timer.c 36.6 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3
/*
 *  linux/kernel/timer.c
 *
4
 *  Kernel internal timers, basic process system calls
L
Linus Torvalds 已提交
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
 *
 *  Copyright (C) 1991, 1992  Linus Torvalds
 *
 *  1997-01-28  Modified by Finn Arne Gangstad to make timers scale better.
 *
 *  1997-09-10  Updated NTP code according to technical memorandum Jan '96
 *              "A Kernel Model for Precision Timekeeping" by Dave Mills
 *  1998-12-24  Fixed a xtime SMP race (we need the xtime_lock rw spinlock to
 *              serialize accesses to xtime/lost_ticks).
 *                              Copyright (C) 1998  Andrea Arcangeli
 *  1999-03-10  Improved NTP compatibility by Ulrich Windl
 *  2002-05-31	Move sys_sysinfo here and make its locking sane, Robert Love
 *  2000-10-05  Implemented scalable SMP per-CPU timer handling.
 *                              Copyright (C) 2000, 2001, 2002  Ingo Molnar
 *              Designed by David S. Miller, Alexey Kuznetsov and Ingo Molnar
 */

#include <linux/kernel_stat.h>
#include <linux/module.h>
#include <linux/interrupt.h>
#include <linux/percpu.h>
#include <linux/init.h>
#include <linux/mm.h>
#include <linux/swap.h>
29
#include <linux/pid_namespace.h>
L
Linus Torvalds 已提交
30 31 32 33 34 35 36
#include <linux/notifier.h>
#include <linux/thread_info.h>
#include <linux/time.h>
#include <linux/jiffies.h>
#include <linux/posix-timers.h>
#include <linux/cpu.h>
#include <linux/syscalls.h>
A
Adrian Bunk 已提交
37
#include <linux/delay.h>
38
#include <linux/tick.h>
39
#include <linux/kallsyms.h>
L
Linus Torvalds 已提交
40 41 42 43 44 45 46

#include <asm/uaccess.h>
#include <asm/unistd.h>
#include <asm/div64.h>
#include <asm/timex.h>
#include <asm/io.h>

T
Thomas Gleixner 已提交
47 48 49 50
u64 jiffies_64 __cacheline_aligned_in_smp = INITIAL_JIFFIES;

EXPORT_SYMBOL(jiffies_64);

L
Linus Torvalds 已提交
51 52 53 54 55 56 57 58 59 60
/*
 * per-CPU timer vector definitions:
 */
#define TVN_BITS (CONFIG_BASE_SMALL ? 4 : 6)
#define TVR_BITS (CONFIG_BASE_SMALL ? 6 : 8)
#define TVN_SIZE (1 << TVN_BITS)
#define TVR_SIZE (1 << TVR_BITS)
#define TVN_MASK (TVN_SIZE - 1)
#define TVR_MASK (TVR_SIZE - 1)

61
struct tvec {
L
Linus Torvalds 已提交
62
	struct list_head vec[TVN_SIZE];
63
};
L
Linus Torvalds 已提交
64

65
struct tvec_root {
L
Linus Torvalds 已提交
66
	struct list_head vec[TVR_SIZE];
67
};
L
Linus Torvalds 已提交
68

69
struct tvec_base {
70 71
	spinlock_t lock;
	struct timer_list *running_timer;
L
Linus Torvalds 已提交
72
	unsigned long timer_jiffies;
73 74 75 76 77
	struct tvec_root tv1;
	struct tvec tv2;
	struct tvec tv3;
	struct tvec tv4;
	struct tvec tv5;
78
} ____cacheline_aligned;
L
Linus Torvalds 已提交
79

80
struct tvec_base boot_tvec_bases;
81
EXPORT_SYMBOL(boot_tvec_bases);
82
static DEFINE_PER_CPU(struct tvec_base *, tvec_bases) = &boot_tvec_bases;
L
Linus Torvalds 已提交
83

84
/*
85
 * Note that all tvec_bases are 2 byte aligned and lower bit of
86 87 88 89 90 91
 * base in timer_list is guaranteed to be zero. Use the LSB for
 * the new flag to indicate whether the timer is deferrable
 */
#define TBASE_DEFERRABLE_FLAG		(0x1)

/* Functions below help us manage 'deferrable' flag */
92
static inline unsigned int tbase_get_deferrable(struct tvec_base *base)
93
{
94
	return ((unsigned int)(unsigned long)base & TBASE_DEFERRABLE_FLAG);
95 96
}

97
static inline struct tvec_base *tbase_get_base(struct tvec_base *base)
98
{
99
	return ((struct tvec_base *)((unsigned long)base & ~TBASE_DEFERRABLE_FLAG));
100 101 102 103
}

static inline void timer_set_deferrable(struct timer_list *timer)
{
104
	timer->base = ((struct tvec_base *)((unsigned long)(timer->base) |
105
				       TBASE_DEFERRABLE_FLAG));
106 107 108
}

static inline void
109
timer_set_base(struct timer_list *timer, struct tvec_base *new_base)
110
{
111
	timer->base = (struct tvec_base *)((unsigned long)(new_base) |
112
				      tbase_get_deferrable(timer->base));
113 114
}

115 116 117 118 119
/**
 * __round_jiffies - function to round jiffies to a full second
 * @j: the time in (absolute) jiffies that should be rounded
 * @cpu: the processor number on which the timeout will happen
 *
120
 * __round_jiffies() rounds an absolute time in the future (in jiffies)
121 122 123 124 125 126 127 128 129 130 131 132
 * up or down to (approximately) full seconds. This is useful for timers
 * for which the exact time they fire does not matter too much, as long as
 * they fire approximately every X seconds.
 *
 * By rounding these timers to whole seconds, all such timers will fire
 * at the same time, rather than at various times spread out. The goal
 * of this is to have the CPU wake up less, which saves power.
 *
 * The exact rounding is skewed for each processor to avoid all
 * processors firing at the exact same time, which could lead
 * to lock contention or spurious cache line bouncing.
 *
133
 * The return value is the rounded version of the @j parameter.
134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176
 */
unsigned long __round_jiffies(unsigned long j, int cpu)
{
	int rem;
	unsigned long original = j;

	/*
	 * We don't want all cpus firing their timers at once hitting the
	 * same lock or cachelines, so we skew each extra cpu with an extra
	 * 3 jiffies. This 3 jiffies came originally from the mm/ code which
	 * already did this.
	 * The skew is done by adding 3*cpunr, then round, then subtract this
	 * extra offset again.
	 */
	j += cpu * 3;

	rem = j % HZ;

