cputime.c 24.3 KB
Newer Older
1 2 3 4 5
#include <linux/export.h>
#include <linux/sched.h>
#include <linux/tsacct_kern.h>
#include <linux/kernel_stat.h>
#include <linux/static_key.h>
6
#include <linux/context_tracking.h>
7
#include "sched.h"
8 9 10
#ifdef CONFIG_PARAVIRT
#include <asm/paravirt.h>
#endif
11 12 13 14 15 16


#ifdef CONFIG_IRQ_TIME_ACCOUNTING

/*
 * There are no locks covering percpu hardirq/softirq time.
17
 * They are only modified in vtime_account, on corresponding CPU
18 19 20
 * with interrupts disabled. So, writes are safe.
 * They are read and saved off onto struct rq in update_rq_clock().
 * This may result in other CPU reading this CPU's irq time and can
21
 * race with irq/vtime_account on this CPU. We would either get old
22 23 24 25
 * or new value with a side effect of accounting a slice of irq time to wrong
 * task when irq is in progress while we read rq->clock. That is a worthy
 * compromise in place of having locks on each irq in account_system_time.
 */
26
DEFINE_PER_CPU(struct irqtime, cpu_irqtime);
27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43

static int sched_clock_irqtime;

void enable_sched_clock_irqtime(void)
{
	sched_clock_irqtime = 1;
}

void disable_sched_clock_irqtime(void)
{
	sched_clock_irqtime = 0;
}

/*
 * Called before incrementing preempt_count on {soft,}irq_enter
 * and before decrementing preempt_count on {soft,}irq_exit.
 */
44
void irqtime_account_irq(struct task_struct *curr)
45
{
46
	struct irqtime *irqtime = this_cpu_ptr(&cpu_irqtime);
47 48 49 50 51 52 53
	s64 delta;
	int cpu;

	if (!sched_clock_irqtime)
		return;

	cpu = smp_processor_id();
54 55
	delta = sched_clock_cpu(cpu) - irqtime->irq_start_time;
	irqtime->irq_start_time += delta;
56

57
	u64_stats_update_begin(&irqtime->sync);
58 59 60 61 62 63 64
	/*
	 * We do not account for softirq time from ksoftirqd here.
	 * We want to continue accounting softirq time to ksoftirqd thread
	 * in that case, so as not to confuse scheduler with a special task
	 * that do not consume any time, but still wants to run.
	 */
	if (hardirq_count())
65
		irqtime->hardirq_time += delta;
66
	else if (in_serving_softirq() && curr != this_cpu_ksoftirqd())
67
		irqtime->softirq_time += delta;
68

69
	u64_stats_update_end(&irqtime->sync);
70
}
71
EXPORT_SYMBOL_GPL(irqtime_account_irq);
72

73
static cputime_t irqtime_account_update(u64 irqtime, int idx, cputime_t maxtime)
74 75
{
	u64 *cpustat = kcpustat_this_cpu->cpustat;
76
	cputime_t irq_cputime;
77

78
	irq_cputime = nsecs_to_cputime64(irqtime) - cpustat[idx];
79
	irq_cputime = min(irq_cputime, maxtime);
80
	cpustat[idx] += irq_cputime;
81

82
	return irq_cputime;
83 84
}

85
static cputime_t irqtime_account_hi_update(cputime_t maxtime)
86
{
87 88 89
	return irqtime_account_update(__this_cpu_read(cpu_irqtime.hardirq_time),
				      CPUTIME_IRQ, maxtime);
}
90

91 92 93 94
static cputime_t irqtime_account_si_update(cputime_t maxtime)
{
	return irqtime_account_update(__this_cpu_read(cpu_irqtime.softirq_time),
				      CPUTIME_SOFTIRQ, maxtime);
95 96 97 98 99 100
}

#else /* CONFIG_IRQ_TIME_ACCOUNTING */

#define sched_clock_irqtime	(0)

