diff options
Diffstat (limited to 'kernel/delayacct.c')
| -rw-r--r-- | kernel/delayacct.c | 312 |
1 files changed, 214 insertions, 98 deletions
diff --git a/kernel/delayacct.c b/kernel/delayacct.c index d473988c1d0b..30e7912ebb0d 100644 --- a/kernel/delayacct.c +++ b/kernel/delayacct.c @@ -1,129 +1,160 @@ +// SPDX-License-Identifier: GPL-2.0-or-later /* delayacct.c - per-task delay accounting * * Copyright (C) Shailabh Nagar, IBM Corp. 2006 - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. - * - * This program is distributed in the hope that it would be useful, but - * WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See - * the GNU General Public License for more details. */ #include <linux/sched.h> +#include <linux/sched/task.h> +#include <linux/sched/cputime.h> +#include <linux/sched/clock.h> #include <linux/slab.h> #include <linux/taskstats.h> -#include <linux/time.h> #include <linux/sysctl.h> #include <linux/delayacct.h> #include <linux/module.h> -int delayacct_on __read_mostly = 1; /* Delay accounting turned on/off */ -EXPORT_SYMBOL_GPL(delayacct_on); +#define UPDATE_DELAY(type) \ +do { \ + d->type##_delay_max = tsk->delays->type##_delay_max; \ + d->type##_delay_min = tsk->delays->type##_delay_min; \ + tmp = d->type##_delay_total + tsk->delays->type##_delay; \ + d->type##_delay_total = (tmp < d->type##_delay_total) ? 0 : tmp; \ + d->type##_count += tsk->delays->type##_count; \ +} while (0) + +DEFINE_STATIC_KEY_FALSE(delayacct_key); +int delayacct_on __read_mostly; /* Delay accounting turned on/off */ struct kmem_cache *delayacct_cache; -static int __init delayacct_setup_disable(char *str) +static void set_delayacct(bool enabled) { - delayacct_on = 0; + if (enabled) { + static_branch_enable(&delayacct_key); + delayacct_on = 1; + } else { + delayacct_on = 0; + static_branch_disable(&delayacct_key); + } +} + +static int __init delayacct_setup_enable(char *str) +{ + delayacct_on = 1; return 1; } -__setup("nodelayacct", delayacct_setup_disable); +__setup("delayacct", delayacct_setup_enable); void delayacct_init(void) { - delayacct_cache = KMEM_CACHE(task_delay_info, SLAB_PANIC); + delayacct_cache = KMEM_CACHE(task_delay_info, SLAB_PANIC|SLAB_ACCOUNT); delayacct_tsk_init(&init_task); + set_delayacct(delayacct_on); } -void __delayacct_tsk_init(struct task_struct *tsk) +#ifdef CONFIG_PROC_SYSCTL +static int sysctl_delayacct(const struct ctl_table *table, int write, void *buffer, + size_t *lenp, loff_t *ppos) { - tsk->delays = kmem_cache_zalloc(delayacct_cache, GFP_KERNEL); - if (tsk->delays) - spin_lock_init(&tsk->delays->lock); + int state = delayacct_on; + struct ctl_table t; + int err; + + if (write && !capable(CAP_SYS_ADMIN)) + return -EPERM; + + t = *table; + t.data = &state; + err = proc_dointvec_minmax(&t, write, buffer, lenp, ppos); + if (err < 0) + return err; + if (write) + set_delayacct(state); + return err; } -/* - * Start accounting for a delay statistic using - * its starting timestamp (@start) - */ +static const struct ctl_table kern_delayacct_table[] = { + { + .procname = "task_delayacct", + .data = NULL, + .maxlen = sizeof(unsigned int), + .mode = 0644, + .proc_handler = sysctl_delayacct, + .extra1 = SYSCTL_ZERO, + .extra2 = SYSCTL_ONE, + }, +}; -static inline void delayacct_start(struct timespec *start) +static __init int kernel_delayacct_sysctls_init(void) +{ + register_sysctl_init("kernel", kern_delayacct_table); + return 0; +} +late_initcall(kernel_delayacct_sysctls_init); +#endif + +void __delayacct_tsk_init(struct task_struct *tsk) { - do_posix_clock_monotonic_gettime(start); + tsk->delays = kmem_cache_zalloc(delayacct_cache, GFP_KERNEL); + if (tsk->delays) + raw_spin_lock_init(&tsk->delays->lock); } /* - * Finish