
);
1194 dl_se
= pick_next_dl_entity(rq
, dl_rq
);
1197 p
= dl_task_of(dl_se
);
1198 p
->se
.exec_start
= rq_clock_task(rq
);
1200 /* Running task will never be pushed. */
1201 dequeue_pushable_dl_task(rq
, p
);
1203 if (hrtick_enabled(rq
))
1204 start_hrtick_dl(rq
, p
);
1206 queue_push_tasks(rq
);
1211 static void put_prev_task_dl(struct rq
*rq
, struct task_struct
*p
)
1215 if (on_dl_rq(&p
->dl
) && p
->nr_cpus_allowed
> 1)
1216 enqueue_pushable_dl_task(rq
, p
);
1219 static void task_tick_dl(struct rq
*rq
, struct task_struct
*p
, int queued
)
1224 * Even when we have runtime, update_curr_dl() might have resulted in us
1225 * not being the leftmost task anymore. In that case NEED_RESCHED will
1226 * be set and schedule() will start a new hrtick for the next task.
1228 if (hrtick_enabled(rq
) && queued
&& p
->dl
.runtime
> 0 &&
1229 is_leftmost(p
, &rq
->dl
))
1230 start_hrtick_dl(rq
, p
);
1233 static void task_fork_dl(struct task_struct
*p
)
1236 * SCHED_DEADLINE tasks cannot fork and this is achieved through
1241 static void task_dead_dl(struct task_struct
*p
)
1243 struct dl_bw
*dl_b
= dl_bw_of(task_cpu(p
));
1246 * Since we are TASK_DEAD we won't slip out of the domain!
1248 raw_spin_lock_irq(&dl_b
->lock
);
1249 /* XXX we should retain the bw until 0-lag */
1250 dl_b
->total_bw
-= p
->dl
.dl_bw
;
1251 raw_spin_unlock_irq(&dl_b
->lock
);
1254 static void set_curr_task_dl(struct rq
*rq
)
1256 struct task_struct
*p
= rq
->curr
;
1258 p
->se
.exec_start
= rq_clock_task(rq
);
1260 /* You can't push away the running task */
1261 dequeue_pushable_dl_task(rq
, p
);
1266 /* Only try algorithms three times */
1267 #define DL_MAX_TRIES 3
1269 static int pick_dl_task(struct rq
*rq
, struct task_struct
*p
, int cpu
)
1271 if (!task_running(rq
, p
) &&
1272 cpumask_test_cpu(cpu
, tsk_cpus_allowed(p
)))
1277 /* Returns the second earliest -deadline task, NULL otherwise */
1278 static struct task_struct
*pick_next_earliest_dl_task(struct rq
*rq
, int cpu
)
1280 struct rb_node
*next_node
= rq
->dl
.rb_leftmost
;
1281 struct sched_dl_entity
*dl_se
;
1282 struct task_struct
*p
= NULL
;
1285 next_node
= rb_next(next_node
);
1287 dl_se
= rb_entry(next_node
, struct sched_dl_entity
, rb_node
);
1288 p
= dl_task_of(dl_se
);
1290 if (pick_dl_task(rq
, p
, cpu
))
1300 * Return the earliest pushable rq's task, which is suitable to be executed
1301 * on the CPU, NULL otherwise:
1303 static struct task_struct
*pick_earliest_pushable_dl_task(struct rq
*rq
, int cpu
)
1305 struct rb_node
*next_node
= rq
->dl
.pushable_dl_tasks_leftmost
;
1306 struct task_struct
*p
= NULL
;
1308 if (!has_pushable_dl_tasks(rq
))
1313 p
= rb_entry(next_node
, struct task_struct
, pushable_dl_tasks
);
1315 if (pick_dl_task(rq
, p
, cpu
))
1318 next_node
= rb_next(next_node
);
1325 static DEFINE_PER_CPU(cpumask_var_t
, local_cpu_mask_dl
);
1327 static int find_later_rq(struct task_struct
*task
)
1329 struct sched_domain
*sd
;
1330 struct cpumask
*later_mask
= this_cpu_cpumask_var_ptr(local_cpu_mask_dl
);
1331 int this_cpu
= smp_processor_id();
1332 int best_cpu
, cpu
= task_cpu(task
);
1334 /* Make sure the mask is initialized first */
1335 if (unlikely(!later_mask
))
1338 if (task
->nr_cpus_allowed
== 1)
1342 * We have to consider system topology and task affinity
1343 * first, then we can look for a suitable cpu.