101 102 103 104 105 106 107 108 109 110
static cputime_t irqtime_account_hi_update(cputime_t dummy)
{
	return 0;
}

static cputime_t irqtime_account_si_update(cputime_t dummy)
{
	return 0;
}

111 112 113 114 115 116 117 118 119 120 121
#endif /* !CONFIG_IRQ_TIME_ACCOUNTING */

static inline void task_group_account_field(struct task_struct *p, int index,
					    u64 tmp)
{
	/*
	 * Since all updates are sure to touch the root cgroup, we
	 * get ourselves ahead and touch it first. If the root cgroup
	 * is the only cgroup, then nothing else should be necessary.
	 *
	 */
122
	__this_cpu_add(kernel_cpustat.cpustat[index], tmp);
123

124
	cpuacct_account_field(p, index, tmp);
125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142
}

/*
 * Account user cpu time to a process.
 * @p: the process that the cpu time gets accounted to
 * @cputime: the cpu time spent in user space since the last update
 * @cputime_scaled: cputime scaled by cpu frequency
 */
void account_user_time(struct task_struct *p, cputime_t cputime,
		       cputime_t cputime_scaled)
{
	int index;

	/* Add user time to process. */
	p->utime += cputime;
	p->utimescaled += cputime_scaled;
	account_group_user_time(p, cputime);

143
	index = (task_nice(p) > 0) ? CPUTIME_NICE : CPUTIME_USER;
144 145 146 147 148

	/* Add user time to cpustat. */
	task_group_account_field(p, index, (__force u64) cputime);

	/* Account for user time used */
149
	acct_account_cputime(p);
150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169
}

/*
 * Account guest cpu time to a process.
 * @p: the process that the cpu time gets accounted to
 * @cputime: the cpu time spent in virtual machine since the last update
 * @cputime_scaled: cputime scaled by cpu frequency
 */
static void account_guest_time(struct task_struct *p, cputime_t cputime,
			       cputime_t cputime_scaled)
{
	u64 *cpustat = kcpustat_this_cpu->cpustat;

	/* Add guest time to process. */
	p->utime += cputime;
	p->utimescaled += cputime_scaled;
	account_group_user_time(p, cputime);
	p->gtime += cputime;

	/* Add guest time to cpustat. */
170
	if (task_nice(p) > 0) {
171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198
		cpustat[CPUTIME_NICE] += (__force u64) cputime;
		cpustat[CPUTIME_GUEST_NICE] += (__force u64) cputime;
	} else {
		cpustat[CPUTIME_USER] += (__force u64) cputime;
		cpustat[CPUTIME_GUEST] += (__force u64) cputime;
	}
}

/*
 * Account system cpu time to a process and desired cpustat field
 * @p: the process that the cpu time gets accounted to
 * @cputime: the cpu time spent in kernel space since the last update
 * @cputime_scaled: cputime scaled by cpu frequency
 * @target_cputime64: pointer to cpustat field that has to be updated
 */
static inline
void __account_system_time(struct task_struct *p, cputime_t cputime,
			cputime_t cputime_scaled, int index)
{
	/* Add system time to process. */
	p->stime += cputime;
	p->stimescaled += cputime_scaled;
	account_group_system_time(p, cputime);

	/* Add system time to cpustat. */
	task_group_account_field(p, index, (__force u64) cputime);

	/* Account for system time used */
199
	acct_account_cputime(p);
200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254
}

/*
 * Account system cpu time to a process.
 * @p: the process that the cpu time gets accounted to
 * @hardirq_offset: the offset to subtract from hardirq_count()
 * @cputime: the cpu time spent in kernel space since the last update
 * @cputime_scaled: cputime scaled by cpu frequency
 */
void account_system_time(struct task_struct *p, int hardirq_offset,
			 cputime_t cputime, cputime_t cputime_scaled)
{
	int index;

	if ((p->flags & PF_VCPU) && (irq_count() - hardirq_offset == 0)) {
		account_guest_time(p, cputime, cputime_scaled);
		return;
	}

	if (hardirq_count() - hardirq_offset)
		index = CPUTIME_IRQ;
	else if (in_serving_softirq())
		index = CPUTIME_SOFTIRQ;
	else
		index = CPUTIME_SYSTEM;