delay accounting for a statistic using - * its timestamps (@start, @end), accumalator (@total) and @count + * Finish delay accounting for a statistic using its timestamps (@start), + * accumulator (@total) and @count */ - -static void delayacct_end(struct timespec *start, struct timespec *end, - u64 *total, u32 *count) +static void delayacct_end(raw_spinlock_t *lock, u64 *start, u64 *total, u32 *count, u64 *max, u64 *min) { - struct timespec ts; - s64 ns; + s64 ns = local_clock() - *start; unsigned long flags; - do_posix_clock_monotonic_gettime(end); - ts = timespec_sub(*end, *start); - ns = timespec_to_ns(&ts); - if (ns < 0) - return; - - spin_lock_irqsave(¤t->delays->lock, flags); - *total += ns; - (*count)++; - spin_unlock_irqrestore(¤t->delays->lock, flags); + if (ns > 0) { + raw_spin_lock_irqsave(lock, flags); + *total += ns; + (*count)++; + if (ns > *max) + *max = ns; + if (*min == 0 || ns < *min) + *min = ns; + raw_spin_unlock_irqrestore(lock, flags); + } } void __delayacct_blkio_start(void) { - delayacct_start(¤t->delays->blkio_start); + current->delays->blkio_start = local_clock(); } -void __delayacct_blkio_end(void) +/* + * We cannot rely on the `current` macro, as we haven't yet switched back to + * the process being woken. + */ +void __delayacct_blkio_end(struct task_struct *p) { - if (current->delays->flags & DELAYACCT_PF_SWAPIN) - /* Swapin block I/O */ - delayacct_end(¤t->delays->blkio_start, - ¤t->delays->blkio_end, - ¤t->delays->swapin_delay, - ¤t->delays->swapin_count); - else /* Other block I/O */ - delayacct_end(¤t->delays->blkio_start, - ¤t->delays->blkio_end, - ¤t->delays->blkio_delay, - ¤t->delays->blkio_count); + delayacct_end(&p->delays->lock, + &p->delays->blkio_start, + &p->delays->blkio_delay, + &p->delays->blkio_count, + &p->delays->blkio_delay_max, + &p->delays->blkio_delay_min); } -int __delayacct_add_tsk(struct taskstats *d, struct task_struct *tsk) +int delayacct_add_tsk(struct taskstats *d, struct task_struct *tsk) { - s64 tmp; - unsigned long t1; + u64 utime, stime, stimescaled, utimescaled; unsigned long long t2, t3; - unsigned long flags; - struct timespec ts; - cputime_t utime, stime, stimescaled, utimescaled; - - /* Though tsk->delays accessed later, early exit avoids - * unnecessary returning of other data - */ - if (!tsk->delays) - goto done; + unsigned long flags, t1; + s64 tmp; - tmp = (s64)d->cpu_run_real_total; task_cputime(tsk, &utime, &stime); - cputime_to_timespec(utime + stime, &ts); - tmp += timespec_to_ns(&ts); + tmp = (s64)d->cpu_run_real_total; + tmp += utime + stime; d->cpu_run_real_total = (tmp < (s64)d->cpu_run_real_total) ? 0 : tmp; - tmp = (s64)d->cpu_scaled_run_real_total; task_cputime_scaled(tsk, &utimescaled, &stimescaled); - cputime_to_timespec(utimescaled + stimescaled, &ts); - tmp += timespec_to_ns(&ts); + tmp = (s64)d->cpu_scaled_run_real_total; + tmp += utimescaled + stimescaled; d->cpu_scaled_run_real_total = (tmp < (s64)d->cpu_scaled_run_real_total) ? 0 : tmp; @@ -137,28 +168,29 @@ int __delayacct_add_tsk(struct taskstats *d, struct task_struct *tsk) d->cpu_count += t1; + d->cpu_delay_max = tsk->sched_info.max_run_delay; + d->cpu_delay_min = tsk->sched_info.min_run_delay; tmp = (s64)d->cpu_delay_total + t2; d->cpu_delay_total = (tmp < (s64)d->cpu_delay_total) ? 0 : tmp; - tmp = (s64)d->cpu_run_virtual_total + t3; + d->cpu_run_virtual_total = (tmp < (s64)d->cpu_run_virtual_total) ? 0 : tmp; + if (!tsk->delays) + return 0; + /* zero XXX_total, non-zero XXX_count implies XXX stat overflowed */ + raw_spin_lock_irqsave(&tsk->delays->lock, flags); + UPDATE_DELAY(blkio); + UPDATE_DELAY(swapin); + UPDATE_DELAY(freepages); + UPDATE_DELAY(thrashing); + UPDATE_DELAY(compact); + UPDATE_DELAY(wpcopy); + UPDATE_DELAY(irq); + raw_spin_unlock_irqrestore(&tsk->delays->lock, flags); - spin_lock_irqsave(&tsk->delays->lock, flags); - tmp = d->blkio_delay_total + tsk->delays->blkio_delay; - d->blkio_delay_total = (tmp < d->blkio_delay_total) ? 