1345 best_cpu
= cpudl_find(&task_rq(task
)->rd
->cpudl
,
1351 * If we are here, some target has been found,
1352 * the most suitable of which is cached in best_cpu.
1353 * This is, among the runqueues where the current tasks
1354 * have later deadlines than the task's one, the rq
1355 * with the latest possible one.
1357 * Now we check how well this matches with task's
1358 * affinity and system topology.
1360 * The last cpu where the task run is our first
1361 * guess, since it is most likely cache-hot there.
1363 if (cpumask_test_cpu(cpu
, later_mask
))
1366 * Check if this_cpu is to be skipped (i.e., it is
1367 * not in the mask) or not.
1369 if (!cpumask_test_cpu(this_cpu
, later_mask
))
1373 for_each_domain(cpu
, sd
) {
1374 if (sd
->flags
& SD_WAKE_AFFINE
) {
1377 * If possible, preempting this_cpu is
1378 * cheaper than migrating.
1380 if (this_cpu
!= -1 &&
1381 cpumask_test_cpu(this_cpu
, sched_domain_span(sd
))) {
1387 * Last chance: if best_cpu is valid and is
1388 * in the mask, that becomes our choice.
1390 if (best_cpu
< nr_cpu_ids
&&
1391 cpumask_test_cpu(best_cpu
, sched_domain_span(sd
))) {
1400 * At this point, all our guesses failed, we just return
1401 * 'something', and let the caller sort the things out.
1406 cpu
= cpumask_any(later_mask
);
1407 if (cpu
< nr_cpu_ids
)
1413 /* Locks the rq it finds */
1414 static struct rq
*find_lock_later_rq(struct task_struct
*task
, struct rq
*rq
)
1416 struct rq
*later_rq
= NULL
;
1420 for (tries
= 0; tries
< DL_MAX_TRIES
; tries
++) {
1421 cpu
= find_later_rq(task
);
1423 if ((cpu
== -1) || (cpu
== rq
->cpu
))
1426 later_rq
= cpu_rq(cpu
);
1428 if (later_rq
->dl
.dl_nr_running
&&
1429 !dl_time_before(task
->dl
.deadline
,
1430 later_rq
->dl
.earliest_dl
.curr
)) {
1432 * Target rq has tasks of equal or earlier deadline,
1433 * retrying does not release any lock and is unlikely
1434 * to yield a different result.
1440 /* Retry if something changed. */
1441 if (double_lock_balance(rq
, later_rq
)) {
1442 if (unlikely(task_rq(task
) != rq
||
1443 !cpumask_test_cpu(later_rq
->cpu
,
1444 &task
->cpus_allowed
) ||
1445 task_running(rq
, task
) ||
1446 !task_on_rq_queued(task
))) {
1447 double_unlock_balance(rq
, later_rq
);
1454 * If the rq we found has no -deadline task, or
1455 * its earliest one has a later deadline than our
1456 * task, the rq is a good one.