	__account_system_time(p, cputime, cputime_scaled, index);
}

/*
 * Account for involuntary wait time.
 * @cputime: the cpu time spent in involuntary wait
 */
void account_steal_time(cputime_t cputime)
{
	u64 *cpustat = kcpustat_this_cpu->cpustat;

	cpustat[CPUTIME_STEAL] += (__force u64) cputime;
}

/*
 * Account for idle time.
 * @cputime: the cpu time spent in idle wait
 */
void account_idle_time(cputime_t cputime)
{
	u64 *cpustat = kcpustat_this_cpu->cpustat;
	struct rq *rq = this_rq();

	if (atomic_read(&rq->nr_iowait) > 0)
		cpustat[CPUTIME_IOWAIT] += (__force u64) cputime;
	else
		cpustat[CPUTIME_IDLE] += (__force u64) cputime;
}

255 256 257 258 259
/*
 * When a guest is interrupted for a longer amount of time, missed clock
 * ticks are not redelivered later. Due to that, this function may on
 * occasion account more time than the calling functions think elapsed.
 */
260
static __always_inline cputime_t steal_account_process_time(cputime_t maxtime)
261 262 263
{
#ifdef CONFIG_PARAVIRT
	if (static_key_false(&paravirt_steal_enabled)) {
264
		cputime_t steal_cputime;
265
		u64 steal;
266 267 268 269

		steal = paravirt_steal_clock(smp_processor_id());
		steal -= this_rq()->prev_steal_time;

270 271 272
		steal_cputime = min(nsecs_to_cputime(steal), maxtime);
		account_steal_time(steal_cputime);
		this_rq()->prev_steal_time += cputime_to_nsecs(steal_cputime);
273

274
		return steal_cputime;
275 276
	}
#endif
277
	return 0;
278 279
}

280 281 282 283 284 285 286
/*
 * Account how much elapsed time was spent in steal, irq, or softirq time.
 */
static inline cputime_t account_other_time(cputime_t max)
{
	cputime_t accounted;

287 288 289
	/* Shall be converted to a lockdep-enabled lightweight check */
	WARN_ON_ONCE(!irqs_disabled());

290 291 292 293 294 295 296 297 298 299 300
	accounted = steal_account_process_time(max);

	if (accounted < max)
		accounted += irqtime_account_hi_update(max - accounted);

	if (accounted < max)
		accounted += irqtime_account_si_update(max - accounted);

	return accounted;
}

301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320
#ifdef CONFIG_64BIT
static inline u64 read_sum_exec_runtime(struct task_struct *t)
{
	return t->se.sum_exec_runtime;
}
#else
static u64 read_sum_exec_runtime(struct task_struct *t)
{
	u64 ns;
	struct rq_flags rf;
	struct rq *rq;

	rq = task_rq_lock(t, &rf);
	ns = t->se.sum_exec_runtime;
	task_rq_unlock(rq, t, &rf);

	return ns;
}
#endif

321 322 323 324 325 326 327
/*
 * Accumulate raw cputime values of dead tasks (sig->[us]time) and live
 * tasks (sum on group iteration) belonging to @tsk's group.
 */
void thread_group_cputime(struct task_struct *tsk, struct task_cputime *times)
{
	struct signal_struct *sig = tsk->signal;
328
	cputime_t utime, stime;
329
	struct task_struct *t;
330
	unsigned int seq, nextseq;
331
	unsigned long flags;
332

333 334 335 336 337 338 339 340 341 342 343
	/*
	 * Update current task runtime to account pending time since last
	 * scheduler action or thread_group_cputime() call. This thread group
	 * might have other running tasks on different CPUs, but updating
	 * their runtime can affect syscall performance, so we skip account
	 * those pending times and rely only on values updated on tick or
	 * other scheduler action.
	 */
	if (same_thread_group(current, tsk))
		(void) task_sched_runtime(current);

344
	rcu_read_lock();
345 346 347 348
	/* Attempt a lockless read on the first round. */
	nextseq = 0;
	do {
		seq = nextseq;
349
		flags = read_seqbegin_or_lock_irqsave(&sig->stats_lock, &seq);
350 351 352 353 354 355 356 357
		times->utime = sig->utime;
		times->stime = sig->stime;
		times->sum_exec_runtime = sig->sum_sched_runtime;

		for_each_thread(tsk, t) {
			task_cputime(t, &utime, &stime);
			times->utime += utime;
			times->stime += stime;
358
			times->sum_exec_runtime += read_sum_exec_runtime(t);
359 360 361 362
		}
		/* If lockless access failed, take the lock. */
		nextseq = 1;
	} while (need_seqretry(&sig->stats_lock, seq));
363
	done_seqretry_irqrestore(&sig->stats_lock, seq, flags);
364 365 366
	rcu_read_unlock();
}