	/*
	 * If the target jiffie is just after a whole second (which can happen
	 * due to delays of the timer irq, long irq off times etc etc) then
	 * we should round down to the whole second, not up. Use 1/4th second
	 * as cutoff for this rounding as an extreme upper bound for this.
	 */
	if (rem < HZ/4) /* round down */
		j = j - rem;
	else /* round up */
		j = j - rem + HZ;

	/* now that we have rounded, subtract the extra skew again */
	j -= cpu * 3;

	if (j <= jiffies) /* rounding ate our timeout entirely; */
		return original;
	return j;
}
EXPORT_SYMBOL_GPL(__round_jiffies);

/**
 * __round_jiffies_relative - function to round jiffies to a full second
 * @j: the time in (relative) jiffies that should be rounded
 * @cpu: the processor number on which the timeout will happen
 *
177
 * __round_jiffies_relative() rounds a time delta  in the future (in jiffies)
178 179 180 181 182 183 184 185 186 187 188 189
 * up or down to (approximately) full seconds. This is useful for timers
 * for which the exact time they fire does not matter too much, as long as
 * they fire approximately every X seconds.
 *
 * By rounding these timers to whole seconds, all such timers will fire
 * at the same time, rather than at various times spread out. The goal
 * of this is to have the CPU wake up less, which saves power.
 *
 * The exact rounding is skewed for each processor to avoid all
 * processors firing at the exact same time, which could lead
 * to lock contention or spurious cache line bouncing.
 *
190
 * The return value is the rounded version of the @j parameter.
191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207
 */
unsigned long __round_jiffies_relative(unsigned long j, int cpu)
{
	/*
	 * In theory the following code can skip a jiffy in case jiffies
	 * increments right between the addition and the later subtraction.
	 * However since the entire point of this function is to use approximate
	 * timeouts, it's entirely ok to not handle that.
	 */
	return  __round_jiffies(j + jiffies, cpu) - jiffies;
}
EXPORT_SYMBOL_GPL(__round_jiffies_relative);

/**
 * round_jiffies - function to round jiffies to a full second
 * @j: the time in (absolute) jiffies that should be rounded
 *
208
 * round_jiffies() rounds an absolute time in the future (in jiffies)
209 210 211 212 213 214 215 216
 * up or down to (approximately) full seconds. This is useful for timers
 * for which the exact time they fire does not matter too much, as long as
 * they fire approximately every X seconds.
 *
 * By rounding these timers to whole seconds, all such timers will fire
 * at the same time, rather than at various times spread out. The goal
 * of this is to have the CPU wake up less, which saves power.
 *
217
 * The return value is the rounded version of the @j parameter.
218 219 220 221 222 223 224 225 226 227 228
 */
unsigned long round_jiffies(unsigned long j)
{
	return __round_jiffies(j, raw_smp_processor_id());
}
EXPORT_SYMBOL_GPL(round_jiffies);

/**
 * round_jiffies_relative - function to round jiffies to a full second
 * @j: the time in (relative) jiffies that should be rounded
 *
229
 * round_jiffies_relative() rounds a time delta  in the future (in jiffies)
230 231 232 233 234 235 236 237
 * up or down to (approximately) full seconds. This is useful for timers
 * for which the exact time they fire does not matter too much, as long as
 * they fire approximately every X seconds.
 *
 * By rounding these timers to whole seconds, all such timers will fire
 * at the same time, rather than at various times spread out. The goal
 * of this is to have the CPU wake up less, which saves power.
 *
238
 * The return value is the rounded version of the @j parameter.
239 240 241 242 243 244 245 246
 */
unsigned long round_jiffies_relative(unsigned long j)
{
	return __round_jiffies_relative(j, raw_smp_processor_id());
}
EXPORT_SYMBOL_GPL(round_jiffies_relative);


247
static inline void set_running_timer(struct tvec_base *base,
L
Linus Torvalds 已提交
248 249 250
					struct timer_list *timer)
{
#ifdef CONFIG_SMP
251
	base->running_timer = timer;
L
Linus Torvalds 已提交
252 253 254
#endif
}

255
static void internal_add_timer(struct tvec_base *base, struct timer_list *timer)
L
Linus Torvalds 已提交
256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296
{
	unsigned long expires = timer->expires;
	unsigned long idx = expires - base->timer_jiffies;
	struct list_head *vec;

	if (idx < TVR_SIZE) {
		int i = expires & TVR_MASK;
		vec = base->tv1.vec + i;
	} else if (idx < 1 << (TVR_BITS + TVN_BITS)) {
		int i = (expires >> TVR_BITS) & TVN_MASK;
		vec = base->tv2.vec + i;
	} else if (idx < 1 << (TVR_BITS + 2 * TVN_BITS)) {
		int i = (expires >> (TVR_BITS + TVN_BITS)) & TVN_MASK;
		vec = base->tv3.vec + i;
	} else if (idx < 1 << (TVR_BITS + 3 * TVN_BITS)) {
		int i = (expires >> (TVR_BITS + 2 * TVN_BITS)) & TVN_MASK;
		vec = base->tv4.vec + i;
	} else if ((signed long) idx < 0) {
		/*
		 * Can happen if you add a timer with expires == jiffies,
		 * or you set a timer to go off in the past
		 */
		vec = base->tv1.vec + (base->timer_jiffies & TVR_MASK);
	} else {
		int i;
		/* If the timeout is larger than 0xffffffff on 64-bit
		 * architectures then we use the maximum timeout:
		 */
		if (idx > 0xffffffffUL) {
			idx = 0xffffffffUL;
			expires = idx + base->timer_jiffies;
		}
		i = (expires >> (TVR_BITS + 3 * TVN_BITS)) & TVN_MASK;
		vec = base->tv5.vec + i;
	}
	/*
	 * Timers are FIFO:
	 */
	list_add_tail(&timer->entry, vec);
}

297 298 299 300 301 302 303 304 305 306
#ifdef CONFIG_TIMER_STATS
void __timer_stats_timer_set_start_info(struct timer_list *timer, void *addr)
{
	if (timer->start_site)
		return;

	timer->start_site = addr;
	memcpy(timer->start_comm, current->comm, TASK_COMM_LEN);
	timer->start_pid = current->pid;
}
307 308 309 310 311 312 313 314 315 316 317 318 319 320

static void timer_stats_account_timer(struct timer_list *timer)
{
	unsigned int flag = 0;

	if (unlikely(tbase_get_deferrable(timer->base)))
		flag |= TIMER_STATS_FLAG_DEFERRABLE;

	timer_stats_update_stats(timer, timer->start_pid, timer->start_site,
				 timer->function, timer->start_comm, flag);
}