1458 if (!later_rq
->dl
.dl_nr_running
||
1459 dl_time_before(task
->dl
.deadline
,
1460 later_rq
->dl
.earliest_dl
.curr
))
1463 /* Otherwise we try again. */
1464 double_unlock_balance(rq
, later_rq
);
1471 static struct task_struct
*pick_next_pushable_dl_task(struct rq
*rq
)
1473 struct task_struct
*p
;
1475 if (!has_pushable_dl_tasks(rq
))
1478 p
= rb_entry(rq
->dl
.pushable_dl_tasks_leftmost
,
1479 struct task_struct
, pushable_dl_tasks
);
1481 BUG_ON(rq
->cpu
!= task_cpu(p
));
1482 BUG_ON(task_current(rq
, p
));
1483 BUG_ON(p
->nr_cpus_allowed
<= 1);
1485 BUG_ON(!task_on_rq_queued(p
));
1486 BUG_ON(!dl_task(p
));
1492 * See if the non running -deadline tasks on this rq
1493 * can be sent to some other CPU where they can preempt
1494 * and start executing.
1496 static int push_dl_task(struct rq
*rq
)
1498 struct task_struct
*next_task
;
1499 struct rq
*later_rq
;
1502 if (!rq
->dl
.overloaded
)
1505 next_task
= pick_next_pushable_dl_task(rq
);
1510 if (unlikely(next_task
== rq
->curr
)) {
1516 * If next_task preempts rq->curr, and rq->curr
1517 * can move away, it makes sense to just reschedule
1518 * without going further in pushing next_task.
1520 if (dl_task(rq
->curr
) &&
1521 dl_time_before(next_task
->dl
.deadline
, rq
->curr
->dl
.deadline
) &&
1522 rq
->curr
->nr_cpus_allowed
> 1) {
1527 /* We might release rq lock */
1528 get_task_struct(next_task
);
1530 /* Will lock the rq it'll find */
1531 later_rq
= find_lock_later_rq(next_task
, rq
);
1533 struct task_struct
*task
;
1536 * We must check all this again, since
1537 * find_lock_later_rq releases rq->lock and it is
1538 * then possible that next_task has migrated.
1540 task
= pick_next_pushable_dl_task(rq
);
1541 if (task_cpu(next_task
) == rq
->cpu
&& task
== next_task
) {
1543 * The task is still there. We don't try
1544 * again, some other cpu will pull it when ready.
1553 put_task_struct(next_task
);
1558 deactivate_task(rq
, next_task
, 0);
1559 set_task_cpu(next_task
, later_rq
->cpu
);
1560 activate_task(later_rq
, next_task
, 0);
1563 resched_curr(later_rq
);
1565 double_unlock_balance(rq
, later_rq
);
1568 put_task_struct(next_task
);
1573 static void push_dl_tasks(struct rq
*rq
)
1575 /* push_dl_task() will return true if it moved a -deadline task */
1576 while (push_dl_task(rq
))
1580 static void pull_dl_task(struct rq
*this_rq
)
1582 int this_cpu
= this_rq
->cpu
, cpu
;
1583 struct task_struct
*p
;
1584 bool resched
= false;
1586 u64 dmin
= LONG_MAX
;
1588 if (likely(!dl_overloaded(this_rq
)))
1592 * Match the barrier from dl_set_overloaded; this guarantees that if we
1593 * see overloaded we must also see the dlo_mask bit.
1597 for_each_cpu(cpu
, this_rq
->rd
->dlo_mask
) {
1598 if (this_cpu
== cpu
)
1601 src_rq
= cpu_rq(cpu
);
1604 * It looks racy, abd it is! However, as in sched_rt.c,
1605 * we are fine with this.
1607 if (this_rq
->dl
.dl_nr_running
&&
1608 dl_time_before(this_rq
->dl
.earliest_dl
.curr
,
1609 src_rq
->dl
.earliest_dl
.next
))
1612 /* Might drop this_rq->lock */
1613 double_lock_balance(this_rq
, src_rq
);
1616 * If there are no more pullable tasks on the
1617 * rq, we're done with it.
1619 if (src_rq
->dl
.dl_nr_running
<= 1)
1622 p
= pick_earliest_pushable_dl_task(src_rq
, this_cpu
);
1625 * We found a task to be pulled if:
1626 * - it preempts our current (if there's one),
1627 * - it will preempt the last one we pulled (if any).