367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389
#ifdef CONFIG_IRQ_TIME_ACCOUNTING
/*
 * Account a tick to a process and cpustat
 * @p: the process that the cpu time gets accounted to
 * @user_tick: is the tick from userspace
 * @rq: the pointer to rq
 *
 * Tick demultiplexing follows the order
 * - pending hardirq update
 * - pending softirq update
 * - user_time
 * - idle_time
 * - system time
 *   - check for guest_time
 *   - else account as system_time
 *
 * Check for hardirq is done both for system and user time as there is
 * no timer going off while we are on hardirq and hence we may never get an
 * opportunity to update it solely in system time.
 * p->stime and friends are only updated on system time and not on irq
 * softirq as those do not count in task exec_runtime any more.
 */
static void irqtime_account_process_tick(struct task_struct *p, int user_tick,
390
					 struct rq *rq, int ticks)
391
{
392
	u64 cputime = (__force u64) cputime_one_jiffy * ticks;
393
	cputime_t other;
394

395 396 397 398 399 400 401
	/*
	 * When returning from idle, many ticks can get accounted at
	 * once, including some ticks of steal, irq, and softirq time.
	 * Subtract those ticks from the amount of time accounted to
	 * idle, or potentially user or system time. Due to rounding,
	 * other time can exceed ticks occasionally.
	 */
402
	other = account_other_time(ULONG_MAX);
403
	if (other >= cputime)
404
		return;
405
	cputime -= other;
406

407
	if (this_cpu_ksoftirqd() == p) {
408 409 410 411 412
		/*
		 * ksoftirqd time do not get accounted in cpu_softirq_time.
		 * So, we have to handle it separately here.
		 * Also, p->stime needs to be updated for ksoftirqd.
		 */
413
		__account_system_time(p, cputime, cputime, CPUTIME_SOFTIRQ);
414
	} else if (user_tick) {
415
		account_user_time(p, cputime, cputime);
416
	} else if (p == rq->idle) {
417
		account_idle_time(cputime);
418
	} else if (p->flags & PF_VCPU) { /* System time or guest time */
419
		account_guest_time(p, cputime, cputime);
420
	} else {
421
		__account_system_time(p, cputime, cputime, CPUTIME_SYSTEM);
422 423 424 425 426 427 428
	}
}

static void irqtime_account_idle_ticks(int ticks)
{
	struct rq *rq = this_rq();

429
	irqtime_account_process_tick(current, 0, rq, ticks);
430 431
}
#else /* CONFIG_IRQ_TIME_ACCOUNTING */
432 433
static inline void irqtime_account_idle_ticks(int ticks) {}
static inline void irqtime_account_process_tick(struct task_struct *p, int user_tick,
434
						struct rq *rq, int nr_ticks) {}
435 436 437 438 439 440
#endif /* CONFIG_IRQ_TIME_ACCOUNTING */

/*
 * Use precise platform statistics if available:
 */
#ifdef CONFIG_VIRT_CPU_ACCOUNTING
441

442
#ifndef __ARCH_HAS_VTIME_TASK_SWITCH
443
void vtime_common_task_switch(struct task_struct *prev)
444 445 446 447 448 449
{
	if (is_idle_task(prev))
		vtime_account_idle(prev);
	else
		vtime_account_system(prev);

450
#ifdef CONFIG_VIRT_CPU_ACCOUNTING_NATIVE
451
	vtime_account_user(prev);
452
#endif
453 454 455
	arch_vtime_task_switch(prev);
}
#endif
456

