Home
last modified time | relevance | path

Searched refs:util_avg (Results 1 – 8 of 8) sorted by relevance

/linux/Documentation/translations/zh_CN/scheduler/
H A Dsched-energy.rst131 调度器必须决定将任务P放在哪个CPU上，这个任务的util_avg = 200（平均利用率），
134 目前CPU的利用率情况如下图所示。CPU 0-3的util_avg分别为400、100、600和500。
136 模型表中。P的util_avg在图中显示为"PP"::
154 Current OPP: ===== Other OPP: - - - util_avg (100 each): ##
/linux/Documentation/scheduler/
H A Dsched-util-clamp.rst20 Since the scheduler and schedutil are both driven by PELT (util_avg) signals,
629 5.2. UCLAMP_MAX can break PELT (util_avg) signal
636 which would result in util_avg being 1024.
645 p0->util_avg = 300
664 _actual_ util_avg will not be 900 though, but somewhere between 300 and 900. As
665 long as there's idle time, p->util_avg updates will be off by a some margin,
670 p0->util_avg = 300 + small_error
679 this happens, then the _actual_ util_avg will become:
683 p0->util_avg = 1024
689 p1->util_avg = 200
[all …]
H A Dsched-energy.rst155 The scheduler must decide where to place a task P whose util_avg = 200
159 below. CPUs 0-3 have a util_avg of 400, 100, 600 and 500 respectively
162 the Energy Model table. The util_avg of P is shown on the figures
181 Current OPP: ===== Other OPP: - - - util_avg (100 each): ##
/linux/include/linux/sched/
H A Dtopology.h86 unsigned long util_avg; member
/linux/kernel/sched/
H A Dfair.c1316 long cap = (long)(cpu_scale - cfs_rq->avg.util_avg) / 2; in post_init_entity_util_avg()
1334 if (cfs_rq->avg.util_avg != 0) { in post_init_entity_util_avg()
1335 sa->util_avg = cfs_rq->avg.util_avg * se_weight(se); in post_init_entity_util_avg()
1336 sa->util_avg /= (cfs_rq->avg.load_avg + 1); in post_init_entity_util_avg()
1338 if (sa->util_avg > cap) in post_init_entity_util_avg()
1339 sa->util_avg = cap; in post_init_entity_util_avg()
1341 sa->util_avg = cap; in post_init_entity_util_avg()
1345 sa->runnable_avg = sa->util_avg; in post_init_entity_util_avg()
5258 sa->util_avg || in load_avg_is_decayed()
5511 long delta_sum, delta_avg = gcfs_rq->avg.util_avg - se->avg.util_avg; in update_tg_cfs_util()
[all …]
H A Ddebug.c919 P(se->avg.util_avg); in print_cfs_group_stats()
1099 cfs_rq->avg.util_avg); in print_cfs_rq()
1105 cfs_rq->removed.util_avg); in print_cfs_rq()
1484 P(se.avg.util_avg); in proc_sched_show_task()
H A Dsched.h719 unsigned long util_avg; member
3691 return READ_ONCE(rq->avg_dl.util_avg); in cpu_util_dl()
3700 return READ_ONCE(rq->avg_rt.util_avg); in cpu_util_rt()
3838 return READ_ONCE(rq->avg_irq.util_avg); in cpu_util_irq()
/linux/include/linux/
H A Dsched.h448 * has a few: load, load_avg, util_avg, freq, and capacity.
466 * The load/runnable/util_avg accumulates an infinite geometric series
477 * [util_avg definition]
479 * util_avg = running% * SCHED_CAPACITY_SCALE
487 * The load/runnable/util_avg doesn't directly factor frequency scaling and CPU
494 * util_avg's SCHED_CAPACITY_SCALE.
518 unsigned long util_avg; member
523 * The UTIL_AVG_UNCHANGED flag is used to synchronize util_est with util_avg
525 * util_avg has not been updated in the meantime.
527 * Since max value of util_est for a task is 1024 (PELT util_avg fo
[all...]