#else
static void timer_stats_account_timer(struct timer_list *timer) {}
321 322
#endif

323
/**
324 325 326 327 328 329
 * init_timer - initialize a timer.
 * @timer: the timer to be initialized
 *
 * init_timer() must be done to a timer prior calling *any* of the
 * other timer functions.
 */
330
void init_timer(struct timer_list *timer)
331 332
{
	timer->entry.next = NULL;
333
	timer->base = __raw_get_cpu_var(tvec_bases);
334 335 336 337 338
#ifdef CONFIG_TIMER_STATS
	timer->start_site = NULL;
	timer->start_pid = -1;
	memset(timer->start_comm, 0, TASK_COMM_LEN);
#endif
339 340 341
}
EXPORT_SYMBOL(init_timer);

342
void init_timer_deferrable(struct timer_list *timer)
343 344 345 346 347 348
{
	init_timer(timer);
	timer_set_deferrable(timer);
}
EXPORT_SYMBOL(init_timer_deferrable);

349
static inline void detach_timer(struct timer_list *timer,
350
				int clear_pending)
351 352 353 354 355 356 357 358 359 360
{
	struct list_head *entry = &timer->entry;

	__list_del(entry->prev, entry->next);
	if (clear_pending)
		entry->next = NULL;
	entry->prev = LIST_POISON2;
}

/*
361
 * We are using hashed locking: holding per_cpu(tvec_bases).lock
362 363 364 365 366 367 368 369 370 371
 * means that all timers which are tied to this base via timer->base are
 * locked, and the base itself is locked too.
 *
 * So __run_timers/migrate_timers can safely modify all timers which could
 * be found on ->tvX lists.
 *
 * When the timer's base is locked, and the timer removed from list, it is
 * possible to set timer->base = NULL and drop the lock: the timer remains
 * locked.
 */
372
static struct tvec_base *lock_timer_base(struct timer_list *timer,
373
					unsigned long *flags)
374
	__acquires(timer->base->lock)
375
{
376
	struct tvec_base *base;
377 378

	for (;;) {
379
		struct tvec_base *prelock_base = timer->base;
380
		base = tbase_get_base(prelock_base);
381 382
		if (likely(base != NULL)) {
			spin_lock_irqsave(&base->lock, *flags);
383
			if (likely(prelock_base == timer->base))
384 385 386 387 388 389 390 391
				return base;
			/* The timer has migrated to another CPU */
			spin_unlock_irqrestore(&base->lock, *flags);
		}
		cpu_relax();
	}
}

L
Linus Torvalds 已提交
392 393
int __mod_timer(struct timer_list *timer, unsigned long expires)
{
394
	struct tvec_base *base, *new_base;
L
Linus Torvalds 已提交
395 396 397
	unsigned long flags;
	int ret = 0;

398
	timer_stats_timer_set_start_info(timer);
L
Linus Torvalds 已提交
399 400
	BUG_ON(!timer->function);

401 402 403 404 405 406 407
	base = lock_timer_base(timer, &flags);

	if (timer_pending(timer)) {
		detach_timer(timer, 0);
		ret = 1;
	}

408
	new_base = __get_cpu_var(tvec_bases);
L
Linus Torvalds 已提交
409

410
	if (base != new_base) {
L
Linus Torvalds 已提交
411
		/*
412 413 414 415 416
		 * We are trying to schedule the timer on the local CPU.
		 * However we can't change timer's base while it is running,
		 * otherwise del_timer_sync() can't detect that the timer's
		 * handler yet has not finished. This also guarantees that
		 * the timer is serialized wrt itself.
L
Linus Torvalds 已提交
417
		 */
418
		if (likely(base->running_timer != timer)) {
419
			/* See the comment in lock_timer_base() */
420
			timer_set_base(timer, NULL);
421
			spin_unlock(&base->lock);
422 423
			base = new_base;
			spin_lock(&base->lock);
424
			timer_set_base(timer, base);
L
Linus Torvalds 已提交
425 426 427 428
		}
	}

	timer->expires = expires;
429 430
	internal_add_timer(base, timer);
	spin_unlock_irqrestore(&base->lock, flags);
L
Linus Torvalds 已提交
431 432 433 434 435 436

	return ret;
}

EXPORT_SYMBOL(__mod_timer);

437
/**
L
Linus Torvalds 已提交
438 439 440 441 442 443 444 445
 * add_timer_on - start a timer on a particular CPU
 * @timer: the timer to be added
 * @cpu: the CPU to start it on
 *
 * This is not very scalable on SMP. Double adds are not possible.
 */
void add_timer_on(struct timer_list *timer, int cpu)
{
446
	struct tvec_base *base = per_cpu(tvec_bases, cpu);
447
	unsigned long flags;
448

449
	timer_stats_timer_set_start_info(timer);
450
	BUG_ON(timer_pending(timer) || !timer->function);
451
	spin_lock_irqsave(&base->lock, flags);
452
	timer_set_base(timer, base);
L
Linus Torvalds 已提交
453
	internal_add_timer(base, timer);
454 455 456 457 458 459 460 461 462
	/*
	 * Check whether the other CPU is idle and needs to be
	 * triggered to reevaluate the timer wheel when nohz is
	 * active. We are protected against the other CPU fiddling
	 * with the timer by holding the timer base lock. This also
	 * makes sure that a CPU on the way to idle can not evaluate
	 * the timer wheel.
	 */
	wake_up_idle_cpu(cpu);
463
	spin_unlock_irqrestore(&base->lock, flags);
L
Linus Torvalds 已提交
464 465
}

466
/**
L
Linus Torvalds 已提交
467 468
 * mod_timer - modify a timer's timeout
 * @timer: the timer to be modified
469
 * @expires: new timeout in jiffies
L
Linus Torvalds 已提交
470
 *
471
 * mod_timer() is a more efficient way to update the expire field of an
L
Linus Torvalds 已提交
472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489
 * active timer (if the timer is inactive it will be activated)
 *
 * mod_timer(timer, expires) is equivalent to:
 *
 *     del_timer(timer); timer->expires = expires; add_timer(timer);
 *
 * Note that if there are multiple unserialized concurrent users of the
 * same timer, then mod_timer() is the only safe way to modify the timeout,
 * since add_timer() cannot modify an already running timer.
 *
 * The function returns whether it has modified a pending timer or not.
 * (ie. mod_timer() of an inactive timer returns 0, mod_timer() of an
 * active timer returns 1.)
 */
int mod_timer(struct timer_list *timer, unsigned long expires)
{
	BUG_ON(!timer->function);