457 458 459 460
#endif /* CONFIG_VIRT_CPU_ACCOUNTING */


#ifdef CONFIG_VIRT_CPU_ACCOUNTING_NATIVE
461 462 463
/*
 * Archs that account the whole time spent in the idle task
 * (outside irq) as idle time can rely on this and just implement
464
 * vtime_account_system() and vtime_account_idle(). Archs that
465 466 467 468 469
 * have other meaning of the idle time (s390 only includes the
 * time spent by the CPU when it's in low power mode) must override
 * vtime_account().
 */
#ifndef __ARCH_HAS_VTIME_ACCOUNT
470
void vtime_account_irq_enter(struct task_struct *tsk)
471
{
472 473 474 475
	if (!in_interrupt() && is_idle_task(tsk))
		vtime_account_idle(tsk);
	else
		vtime_account_system(tsk);
476
}
477
EXPORT_SYMBOL_GPL(vtime_account_irq_enter);
478
#endif /* __ARCH_HAS_VTIME_ACCOUNT */
479 480 481 482 483 484

void task_cputime_adjusted(struct task_struct *p, cputime_t *ut, cputime_t *st)
{
	*ut = p->utime;
	*st = p->stime;
}
485
EXPORT_SYMBOL_GPL(task_cputime_adjusted);
486

487 488 489
void thread_group_cputime_adjusted(struct task_struct *p, cputime_t *ut, cputime_t *st)
{
	struct task_cputime cputime;
490

491 492 493 494 495 496 497 498 499 500 501 502
	thread_group_cputime(p, &cputime);

	*ut = cputime.utime;
	*st = cputime.stime;
}
#else /* !CONFIG_VIRT_CPU_ACCOUNTING_NATIVE */
/*
 * Account a single tick of cpu time.
 * @p: the process that the cpu time gets accounted to
 * @user_tick: indicates if the tick is a user or a system tick
 */
void account_process_tick(struct task_struct *p, int user_tick)
503
{
504
	cputime_t cputime, steal;
505
	struct rq *rq = this_rq();
506

507
	if (vtime_accounting_cpu_enabled())
508 509 510
		return;

	if (sched_clock_irqtime) {
511
		irqtime_account_process_tick(p, user_tick, rq, 1);
512 513 514
		return;
	}

515
	cputime = cputime_one_jiffy;
516
	steal = steal_account_process_time(ULONG_MAX);
517 518

	if (steal >= cputime)
519
		return;
520

521 522
	cputime -= steal;

523
	if (user_tick)
524
		account_user_time(p, cputime, cputime);
525
	else if ((p != rq->idle) || (irq_count() != HARDIRQ_OFFSET))
526
		account_system_time(p, HARDIRQ_OFFSET, cputime, cputime);
527
	else
528
		account_idle_time(cputime);
529
}
530

531 532 533 534 535 536
/*
 * Account multiple ticks of idle time.
 * @ticks: number of stolen ticks
 */
void account_idle_ticks(unsigned long ticks)
{
537
	cputime_t cputime, steal;
538

539 540 541 542 543
	if (sched_clock_irqtime) {
		irqtime_account_idle_ticks(ticks);
		return;
	}

544
	cputime = jiffies_to_cputime(ticks);
545
	steal = steal_account_process_time(ULONG_MAX);
546 547 548 549 550 551

	if (steal >= cputime)
		return;

	cputime -= steal;
	account_idle_time(cputime);
552
}
553

554
/*
555 556
 * Perform (stime * rtime) / total, but avoid multiplication overflow by
 * loosing precision when the numbers are big.
557 558
 */
static cputime_t scale_stime(u64 stime, u64 rtime, u64 total)
559
{
560
	u64 scaled;
561

562 563
	for (;;) {
		/* Make sure "rtime" is the bigger of stime/rtime */
564 565
		if (stime > rtime)
			swap(rtime, stime);
566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587

		/* Make sure 'total' fits in 32 bits */
		if (total >> 32)
			goto drop_precision;

		/* Does rtime (and thus stime) fit in 32 bits? */
		if (!(rtime >> 32))
			break;

		/* Can we just balance rtime/stime rather than dropping bits? */
		if (stime >> 31)
			goto drop_precision;

		/* We can grow stime and shrink rtime and try to make them both fit */
		stime <<= 1;
		rtime >>= 1;
		continue;

drop_precision:
		/* We drop from rtime, it has more bits than stime */
		rtime >>= 1;
		total >>= 1;
588
	}
589

590 591 592 593 594
	/*
	 * Make sure gcc understands that this is a 32x32->64 multiply,
	 * followed by a 64/32->64 divide.
	 */
	scaled = div_u64((u64) (u32) stime * (u64) (u32) rtime, (u32)total);
595
	return (__force cputime_t) scaled;
596 597
}

