Searched refs:util_avg (Results 1 - 5 of 5) sorted by relevance
/linux-master/kernel/sched/ |
H A D | pelt.c | 266 WRITE_ONCE(sa->util_avg, sa->util_sum / divider); 393 * util_avg and runnable_load_avg are not supported and meaningless. 396 * running a rt/dl task through util_avg, the average thermal pressure is 398 * time weighted "delta" capacity unlike util_avg which is binary.
|
H A D | debug.c | 528 P(se->avg.util_avg); 685 SEQ_printf(m, " .%-30s: %lu\n", "util_avg", 686 cfs_rq->avg.util_avg); 691 SEQ_printf(m, " .%-30s: %ld\n", "removed.util_avg", 692 cfs_rq->removed.util_avg); 1076 P(se.avg.util_avg);
|
H A D | fair.c | 1038 * based on the cfs_rq's current util_avg: 1040 * util_avg = cfs_rq->util_avg / (cfs_rq->load_avg + 1) * se.load.weight 1042 * However, in many cases, the above util_avg does not give a desired 1046 * To solve this problem, we also cap the util_avg of successive tasks to 1049 * util_avg_cap = (cpu_scale - cfs_rq->avg.util_avg) / 2^n 1056 * task util_avg: 512, 256, 128, 64, 32, 16, 8, ... 1057 * cfs_rq util_avg: 512, 768, 896, 960, 992, 1008, 1016, ... 1059 * Finally, that extrapolated util_avg is clamped to the cap (util_avg_cap) 1060 * if util_avg > util_avg_ca [all...] |
H A D | sched.h | 619 unsigned long util_avg; member in struct:cfs_rq::__anon103 3041 return READ_ONCE(rq->avg_dl.util_avg); 3050 return READ_ONCE(rq->avg_rt.util_avg); 3141 return READ_ONCE(rq->avg_irq.util_avg);
|
/linux-master/include/linux/ |
H A D | sched.h | 409 * has a few: load, load_avg, util_avg, freq, and capacity. 427 * The load/runnable/util_avg accumulates an infinite geometric series 438 * [util_avg definition] 440 * util_avg = running% * SCHED_CAPACITY_SCALE 448 * The load/runnable/util_avg doesn't directly factor frequency scaling and CPU 455 * util_avg's SCHED_CAPACITY_SCALE. 479 unsigned long util_avg; member in struct:sched_avg 484 * The UTIL_AVG_UNCHANGED flag is used to synchronize util_est with util_avg 486 * util_avg has not been updated in the meantime. 488 * Since max value of util_est for a task is 1024 (PELT util_avg fo [all...] |
Completed in 140 milliseconds