490
	timer_stats_timer_set_start_info(timer);
L
Linus Torvalds 已提交
491 492 493 494 495 496 497 498 499 500 501 502 503
	/*
	 * This is a common optimization triggered by the
	 * networking code - if the timer is re-modified
	 * to be the same thing then just return:
	 */
	if (timer->expires == expires && timer_pending(timer))
		return 1;

	return __mod_timer(timer, expires);
}

EXPORT_SYMBOL(mod_timer);

504
/**
L
Linus Torvalds 已提交
505 506 507 508 509 510 511 512 513 514 515 516
 * del_timer - deactive a timer.
 * @timer: the timer to be deactivated
 *
 * del_timer() deactivates a timer - this works on both active and inactive
 * timers.
 *
 * The function returns whether it has deactivated a pending timer or not.
 * (ie. del_timer() of an inactive timer returns 0, del_timer() of an
 * active timer returns 1.)
 */
int del_timer(struct timer_list *timer)
{
517
	struct tvec_base *base;
L
Linus Torvalds 已提交
518
	unsigned long flags;
519
	int ret = 0;
L
Linus Torvalds 已提交
520

521
	timer_stats_timer_clear_start_info(timer);
522 523 524 525 526 527
	if (timer_pending(timer)) {
		base = lock_timer_base(timer, &flags);
		if (timer_pending(timer)) {
			detach_timer(timer, 1);
			ret = 1;
		}
L
Linus Torvalds 已提交
528 529 530
		spin_unlock_irqrestore(&base->lock, flags);
	}

531
	return ret;
L
Linus Torvalds 已提交
532 533 534 535 536
}

EXPORT_SYMBOL(del_timer);

#ifdef CONFIG_SMP
537 538 539 540
/**
 * try_to_del_timer_sync - Try to deactivate a timer
 * @timer: timer do del
 *
541 542 543 544 545 546 547
 * This function tries to deactivate a timer. Upon successful (ret >= 0)
 * exit the timer is not queued and the handler is not running on any CPU.
 *
 * It must not be called from interrupt contexts.
 */
int try_to_del_timer_sync(struct timer_list *timer)
{
548
	struct tvec_base *base;
549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567
	unsigned long flags;
	int ret = -1;

	base = lock_timer_base(timer, &flags);

	if (base->running_timer == timer)
		goto out;

	ret = 0;
	if (timer_pending(timer)) {
		detach_timer(timer, 1);
		ret = 1;
	}
out:
	spin_unlock_irqrestore(&base->lock, flags);

	return ret;
}

568 569
EXPORT_SYMBOL(try_to_del_timer_sync);

570
/**
L
Linus Torvalds 已提交
571 572 573 574 575 576 577
 * del_timer_sync - deactivate a timer and wait for the handler to finish.
 * @timer: the timer to be deactivated
 *
 * This function only differs from del_timer() on SMP: besides deactivating
 * the timer it also makes sure the handler has finished executing on other
 * CPUs.
 *
578
 * Synchronization rules: Callers must prevent restarting of the timer,
L
Linus Torvalds 已提交
579 580
 * otherwise this function is meaningless. It must not be called from
 * interrupt contexts. The caller must not hold locks which would prevent
581 582 583
 * completion of the timer's handler. The timer's handler must not call
 * add_timer_on(). Upon exit the timer is not queued and the handler is
 * not running on any CPU.
L
Linus Torvalds 已提交
584 585 586 587 588
 *
 * The function returns whether it has deactivated a pending timer or not.
 */
int del_timer_sync(struct timer_list *timer)
{
589 590 591 592
	for (;;) {
		int ret = try_to_del_timer_sync(timer);
		if (ret >= 0)
			return ret;
593
		cpu_relax();
594
	}
L
Linus Torvalds 已提交
595 596
}

597
EXPORT_SYMBOL(del_timer_sync);
L
Linus Torvalds 已提交
598 599
#endif

600
static int cascade(struct tvec_base *base, struct tvec *tv, int index)
L
Linus Torvalds 已提交
601 602
{
	/* cascade all the timers from tv up one level */
603 604 605 606
	struct timer_list *timer, *tmp;
	struct list_head tv_list;

	list_replace_init(tv->vec + index, &tv_list);
L
Linus Torvalds 已提交
607 608

	/*
609 610
	 * We are removing _all_ timers from the list, so we
	 * don't have to detach them individually.
L
Linus Torvalds 已提交
611
	 */
612
	list_for_each_entry_safe(timer, tmp, &tv_list, entry) {
613
		BUG_ON(tbase_get_base(timer->base) != base);
614
		internal_add_timer(base, timer);
L
Linus Torvalds 已提交
615 616 617 618 619
	}

	return index;
}

620 621 622
#define INDEX(N) ((base->timer_jiffies >> (TVR_BITS + (N) * TVN_BITS)) & TVN_MASK)

/**
L
Linus Torvalds 已提交
623 624 625 626 627 628
 * __run_timers - run all expired timers (if any) on this CPU.
 * @base: the timer vector to be processed.
 *
 * This function cascades all vectors and executes all expired timer
 * vectors.
 */
629
static inline void __run_timers(struct tvec_base *base)
L
Linus Torvalds 已提交
630 631 632
{
	struct timer_list *timer;

633
	spin_lock_irq(&base->lock);
L
Linus Torvalds 已提交
634
	while (time_after_eq(jiffies, base->timer_jiffies)) {
635
		struct list_head work_list;
L
Linus Torvalds 已提交
636
		struct list_head *head = &work_list;
637
		int index = base->timer_jiffies & TVR_MASK;
638

L
Linus Torvalds 已提交
639 640 641 642 643 644 645 646
		/*
		 * Cascade timers:
		 */
		if (!index &&
			(!cascade(base, &base->tv2, INDEX(0))) &&
				(!cascade(base, &base->tv3, INDEX(1))) &&
					!cascade(base, &base->tv4, INDEX(2)))
			cascade(base, &base->tv5, INDEX(3));
647 648
		++base->timer_jiffies;
		list_replace_init(base->tv1.vec + index, &work_list);
649
		while (!list_empty(head)) {
L
Linus Torvalds 已提交
650 651 652
			void (*fn)(unsigned long);
			unsigned long data;

653
			timer = list_first_entry(head, struct timer_list,entry);
654 655
			fn = timer->function;
			data = timer->data;
L
Linus Torvalds 已提交
656

657 658
			timer_stats_account_timer(timer);