1629 if (p
&& dl_time_before(p
->dl
.deadline
, dmin
) &&
1630 (!this_rq
->dl
.dl_nr_running
||
1631 dl_time_before(p
->dl
.deadline
,
1632 this_rq
->dl
.earliest_dl
.curr
))) {
1633 WARN_ON(p
== src_rq
->curr
);
1634 WARN_ON(!task_on_rq_queued(p
));
1637 * Then we pull iff p has actually an earlier
1638 * deadline than the current task of its runqueue.
1640 if (dl_time_before(p
->dl
.deadline
,
1641 src_rq
->curr
->dl
.deadline
))
1646 deactivate_task(src_rq
, p
, 0);
1647 set_task_cpu(p
, this_cpu
);
1648 activate_task(this_rq
, p
, 0);
1649 dmin
= p
->dl
.deadline
;
1651 /* Is there any other task even earlier? */
1654 double_unlock_balance(this_rq
, src_rq
);
1658 resched_curr(this_rq
);
1662 * Since the task is not running and a reschedule is not going to happen
1663 * anytime soon on its runqueue, we try pushing it away now.
1665 static void task_woken_dl(struct rq
*rq
, struct task_struct
*p
)
1667 if (!task_running(rq
, p
) &&
1668 !test_tsk_need_resched(rq
->curr
) &&
1669 p
->nr_cpus_allowed
> 1 &&
1670 dl_task(rq
->curr
) &&
1671 (rq
->curr
->nr_cpus_allowed
< 2 ||
1672 !dl_entity_preempt(&p
->dl
, &rq
->curr
->dl
))) {
1677 static void set_cpus_allowed_dl(struct task_struct
*p
,
1678 const struct cpumask
*new_mask
)
1680 struct root_domain
*src_rd
;
1683 BUG_ON(!dl_task(p
));
1688 * Migrating a SCHED_DEADLINE task between exclusive
1689 * cpusets (different root_domains) entails a bandwidth
1690 * update. We already made space for us in the destination
1691 * domain (see cpuset_can_attach()).
1693 if (!cpumask_intersects(src_rd
->span
, new_mask
)) {
1694 struct dl_bw
*src_dl_b
;
1696 src_dl_b
= dl_bw_of(cpu_of(rq
));
1698 * We now free resources of the root_domain we are migrating
1699 * off. In the worst case, sched_setattr() may temporary fail
1700 * until we complete the update.
1702 raw_spin_lock(&src_dl_b
->lock
);
1703 __dl_clear(src_dl_b
, p
->dl
.dl_bw
);
1704 raw_spin_unlock(&src_dl_b
->lock
);
1707 set_cpus_allowed_common(p
, new_mask
);
1710 /* Assumes rq->lock is held */
1711 static void rq_online_dl(struct rq
*rq
)
1713 if (rq
->dl
.overloaded
)
1714 dl_set_overload(rq
);
1716 cpudl_set_freecpu(&rq
->rd
->cpudl
, rq
->cpu
);
1717 if (rq
->dl
.dl_nr_running
> 0)
1718 cpudl_set(&rq
->rd
->cpudl
, rq
->cpu
, rq
->dl
.earliest_dl
.curr
, 1);
1721 /* Assumes rq->lock is held */
1722 static void rq_offline_dl(struct rq
*rq
)
1724 if (rq
->dl
.overloaded
)
1725 dl_clear_overload(rq
);
1727 cpudl_set(&rq
->rd
->cpudl
, rq
->cpu
, 0, 0);
1728 cpudl_clear_freecpu(&rq
->rd
->cpudl
, rq
->cpu
);
1731 void __init
init_sched_dl_class(void)
1735 for_each_possible_cpu(i
)
1736 zalloc_cpumask_var_node(&per_cpu(local_cpu_mask_dl
, i
),
1737 GFP_KERNEL
, cpu_to_node(i
));
1740 #endif /* CONFIG_SMP */
1742 static void switched_from_dl(struct rq
*rq
, struct task_struct
*p
)
1745 * Start the deadline timer; if we switch back to dl before this we'll
1746 * continue consuming our current CBS slice. If we stay outside of
1747 * SCHED_DEADLINE until the deadline passes, the timer will reset the
1750 if (!start_dl_timer(p
))
1751 __dl_clear_params(p
);
1754 * Since this might be the only -deadline task on the rq,
1755 * this is the right place to try to pull some other one
1756 * from an overloaded cpu, if any.