598
/*
599 600
 * Adjust tick based cputime random precision against scheduler runtime
 * accounting.
601
 *
602 603 604 605 606 607 608 609 610 611 612 613 614 615 616
 * Tick based cputime accounting depend on random scheduling timeslices of a
 * task to be interrupted or not by the timer.  Depending on these
 * circumstances, the number of these interrupts may be over or
 * under-optimistic, matching the real user and system cputime with a variable
 * precision.
 *
 * Fix this by scaling these tick based values against the total runtime
 * accounted by the CFS scheduler.
 *
 * This code provides the following guarantees:
 *
 *   stime + utime == rtime
 *   stime_i+1 >= stime_i, utime_i+1 >= utime_i
 *
 * Assuming that rtime_i+1 >= rtime_i.
617
 */
618
static void cputime_adjust(struct task_cputime *curr,
619
			   struct prev_cputime *prev,
620
			   cputime_t *ut, cputime_t *st)
621
{
622
	cputime_t rtime, stime, utime;
623
	unsigned long flags;
624

625 626
	/* Serialize concurrent callers such that we can honour our guarantees */
	raw_spin_lock_irqsave(&prev->lock, flags);
627
	rtime = nsecs_to_cputime(curr->sum_exec_runtime);
628

629
	/*
630 631 632 633 634 635
	 * This is possible under two circumstances:
	 *  - rtime isn't monotonic after all (a bug);
	 *  - we got reordered by the lock.
	 *
	 * In both cases this acts as a filter such that the rest of the code
	 * can assume it is monotonic regardless of anything else.
636 637 638 639
	 */
	if (prev->stime + prev->utime >= rtime)
		goto out;

640 641 642
	stime = curr->stime;
	utime = curr->utime;

643 644 645 646 647 648 649
	/*
	 * If either stime or both stime and utime are 0, assume all runtime is
	 * userspace. Once a task gets some ticks, the monotonicy code at
	 * 'update' will ensure things converge to the observed ratio.
	 */
	if (stime == 0) {
		utime = rtime;
650 651
		goto update;
	}
652

653 654
	if (utime == 0) {
		stime = rtime;
655
		goto update;
656
	}
657

658 659 660
	stime = scale_stime((__force u64)stime, (__force u64)rtime,
			    (__force u64)(stime + utime));

661
update:
662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682
	/*
	 * Make sure stime doesn't go backwards; this preserves monotonicity
	 * for utime because rtime is monotonic.
	 *
	 *  utime_i+1 = rtime_i+1 - stime_i
	 *            = rtime_i+1 - (rtime_i - utime_i)
	 *            = (rtime_i+1 - rtime_i) + utime_i
	 *            >= utime_i
	 */
	if (stime < prev->stime)
		stime = prev->stime;
	utime = rtime - stime;

	/*
	 * Make sure utime doesn't go backwards; this still preserves
	 * monotonicity for stime, analogous argument to above.
	 */
	if (utime < prev->utime) {
		utime = prev->utime;
		stime = rtime - utime;
	}
683

684 685
	prev->stime = stime;
	prev->utime = utime;
686
out:
687 688
	*ut = prev->utime;
	*st = prev->stime;
689
	raw_spin_unlock_irqrestore(&prev->lock, flags);
690
}
691

692 693 694 695 696 697
void task_cputime_adjusted(struct task_struct *p, cputime_t *ut, cputime_t *st)
{
	struct task_cputime cputime = {
		.sum_exec_runtime = p->se.sum_exec_runtime,
	};

698
	task_cputime(p, &cputime.utime, &cputime.stime);
699
	cputime_adjust(&cputime, &p->prev_cputime, ut, st);
700
}
701
EXPORT_SYMBOL_GPL(task_cputime_adjusted);
702

703
void thread_group_cputime_adjusted(struct task_struct *p, cputime_t *ut, cputime_t *st)
704 705 706 707
{
	struct task_cputime cputime;

	thread_group_cputime(p, &cputime);
708
	cputime_adjust(&cputime, &p->signal->prev_cputime, ut, st);
709
}
710
#endif /* !CONFIG_VIRT_CPU_ACCOUNTING_NATIVE */
711 712