L
Linus Torvalds 已提交
659
			set_running_timer(base, timer);
660
			detach_timer(timer, 1);
661
			spin_unlock_irq(&base->lock);
L
Linus Torvalds 已提交
662
			{
663
				int preempt_count = preempt_count();
L
Linus Torvalds 已提交
664 665
				fn(data);
				if (preempt_count != preempt_count()) {
P
Pavel Machek 已提交
666
					printk(KERN_ERR "huh, entered %p "
667 668 669 670
					       "with preempt_count %08x, exited"
					       " with %08x?\n",
					       fn, preempt_count,
					       preempt_count());
L
Linus Torvalds 已提交
671 672 673
					BUG();
				}
			}
674
			spin_lock_irq(&base->lock);
L
Linus Torvalds 已提交
675 676 677
		}
	}
	set_running_timer(base, NULL);
678
	spin_unlock_irq(&base->lock);
L
Linus Torvalds 已提交
679 680
}

681
#if defined(CONFIG_NO_IDLE_HZ) || defined(CONFIG_NO_HZ)
L
Linus Torvalds 已提交
682 683 684 685 686
/*
 * Find out when the next timer event is due to happen. This
 * is used on S/390 to stop all activity when a cpus is idle.
 * This functions needs to be called disabled.
 */
687
static unsigned long __next_timer_interrupt(struct tvec_base *base)
L
Linus Torvalds 已提交
688
{
689
	unsigned long timer_jiffies = base->timer_jiffies;
690
	unsigned long expires = timer_jiffies + NEXT_TIMER_MAX_DELTA;
691
	int index, slot, array, found = 0;
L
Linus Torvalds 已提交
692
	struct timer_list *nte;
693
	struct tvec *varray[4];
L
Linus Torvalds 已提交
694 695

	/* Look for timer events in tv1. */
696
	index = slot = timer_jiffies & TVR_MASK;
L
Linus Torvalds 已提交
697
	do {
698
		list_for_each_entry(nte, base->tv1.vec + slot, entry) {
699 700
			if (tbase_get_deferrable(nte->base))
				continue;
701

702
			found = 1;
L
Linus Torvalds 已提交
703
			expires = nte->expires;
704 705 706 707
			/* Look at the cascade bucket(s)? */
			if (!index || slot < index)
				goto cascade;
			return expires;
L
Linus Torvalds 已提交
708
		}
709 710 711 712 713 714 715 716
		slot = (slot + 1) & TVR_MASK;
	} while (slot != index);

cascade:
	/* Calculate the next cascade event */
	if (index)
		timer_jiffies += TVR_SIZE - index;
	timer_jiffies >>= TVR_BITS;
L
Linus Torvalds 已提交
717 718 719 720 721 722

	/* Check tv2-tv5. */
	varray[0] = &base->tv2;
	varray[1] = &base->tv3;
	varray[2] = &base->tv4;
	varray[3] = &base->tv5;
723 724

	for (array = 0; array < 4; array++) {
725
		struct tvec *varp = varray[array];
726 727

		index = slot = timer_jiffies & TVN_MASK;
L
Linus Torvalds 已提交
728
		do {
729 730
			list_for_each_entry(nte, varp->vec + slot, entry) {
				found = 1;
L
Linus Torvalds 已提交
731 732
				if (time_before(nte->expires, expires))
					expires = nte->expires;
733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749
			}
			/*
			 * Do we still search for the first timer or are
			 * we looking up the cascade buckets ?
			 */
			if (found) {
				/* Look at the cascade bucket(s)? */
				if (!index || slot < index)
					break;
				return expires;
			}
			slot = (slot + 1) & TVN_MASK;
		} while (slot != index);

		if (index)
			timer_jiffies += TVN_SIZE - index;
		timer_jiffies >>= TVN_BITS;
L
Linus Torvalds 已提交
750
	}
751 752
	return expires;
}
753

754 755 756 757 758 759 760 761 762
/*
 * Check, if the next hrtimer event is before the next timer wheel
 * event:
 */
static unsigned long cmp_next_hrtimer_event(unsigned long now,
					    unsigned long expires)
{
	ktime_t hr_delta = hrtimer_get_next_event();
	struct timespec tsdelta;
763
	unsigned long delta;
764 765 766

	if (hr_delta.tv64 == KTIME_MAX)
		return expires;
767

768 769 770 771 772
	/*
	 * Expired timer available, let it expire in the next tick
	 */
	if (hr_delta.tv64 <= 0)
		return now + 1;
773

774
	tsdelta = ktime_to_timespec(hr_delta);
775
	delta = timespec_to_jiffies(&tsdelta);
776 777 778 779 780 781 782 783

	/*
	 * Limit the delta to the max value, which is checked in
	 * tick_nohz_stop_sched_tick():
	 */
	if (delta > NEXT_TIMER_MAX_DELTA)
		delta = NEXT_TIMER_MAX_DELTA;

784 785 786 787 788 789 790 791 792
	/*
	 * Take rounding errors in to account and make sure, that it
	 * expires in the next tick. Otherwise we go into an endless
	 * ping pong due to tick_nohz_stop_sched_tick() retriggering
	 * the timer softirq
	 */
	if (delta < 1)
		delta = 1;
	now += delta;
793 794
	if (time_before(now, expires))
		return now;
L
Linus Torvalds 已提交
795 796
	return expires;
}
797 798

/**
799
 * get_next_timer_interrupt - return the jiffy of the next pending timer
800
 * @now: current time (in jiffies)
801
 */
802
unsigned long get_next_timer_interrupt(unsigned long now)
803
{
804
	struct tvec_base *base = __get_cpu_var(tvec_bases);
805
	unsigned long expires;
806 807 808 809 810 811 812 813 814 815

	spin_lock(&base->lock);
	expires = __next_timer_interrupt(base);
	spin_unlock(&base->lock);

	if (time_before_eq(expires, now))
		return now;

	return cmp_next_hrtimer_event(now, expires);
}
816 817 818 819 820 821 822 823

#ifdef CONFIG_NO_IDLE_HZ
unsigned long next_timer_interrupt(void)
{
	return get_next_timer_interrupt(jiffies);
}
#endif

L
Linus Torvalds 已提交
824 825
#endif

826 827 828
#ifndef CONFIG_VIRT_CPU_ACCOUNTING
void account_process_tick(struct task_struct *p, int user_tick)
{
M
Michael Neuling 已提交
829 830
	cputime_t one_jiffy = jiffies_to_cputime(1);

831
	if (user_tick) {
M
Michael Neuling 已提交
832 833
		account_user_time(p, one_jiffy);
		account_user_time_scaled(p, cputime_to_scaled(one_jiffy));
834
	} else {
M
Michael Neuling 已提交
835 836
		account_system_time(p, HARDIRQ_OFFSET, one_jiffy);
		account_system_time_scaled(p, cputime_to_scaled(one_jiffy));
837 838 839 840
	}
}
#endif