0 : tmp; - tmp = d->swapin_delay_total + tsk->delays->swapin_delay; - d->swapin_delay_total = (tmp < d->swapin_delay_total) ? 0 : tmp; - tmp = d->freepages_delay_total + tsk->delays->freepages_delay; - d->freepages_delay_total = (tmp < d->freepages_delay_total) ? 0 : tmp; - d->blkio_count += tsk->delays->blkio_count; - d->swapin_count += tsk->delays->swapin_count; - d->freepages_count += tsk->delays->freepages_count; - spin_unlock_irqrestore(&tsk->delays->lock, flags); - -done: return 0; } @@ -167,23 +199,107 @@ __u64 __delayacct_blkio_ticks(struct task_struct *tsk) __u64 ret; unsigned long flags; - spin_lock_irqsave(&tsk->delays->lock, flags); - ret = nsec_to_clock_t(tsk->delays->blkio_delay + - tsk->delays->swapin_delay); - spin_unlock_irqrestore(&tsk->delays->lock, flags); + raw_spin_lock_irqsave(&tsk->delays->lock, flags); + ret = nsec_to_clock_t(tsk->delays->blkio_delay); + raw_spin_unlock_irqrestore(&tsk->delays->lock, flags); return ret; } void __delayacct_freepages_start(void) { - delayacct_start(¤t->delays->freepages_start); + current->delays->freepages_start = local_clock(); } void __delayacct_freepages_end(void) { - delayacct_end(¤t->delays->freepages_start, - ¤t->delays->freepages_end, - ¤t->delays->freepages_delay, - ¤t->delays->freepages_count); + delayacct_end(¤t->delays->lock, + ¤t->delays->freepages_start, + ¤t->delays->freepages_delay, + ¤t->delays->freepages_count, + ¤t->delays->freepages_delay_max, + ¤t->delays->freepages_delay_min); +} + +void __delayacct_thrashing_start(bool *in_thrashing) +{ + *in_thrashing = !!current->in_thrashing; + if (*in_thrashing) + return; + + current->in_thrashing = 1; + current->delays->thrashing_start = local_clock(); +} + +void __delayacct_thrashing_end(bool *in_thrashing) +{ + if (*in_thrashing) + return; + + current->in_thrashing = 0; + delayacct_end(¤t->delays->lock, + ¤t->delays->thrashing_start, + ¤t->delays->thrashing_delay, + ¤t->delays->thrashing_count, + ¤t->delays->thrashing_delay_max, + ¤t->delays->thrashing_delay_min); +} + +void __delayacct_swapin_start(void) +{ + current->delays->swapin_start = local_clock(); +} + +void __delayacct_swapin_end(void) +{ + delayacct_end(¤t->delays->lock, + ¤t->delays->swapin_start, + ¤t->delays->swapin_delay, + ¤t->delays->swapin_count, + ¤t->delays->swapin_delay_max, + ¤t->delays->swapin_delay_min); +} + +void __delayacct_compact_start(void) +{ + current->delays->compact_start = local_clock(); +} + +void __delayacct_compact_end(void) +{ + delayacct_end(¤t->delays->lock, + ¤t->delays->compact_start, + ¤t->delays->compact_delay, + ¤t->delays->compact_count, + ¤t->delays->compact_delay_max, + ¤t->delays->compact_delay_min); +} + +void __delayacct_wpcopy_start(void) +{ + current->delays->wpcopy_start = local_clock(); +} + +void __delayacct_wpcopy_end(void) +{ + delayacct_end(¤t->delays->lock, + ¤t->delays->wpcopy_start, + ¤t->delays->wpcopy_delay, + ¤t->delays->wpcopy_count, + ¤t->delays->wpcopy_delay_max, + ¤t->delays->wpcopy_delay_min); +} + +void __delayacct_irq(struct task_struct *task, u32 delta) +{ + unsigned long flags; + + raw_spin_lock_irqsave(&task->delays->lock, flags); + task->delays->irq_delay += delta; + task->delays->irq_count++; + if (delta > task->delays->irq_delay_max) + task->delays->irq_delay_max = delta; + if (delta && (!task->delays->irq_delay_min || delta < task->delays->irq_delay_min)) + task->delays->irq_delay_min = delta; + raw_spin_unlock_irqrestore(&task->delays->lock, flags); } |