#ifdef CONFIG_VIRT_CPU_ACCOUNTING_GEN
713
static cputime_t vtime_delta(struct task_struct *tsk)
714
{
715
	unsigned long now = READ_ONCE(jiffies);
716

717
	if (time_before(now, (unsigned long)tsk->vtime_snap))
718
		return 0;
719

720
	return jiffies_to_cputime(now - tsk->vtime_snap);
721 722 723
}

static cputime_t get_vtime_delta(struct task_struct *tsk)
724
{
725
	unsigned long now = READ_ONCE(jiffies);
726
	cputime_t delta, other;
727

728 729 730 731 732 733 734
	/*
	 * Unlike tick based timing, vtime based timing never has lost
	 * ticks, and no need for steal time accounting to make up for
	 * lost ticks. Vtime accounts a rounded version of actual
	 * elapsed time. Limit account_other_time to prevent rounding
	 * errors from causing elapsed vtime to go negative.
	 */
735
	delta = jiffies_to_cputime(now - tsk->vtime_snap);
736
	other = account_other_time(delta);
737
	WARN_ON_ONCE(tsk->vtime_snap_whence == VTIME_INACTIVE);
738
	tsk->vtime_snap = now;
739

740
	return delta - other;
741 742
}

743 744 745 746
static void __vtime_account_system(struct task_struct *tsk)
{
	cputime_t delta_cpu = get_vtime_delta(tsk);

747
	account_system_time(tsk, irq_count(), delta_cpu, delta_cpu);
748 749
}

750 751
void vtime_account_system(struct task_struct *tsk)
{
752 753 754
	if (!vtime_delta(tsk))
		return;

755
	write_seqcount_begin(&tsk->vtime_seqcount);
756
	__vtime_account_system(tsk);
757
	write_seqcount_end(&tsk->vtime_seqcount);
758
}
759

760 761
void vtime_account_user(struct task_struct *tsk)
{
762 763
	cputime_t delta_cpu;

764
	write_seqcount_begin(&tsk->vtime_seqcount);
765
	tsk->vtime_snap_whence = VTIME_SYS;
766 767
	if (vtime_delta(tsk)) {
		delta_cpu = get_vtime_delta(tsk);
768
		account_user_time(tsk, delta_cpu, delta_cpu);
769
	}
770
	write_seqcount_end(&tsk->vtime_seqcount);
771 772 773 774
}

void vtime_user_enter(struct task_struct *tsk)
{
775
	write_seqcount_begin(&tsk->vtime_seqcount);
776 777
	if (vtime_delta(tsk))
		__vtime_account_system(tsk);
778
	tsk->vtime_snap_whence = VTIME_USER;
779
	write_seqcount_end(&tsk->vtime_seqcount);
780 781 782 783
}

void vtime_guest_enter(struct task_struct *tsk)
{
784 785 786 787 788 789 790
	/*
	 * The flags must be updated under the lock with
	 * the vtime_snap flush and update.
	 * That enforces a right ordering and update sequence
	 * synchronization against the reader (task_gtime())
	 * that can thus safely catch up with a tickless delta.
	 */
791
	write_seqcount_begin(&tsk->vtime_seqcount);
792 793
	if (vtime_delta(tsk))
		__vtime_account_system(tsk);
794
	current->flags |= PF_VCPU;
795
	write_seqcount_end(&tsk->vtime_seqcount);
796
}
797
EXPORT_SYMBOL_GPL(vtime_guest_enter);
798 799 800

void vtime_guest_exit(struct task_struct *tsk)
{
801
	write_seqcount_begin(&tsk->vtime_seqcount);
802 803
	__vtime_account_system(tsk);
	current->flags &= ~PF_VCPU;
804
	write_seqcount_end(&tsk->vtime_seqcount);
805
}
806
EXPORT_SYMBOL_GPL(vtime_guest_exit);
807 808 809

void vtime_account_idle(struct task_struct *tsk)
{
810
	cputime_t delta_cpu = get_vtime_delta(tsk);
811 812 813

	account_idle_time(delta_cpu);
}
814

815 816
void arch_vtime_task_switch(struct task_struct *prev)
{
817
	write_seqcount_begin(&prev->vtime_seqcount);
818
	prev->vtime_snap_whence = VTIME_INACTIVE;
819
	write_seqcount_end(&prev->vtime_seqcount);
820