L
Linus Torvalds 已提交
841
/*
D
Daniel Walker 已提交
842
 * Called from the timer interrupt handler to charge one tick to the current
L
Linus Torvalds 已提交
843 844 845 846 847 848 849 850
 * process.  user_tick is 1 if the tick is user time, 0 for system.
 */
void update_process_times(int user_tick)
{
	struct task_struct *p = current;
	int cpu = smp_processor_id();

	/* Note: this timer irq context must be accounted for as well. */
851
	account_process_tick(p, user_tick);
L
Linus Torvalds 已提交
852 853 854 855
	run_local_timers();
	if (rcu_pending(cpu))
		rcu_check_callbacks(cpu, user_tick);
	scheduler_tick();
856
	run_posix_cpu_timers(p);
L
Linus Torvalds 已提交
857 858 859 860 861 862 863
}

/*
 * Nr of active tasks - counted in fixed-point numbers
 */
static unsigned long count_active_tasks(void)
{
864
	return nr_active() * FIXED_1;
L
Linus Torvalds 已提交
865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887
}

/*
 * Hmm.. Changed this, as the GNU make sources (load.c) seems to
 * imply that avenrun[] is the standard name for this kind of thing.
 * Nothing else seems to be standardized: the fractional size etc
 * all seem to differ on different machines.
 *
 * Requires xtime_lock to access.
 */
unsigned long avenrun[3];

EXPORT_SYMBOL(avenrun);

/*
 * calc_load - given tick count, update the avenrun load estimates.
 * This is called while holding a write_lock on xtime_lock.
 */
static inline void calc_load(unsigned long ticks)
{
	unsigned long active_tasks; /* fixed-point */
	static int count = LOAD_FREQ;

E
Eric Dumazet 已提交
888 889 890 891 892 893 894 895 896
	count -= ticks;
	if (unlikely(count < 0)) {
		active_tasks = count_active_tasks();
		do {
			CALC_LOAD(avenrun[0], EXP_1, active_tasks);
			CALC_LOAD(avenrun[1], EXP_5, active_tasks);
			CALC_LOAD(avenrun[2], EXP_15, active_tasks);
			count += LOAD_FREQ;
		} while (count < 0);
L
Linus Torvalds 已提交
897 898 899 900 901 902 903 904
	}
}

/*
 * This function runs timers and the timer-tq in bottom half context.
 */
static void run_timer_softirq(struct softirq_action *h)
{
905
	struct tvec_base *base = __get_cpu_var(tvec_bases);
L
Linus Torvalds 已提交
906

907
	hrtimer_run_pending();
908

L
Linus Torvalds 已提交
909 910 911 912 913 914 915 916 917
	if (time_after_eq(jiffies, base->timer_jiffies))
		__run_timers(base);
}

/*
 * Called by the local, per-CPU timer interrupt on SMP.
 */
void run_local_timers(void)
{
918
	hrtimer_run_queues();
L
Linus Torvalds 已提交
919
	raise_softirq(TIMER_SOFTIRQ);
920
	softlockup_tick();
L
Linus Torvalds 已提交
921 922 923 924 925 926
}

/*
 * Called by the timer interrupt. xtime_lock must already be taken
 * by the timer IRQ!
 */
927
static inline void update_times(unsigned long ticks)
L
Linus Torvalds 已提交
928
{
929
	update_wall_time();
L
Linus Torvalds 已提交
930 931
	calc_load(ticks);
}
932

L
Linus Torvalds 已提交
933 934 935 936 937 938
/*
 * The 64-bit jiffies value is not atomic - you MUST NOT read it
 * without sampling the sequence number in xtime_lock.
 * jiffies is defined in the linker script...
 */

939
void do_timer(unsigned long ticks)
L
Linus Torvalds 已提交
940
{
941 942
	jiffies_64 += ticks;
	update_times(ticks);
L
Linus Torvalds 已提交
943 944 945 946 947 948 949 950 951 952
}

#ifdef __ARCH_WANT_SYS_ALARM

/*
 * For backwards compatibility?  This can be done in libc so Alpha
 * and all newer ports shouldn't need it.
 */
asmlinkage unsigned long sys_alarm(unsigned int seconds)
{
953
	return alarm_setitimer(seconds);
L
Linus Torvalds 已提交
954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975
}

#endif

#ifndef __alpha__

/*
 * The Alpha uses getxpid, getxuid, and getxgid instead.  Maybe this
 * should be moved into arch/i386 instead?
 */

/**
 * sys_getpid - return the thread group id of the current process
 *
 * Note, despite the name, this returns the tgid not the pid.  The tgid and
 * the pid are identical unless CLONE_THREAD was specified on clone() in
 * which case the tgid is the same in all threads of the same group.
 *
 * This is SMP safe as current->tgid does not change.
 */
asmlinkage long sys_getpid(void)
{
976
	return task_tgid_vnr(current);
L
Linus Torvalds 已提交
977 978 979
}

/*
980 981 982 983
 * Accessing ->real_parent is not SMP-safe, it could
 * change from under us. However, we can use a stale
 * value of ->real_parent under rcu_read_lock(), see
 * release_task()->call_rcu(delayed_put_task_struct).
L
Linus Torvalds 已提交
984 985 986 987 988
 */
asmlinkage long sys_getppid(void)
{
	int pid;

989
	rcu_read_lock();
990
	pid = task_tgid_vnr(current->real_parent);
991
	rcu_read_unlock();
L
Linus Torvalds 已提交
992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023

	return pid;
}

asmlinkage long sys_getuid(void)
{
	/* Only we change this so SMP safe */
	return current->uid;
}

asmlinkage long sys_geteuid(void)
{
	/* Only we change this so SMP safe */
	return current->euid;
}

asmlinkage long sys_getgid(void)
{
	/* Only we change this so SMP safe */
	return current->gid;
}

asmlinkage long sys_getegid(void)
{
	/* Only we change this so SMP safe */
	return  current->egid;
}

#endif

static void process_timeout(unsigned long __data)
{
1024
	wake_up_process((struct task_struct *)__data);
L
Linus Torvalds 已提交
1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052
}