1758 if (!task_on_rq_queued(p
) || rq
->dl
.dl_nr_running
)
1761 queue_pull_task(rq
);
1765 * When switching to -deadline, we may overload the rq, then
1766 * we try to push someone off, if possible.
1768 static void switched_to_dl(struct rq
*rq
, struct task_struct
*p
)
1770 if (task_on_rq_queued(p
) && rq
->curr
!= p
) {
1772 if (p
->nr_cpus_allowed
> 1 && rq
->dl
.overloaded
)
1773 queue_push_tasks(rq
);
1775 if (dl_task(rq
->curr
))
1776 check_preempt_curr_dl(rq
, p
, 0);
1784 * If the scheduling parameters of a -deadline task changed,
1785 * a push or pull operation might be needed.
1787 static void prio_changed_dl(struct rq
*rq
, struct task_struct
*p
,
1790 if (task_on_rq_queued(p
) || rq
->curr
== p
) {
1793 * This might be too much, but unfortunately
1794 * we don't have the old deadline value, and
1795 * we can't argue if the task is increasing
1796 * or lowering its prio, so...
1798 if (!rq
->dl
.overloaded
)
1799 queue_pull_task(rq
);
1802 * If we now have a earlier deadline task than p,
1803 * then reschedule, provided p is still on this
1806 if (dl_time_before(rq
->dl
.earliest_dl
.curr
, p
->dl
.deadline
))
1810 * Again, we don't know if p has a earlier
1811 * or later deadline, so let's blindly set a
1812 * (maybe not needed) rescheduling point.
1815 #endif /* CONFIG_SMP */
1817 switched_to_dl(rq
, p
);
1820 const struct sched_class dl_sched_class
= {
1821 .next
= &rt_sched_class
,
1822 .enqueue_task
= enqueue_task_dl
,
1823 .dequeue_task
= dequeue_task_dl
,
1824 .yield_task
= yield_task_dl
,
1826 .check_preempt_curr
= check_preempt_curr_dl
,
1828 .pick_next_task
= pick_next_task_dl
,
1829 .put_prev_task
= put_prev_task_dl
,
1832 .select_task_rq
= select_task_rq_dl
,
1833 .set_cpus_allowed
= set_cpus_allowed_dl
,
1834 .rq_online
= rq_online_dl
,
1835 .rq_offline
= rq_offline_dl
,
1836 .task_woken
= task_woken_dl
,
1839 .set_curr_task
= set_curr_task_dl
,
1840 .task_tick
= task_tick_dl
,
1841 .task_fork
= task_fork_dl
,
1842 .task_dead
= task_dead_dl
,
1844 .prio_changed
= prio_changed_dl
,
1845 .switched_from
= switched_from_dl
,
1846 .switched_to
= switched_to_dl
,
1848 .update_curr
= update_curr_dl
,
1851 #ifdef CONFIG_SCHED_DEBUG
1852 extern void print_dl_rq(struct seq_file
*m
, int cpu
, struct dl_rq
*dl_rq
);
1854 void print_dl_stats(struct seq_file
*m
, int cpu
)
1856 print_dl_rq(m
, cpu
, &cpu_rq(cpu
)->dl
);
1858 #endif /* CONFIG_SCHED_DEBUG */