821
	write_seqcount_begin(&current->vtime_seqcount);
822
	current->vtime_snap_whence = VTIME_SYS;
823
	current->vtime_snap = jiffies;
824
	write_seqcount_end(&current->vtime_seqcount);
825 826
}

827
void vtime_init_idle(struct task_struct *t, int cpu)
828 829 830
{
	unsigned long flags;

831 832
	local_irq_save(flags);
	write_seqcount_begin(&t->vtime_seqcount);
833
	t->vtime_snap_whence = VTIME_SYS;
834
	t->vtime_snap = jiffies;
835 836
	write_seqcount_end(&t->vtime_seqcount);
	local_irq_restore(flags);
837 838 839 840 841 842 843
}

cputime_t task_gtime(struct task_struct *t)
{
	unsigned int seq;
	cputime_t gtime;

844
	if (!vtime_accounting_enabled())
845 846
		return t->gtime;

847
	do {
848
		seq = read_seqcount_begin(&t->vtime_seqcount);
849 850

		gtime = t->gtime;
851
		if (t->vtime_snap_whence == VTIME_SYS && t->flags & PF_VCPU)
852 853
			gtime += vtime_delta(t);

854
	} while (read_seqcount_retry(&t->vtime_seqcount, seq));
855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876

	return gtime;
}

/*
 * Fetch cputime raw values from fields of task_struct and
 * add up the pending nohz execution time since the last
 * cputime snapshot.
 */
static void
fetch_task_cputime(struct task_struct *t,
		   cputime_t *u_dst, cputime_t *s_dst,
		   cputime_t *u_src, cputime_t *s_src,
		   cputime_t *udelta, cputime_t *sdelta)
{
	unsigned int seq;
	unsigned long long delta;

	do {
		*udelta = 0;
		*sdelta = 0;

877
		seq = read_seqcount_begin(&t->vtime_seqcount);
878 879 880 881 882 883 884

		if (u_dst)
			*u_dst = *u_src;
		if (s_dst)
			*s_dst = *s_src;

		/* Task is sleeping, nothing to add */
885
		if (t->vtime_snap_whence == VTIME_INACTIVE ||
886 887 888 889 890 891 892 893 894 895 896 897 898 899 900
		    is_idle_task(t))
			continue;

		delta = vtime_delta(t);

		/*
		 * Task runs either in user or kernel space, add pending nohz time to
		 * the right place.
		 */
		if (t->vtime_snap_whence == VTIME_USER || t->flags & PF_VCPU) {
			*udelta = delta;
		} else {
			if (t->vtime_snap_whence == VTIME_SYS)
				*sdelta = delta;
		}
901
	} while (read_seqcount_retry(&t->vtime_seqcount, seq));
902 903 904 905 906 907 908
}


void task_cputime(struct task_struct *t, cputime_t *utime, cputime_t *stime)
{
	cputime_t udelta, sdelta;

909
	if (!vtime_accounting_enabled()) {
910 911 912 913 914 915 916
		if (utime)
			*utime = t->utime;
		if (stime)
			*stime = t->stime;
		return;
	}

917 918 919 920 921 922 923 924 925 926 927 928 929
	fetch_task_cputime(t, utime, stime, &t->utime,
			   &t->stime, &udelta, &sdelta);
	if (utime)
		*utime += udelta;
	if (stime)
		*stime += sdelta;
}

void task_cputime_scaled(struct task_struct *t,
			 cputime_t *utimescaled, cputime_t *stimescaled)
{
	cputime_t udelta, sdelta;

930
	if (!vtime_accounting_enabled()) {
931 932 933 934 935 936 937
		if (utimescaled)
			*utimescaled = t->utimescaled;
		if (stimescaled)
			*stimescaled = t->stimescaled;
		return;
	}

938 939 940
	fetch_task_cputime(t, utimescaled, stimescaled,
			   &t->utimescaled, &t->stimescaled, &udelta, &sdelta);
	if (utimescaled)
941
		*utimescaled += udelta;
942
	if (stimescaled)
943
		*stimescaled += sdelta;
944
}
945
#endif /* CONFIG_VIRT_CPU_ACCOUNTING_GEN */