/**
 * schedule_timeout - sleep until timeout
 * @timeout: timeout value in jiffies
 *
 * Make the current task sleep until @timeout jiffies have
 * elapsed. The routine will return immediately unless
 * the current task state has been set (see set_current_state()).
 *
 * You can set the task state as follows -
 *
 * %TASK_UNINTERRUPTIBLE - at least @timeout jiffies are guaranteed to
 * pass before the routine returns. The routine will return 0
 *
 * %TASK_INTERRUPTIBLE - the routine may return early if a signal is
 * delivered to the current task. In this case the remaining time
 * in jiffies will be returned, or 0 if the timer expired in time
 *
 * The current task state is guaranteed to be TASK_RUNNING when this
 * routine returns.
 *
 * Specifying a @timeout value of %MAX_SCHEDULE_TIMEOUT will schedule
 * the CPU away without a bound on the timeout. In this case the return
 * value will be %MAX_SCHEDULE_TIMEOUT.
 *
 * In all cases the return value is guaranteed to be non-negative.
 */
1053
signed long __sched schedule_timeout(signed long timeout)
L
Linus Torvalds 已提交
1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077
{
	struct timer_list timer;
	unsigned long expire;

	switch (timeout)
	{
	case MAX_SCHEDULE_TIMEOUT:
		/*
		 * These two special cases are useful to be comfortable
		 * in the caller. Nothing more. We could take
		 * MAX_SCHEDULE_TIMEOUT from one of the negative value
		 * but I' d like to return a valid offset (>=0) to allow
		 * the caller to do everything it want with the retval.
		 */
		schedule();
		goto out;
	default:
		/*
		 * Another bit of PARANOID. Note that the retval will be
		 * 0 since no piece of kernel is supposed to do a check
		 * for a negative retval of schedule_timeout() (since it
		 * should never happens anyway). You just have the printk()
		 * that will tell you if something is gone wrong and where.
		 */
1078
		if (timeout < 0) {
L
Linus Torvalds 已提交
1079
			printk(KERN_ERR "schedule_timeout: wrong timeout "
1080 1081
				"value %lx\n", timeout);
			dump_stack();
L
Linus Torvalds 已提交
1082 1083 1084 1085 1086 1087 1088
			current->state = TASK_RUNNING;
			goto out;
		}
	}

	expire = timeout + jiffies;

1089 1090
	setup_timer(&timer, process_timeout, (unsigned long)current);
	__mod_timer(&timer, expire);
L
Linus Torvalds 已提交
1091 1092 1093 1094 1095 1096 1097 1098 1099 1100
	schedule();
	del_singleshot_timer_sync(&timer);

	timeout = expire - jiffies;

 out:
	return timeout < 0 ? 0 : timeout;
}
EXPORT_SYMBOL(schedule_timeout);

1101 1102 1103 1104
/*
 * We can use __set_current_state() here because schedule_timeout() calls
 * schedule() unconditionally.
 */
1105 1106
signed long __sched schedule_timeout_interruptible(signed long timeout)
{
A
Andrew Morton 已提交
1107 1108
	__set_current_state(TASK_INTERRUPTIBLE);
	return schedule_timeout(timeout);
1109 1110 1111
}
EXPORT_SYMBOL(schedule_timeout_interruptible);

M
Matthew Wilcox 已提交
1112 1113 1114 1115 1116 1117 1118
signed long __sched schedule_timeout_killable(signed long timeout)
{
	__set_current_state(TASK_KILLABLE);
	return schedule_timeout(timeout);
}
EXPORT_SYMBOL(schedule_timeout_killable);

1119 1120
signed long __sched schedule_timeout_uninterruptible(signed long timeout)
{
A
Andrew Morton 已提交
1121 1122
	__set_current_state(TASK_UNINTERRUPTIBLE);
	return schedule_timeout(timeout);
1123 1124 1125
}
EXPORT_SYMBOL(schedule_timeout_uninterruptible);

L
Linus Torvalds 已提交
1126 1127 1128
/* Thread ID - the internal kernel "pid" */
asmlinkage long sys_gettid(void)
{
1129
	return task_pid_vnr(current);
L
Linus Torvalds 已提交
1130 1131
}

1132
/**
1133
 * do_sysinfo - fill in sysinfo struct
1134
 * @info: pointer to buffer to fill
1135
 */
1136
int do_sysinfo(struct sysinfo *info)
L
Linus Torvalds 已提交
1137 1138 1139 1140 1141
{
	unsigned long mem_total, sav_total;
	unsigned int mem_unit, bitcount;
	unsigned long seq;

1142
	memset(info, 0, sizeof(struct sysinfo));
L
Linus Torvalds 已提交
1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157

	do {
		struct timespec tp;
		seq = read_seqbegin(&xtime_lock);

		/*
		 * This is annoying.  The below is the same thing
		 * posix_get_clock_monotonic() does, but it wants to
		 * take the lock which we want to cover the loads stuff
		 * too.
		 */

		getnstimeofday(&tp);
		tp.tv_sec += wall_to_monotonic.tv_sec;
		tp.tv_nsec += wall_to_monotonic.tv_nsec;
1158
		monotonic_to_bootbased(&tp);
L
Linus Torvalds 已提交
1159 1160 1161 1162
		if (tp.tv_nsec - NSEC_PER_SEC >= 0) {
			tp.tv_nsec = tp.tv_nsec - NSEC_PER_SEC;
			tp.tv_sec++;
		}
1163
		info->uptime = tp.tv_sec + (tp.tv_nsec ? 1 : 0);
L
Linus Torvalds 已提交
1164

1165 1166 1167
		info->loads[0] = avenrun[0] << (SI_LOAD_SHIFT - FSHIFT);
		info->loads[1] = avenrun[1] << (SI_LOAD_SHIFT - FSHIFT);
		info->loads[2] = avenrun[2] << (SI_LOAD_SHIFT - FSHIFT);
L
Linus Torvalds 已提交
1168

1169
		info->procs = nr_threads;
L
Linus Torvalds 已提交
1170 1171
	} while (read_seqretry(&xtime_lock, seq));

1172 1173
	si_meminfo(info);
	si_swapinfo(info);
L
Linus Torvalds 已提交
1174 1175 1176 1177 1178 1179 1180 1181 1182 1183

	/*
	 * If the sum of all the available memory (i.e. ram + swap)
	 * is less than can be stored in a 32 bit unsigned long then
	 * we can be binary compatible with 2.2.x kernels.  If not,
	 * well, in that case 2.2.x was broken anyways...
	 *
	 *  -Erik Andersen <andersee@debian.org>
	 */

1184 1185
	mem_total = info->totalram + info->totalswap;
	if (mem_total < info->totalram || mem_total < info->totalswap)
L
Linus Torvalds 已提交
1186 1187
		goto out;
	bitcount = 0;
1188
	mem_unit = info->mem_unit;
L
Linus Torvalds 已提交
1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199
	while (mem_unit > 1) {
		bitcount++;
		mem_unit >>= 1;
		sav_total = mem_total;
		mem_total <<= 1;
		if (mem_total < sav_total)
			goto out;
	}

