^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1) // SPDX-License-Identifier: GPL-2.0-or-later
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2) /* delayacct.c - per-task delay accounting
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 3) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 4) * Copyright (C) Shailabh Nagar, IBM Corp. 2006
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 5) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 6)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 7) #include <linux/sched.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 8) #include <linux/sched/task.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 9) #include <linux/sched/cputime.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 10) #include <linux/slab.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 11) #include <linux/taskstats.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 12) #include <linux/time.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 13) #include <linux/sysctl.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 14) #include <linux/delayacct.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 15) #include <linux/module.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 16)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 17) int delayacct_on __read_mostly = 1; /* Delay accounting turned on/off */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 18) EXPORT_SYMBOL_GPL(delayacct_on);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 19) struct kmem_cache *delayacct_cache;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 20)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 21) static int __init delayacct_setup_disable(char *str)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 22) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 23) delayacct_on = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 24) return 1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 25) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 26) __setup("nodelayacct", delayacct_setup_disable);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 27)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 28) void delayacct_init(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 29) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 30) delayacct_cache = KMEM_CACHE(task_delay_info, SLAB_PANIC|SLAB_ACCOUNT);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 31) delayacct_tsk_init(&init_task);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 32) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 33)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 34) void __delayacct_tsk_init(struct task_struct *tsk)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 35) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 36) tsk->delays = kmem_cache_zalloc(delayacct_cache, GFP_KERNEL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 37) if (tsk->delays)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 38) raw_spin_lock_init(&tsk->delays->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 39) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 40)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 41) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 42) * Finish delay accounting for a statistic using its timestamps (@start),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 43) * accumalator (@total) and @count
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 44) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 45) static void delayacct_end(raw_spinlock_t *lock, u64 *start, u64 *total,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 46) u32 *count)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 47) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 48) s64 ns = ktime_get_ns() - *start;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 49) unsigned long flags;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 50)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 51) if (ns > 0) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 52) raw_spin_lock_irqsave(lock, flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 53) *total += ns;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 54) (*count)++;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 55) raw_spin_unlock_irqrestore(lock, flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 56) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 57) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 58)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 59) void __delayacct_blkio_start(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 60) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 61) current->delays->blkio_start = ktime_get_ns();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 62) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 63)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 64) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 65) * We cannot rely on the `current` macro, as we haven't yet switched back to
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 66) * the process being woken.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 67) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 68) void __delayacct_blkio_end(struct task_struct *p)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 69) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 70) struct task_delay_info *delays = p->delays;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 71) u64 *total;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 72) u32 *count;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 73)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 74) if (p->delays->flags & DELAYACCT_PF_SWAPIN) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 75) total = &delays->swapin_delay;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 76) count = &delays->swapin_count;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 77) } else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 78) total = &delays->blkio_delay;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 79) count = &delays->blkio_count;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 80) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 81)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 82) delayacct_end(&delays->lock, &delays->blkio_start, total, count);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 83) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 84)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 85) int __delayacct_add_tsk(struct taskstats *d, struct task_struct *tsk)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 86) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 87) u64 utime, stime, stimescaled, utimescaled;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 88) unsigned long long t2, t3;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 89) unsigned long flags, t1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 90) s64 tmp;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 91)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 92) task_cputime(tsk, &utime, &stime);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 93) tmp = (s64)d->cpu_run_real_total;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 94) tmp += utime + stime;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 95) d->cpu_run_real_total = (tmp < (s64)d->cpu_run_real_total) ? 0 : tmp;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 96)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 97) task_cputime_scaled(tsk, &utimescaled, &stimescaled);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 98) tmp = (s64)d->cpu_scaled_run_real_total;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 99) tmp += utimescaled + stimescaled;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 100) d->cpu_scaled_run_real_total =
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 101) (tmp < (s64)d->cpu_scaled_run_real_total) ? 0 : tmp;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 102)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 103) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 104) * No locking available for sched_info (and too expensive to add one)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 105) * Mitigate by taking snapshot of values
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 106) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 107) t1 = tsk->sched_info.pcount;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 108) t2 = tsk->sched_info.run_delay;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 109) t3 = tsk->se.sum_exec_runtime;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 110)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 111) d->cpu_count += t1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 112)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 113) tmp = (s64)d->cpu_delay_total + t2;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 114) d->cpu_delay_total = (tmp < (s64)d->cpu_delay_total) ? 0 : tmp;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 115)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 116) tmp = (s64)d->cpu_run_virtual_total + t3;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 117) d->cpu_run_virtual_total =
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 118) (tmp < (s64)d->cpu_run_virtual_total) ? 0 : tmp;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 119)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 120) /* zero XXX_total, non-zero XXX_count implies XXX stat overflowed */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 121)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 122) raw_spin_lock_irqsave(&tsk->delays->lock, flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 123) tmp = d->blkio_delay_total + tsk->delays->blkio_delay;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 124) d->blkio_delay_total = (tmp < d->blkio_delay_total) ? 0 : tmp;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 125) tmp = d->swapin_delay_total + tsk->delays->swapin_delay;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 126) d->swapin_delay_total = (tmp < d->swapin_delay_total) ? 0 : tmp;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 127) tmp = d->freepages_delay_total + tsk->delays->freepages_delay;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 128) d->freepages_delay_total = (tmp < d->freepages_delay_total) ? 0 : tmp;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 129) tmp = d->thrashing_delay_total + tsk->delays->thrashing_delay;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 130) d->thrashing_delay_total = (tmp < d->thrashing_delay_total) ? 0 : tmp;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 131) d->blkio_count += tsk->delays->blkio_count;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 132) d->swapin_count += tsk->delays->swapin_count;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 133) d->freepages_count += tsk->delays->freepages_count;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 134) d->thrashing_count += tsk->delays->thrashing_count;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 135) raw_spin_unlock_irqrestore(&tsk->delays->lock, flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 136)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 137) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 138) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 139)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 140) __u64 __delayacct_blkio_ticks(struct task_struct *tsk)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 141) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 142) __u64 ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 143) unsigned long flags;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 144)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 145) raw_spin_lock_irqsave(&tsk->delays->lock, flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 146) ret = nsec_to_clock_t(tsk->delays->blkio_delay +
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 147) tsk->delays->swapin_delay);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 148) raw_spin_unlock_irqrestore(&tsk->delays->lock, flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 149) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 150) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 151)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 152) void __delayacct_freepages_start(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 153) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 154) current->delays->freepages_start = ktime_get_ns();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 155) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 156)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 157) void __delayacct_freepages_end(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 158) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 159) delayacct_end(
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 160) ¤t->delays->lock,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 161) ¤t->delays->freepages_start,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 162) ¤t->delays->freepages_delay,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 163) ¤t->delays->freepages_count);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 164) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 165)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 166) void __delayacct_thrashing_start(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 167) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 168) current->delays->thrashing_start = ktime_get_ns();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 169) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 170)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 171) void __delayacct_thrashing_end(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 172) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 173) delayacct_end(¤t->delays->lock,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 174) ¤t->delays->thrashing_start,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 175) ¤t->delays->thrashing_delay,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 176) ¤t->delays->thrashing_count);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 177) }