	/*
	 * If mem_total did not overflow, multiply all memory values by
1200
	 * info->mem_unit and set it to 1.  This leaves things compatible
L
Linus Torvalds 已提交
1201 1202 1203 1204
	 * with 2.2.x, and also retains compatibility with earlier 2.4.x
	 * kernels...
	 */

1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223
	info->mem_unit = 1;
	info->totalram <<= bitcount;
	info->freeram <<= bitcount;
	info->sharedram <<= bitcount;
	info->bufferram <<= bitcount;
	info->totalswap <<= bitcount;
	info->freeswap <<= bitcount;
	info->totalhigh <<= bitcount;
	info->freehigh <<= bitcount;

out:
	return 0;
}

asmlinkage long sys_sysinfo(struct sysinfo __user *info)
{
	struct sysinfo val;

	do_sysinfo(&val);
L
Linus Torvalds 已提交
1224 1225 1226 1227 1228 1229 1230

	if (copy_to_user(info, &val, sizeof(struct sysinfo)))
		return -EFAULT;

	return 0;
}

1231 1232 1233 1234 1235 1236 1237
/*
 * lockdep: we want to track each per-CPU base as a separate lock-class,
 * but timer-bases are kmalloc()-ed, so we need to attach separate
 * keys to them:
 */
static struct lock_class_key base_lock_keys[NR_CPUS];

A
Adrian Bunk 已提交
1238
static int __cpuinit init_timers_cpu(int cpu)
L
Linus Torvalds 已提交
1239 1240
{
	int j;
1241
	struct tvec_base *base;
A
Adrian Bunk 已提交
1242
	static char __cpuinitdata tvec_base_done[NR_CPUS];
1243

A
Andrew Morton 已提交
1244
	if (!tvec_base_done[cpu]) {
1245 1246 1247
		static char boot_done;

		if (boot_done) {
A
Andrew Morton 已提交
1248 1249 1250
			/*
			 * The APs use this path later in boot
			 */
1251 1252
			base = kmalloc_node(sizeof(*base),
						GFP_KERNEL | __GFP_ZERO,
1253 1254 1255
						cpu_to_node(cpu));
			if (!base)
				return -ENOMEM;
1256 1257 1258 1259 1260 1261 1262

			/* Make sure that tvec_base is 2 byte aligned */
			if (tbase_get_deferrable(base)) {
				WARN_ON(1);
				kfree(base);
				return -ENOMEM;
			}
A
Andrew Morton 已提交
1263
			per_cpu(tvec_bases, cpu) = base;
1264
		} else {
A
Andrew Morton 已提交
1265 1266 1267 1268 1269 1270
			/*
			 * This is for the boot CPU - we use compile-time
			 * static initialisation because per-cpu memory isn't
			 * ready yet and because the memory allocators are not
			 * initialised either.
			 */
1271
			boot_done = 1;
A
Andrew Morton 已提交
1272
			base = &boot_tvec_bases;
1273
		}
A
Andrew Morton 已提交
1274 1275 1276
		tvec_base_done[cpu] = 1;
	} else {
		base = per_cpu(tvec_bases, cpu);
1277
	}
A
Andrew Morton 已提交
1278

1279
	spin_lock_init(&base->lock);
1280 1281
	lockdep_set_class(&base->lock, base_lock_keys + cpu);

L
Linus Torvalds 已提交
1282 1283 1284 1285 1286 1287 1288 1289 1290 1291
	for (j = 0; j < TVN_SIZE; j++) {
		INIT_LIST_HEAD(base->tv5.vec + j);
		INIT_LIST_HEAD(base->tv4.vec + j);
		INIT_LIST_HEAD(base->tv3.vec + j);
		INIT_LIST_HEAD(base->tv2.vec + j);
	}
	for (j = 0; j < TVR_SIZE; j++)
		INIT_LIST_HEAD(base->tv1.vec + j);

	base->timer_jiffies = jiffies;
1292
	return 0;
L
Linus Torvalds 已提交
1293 1294 1295
}

#ifdef CONFIG_HOTPLUG_CPU
1296
static void migrate_timer_list(struct tvec_base *new_base, struct list_head *head)
L
Linus Torvalds 已提交
1297 1298 1299 1300
{
	struct timer_list *timer;

	while (!list_empty(head)) {
1301
		timer = list_first_entry(head, struct timer_list, entry);
1302
		detach_timer(timer, 0);
1303
		timer_set_base(timer, new_base);
L
Linus Torvalds 已提交
1304 1305 1306 1307
		internal_add_timer(new_base, timer);
	}
}

R
Randy Dunlap 已提交
1308
static void __cpuinit migrate_timers(int cpu)
L
Linus Torvalds 已提交
1309
{
1310 1311
	struct tvec_base *old_base;
	struct tvec_base *new_base;
L
Linus Torvalds 已提交
1312 1313 1314
	int i;

	BUG_ON(cpu_online(cpu));
1315 1316
	old_base = per_cpu(tvec_bases, cpu);
	new_base = get_cpu_var(tvec_bases);
L
Linus Torvalds 已提交
1317 1318

	local_irq_disable();
1319 1320
	double_spin_lock(&new_base->lock, &old_base->lock,
			 smp_processor_id() < cpu);
1321 1322

	BUG_ON(old_base->running_timer);
L
Linus Torvalds 已提交
1323 1324

	for (i = 0; i < TVR_SIZE; i++)
1325 1326 1327 1328 1329 1330 1331 1332
		migrate_timer_list(new_base, old_base->tv1.vec + i);
	for (i = 0; i < TVN_SIZE; i++) {
		migrate_timer_list(new_base, old_base->tv2.vec + i);
		migrate_timer_list(new_base, old_base->tv3.vec + i);
		migrate_timer_list(new_base, old_base->tv4.vec + i);
		migrate_timer_list(new_base, old_base->tv5.vec + i);
	}

1333 1334
	double_spin_unlock(&new_base->lock, &old_base->lock,
			   smp_processor_id() < cpu);
L
Linus Torvalds 已提交
1335 1336 1337 1338 1339
	local_irq_enable();
	put_cpu_var(tvec_bases);
}
#endif /* CONFIG_HOTPLUG_CPU */

1340
static int __cpuinit timer_cpu_notify(struct notifier_block *self,
L
Linus Torvalds 已提交
1341 1342 1343 1344 1345
				unsigned long action, void *hcpu)
{
	long cpu = (long)hcpu;
	switch(action) {
	case CPU_UP_PREPARE:
1346
	case CPU_UP_PREPARE_FROZEN:
1347 1348
		if (init_timers_cpu(cpu) < 0)
			return NOTIFY_BAD;
L
Linus Torvalds 已提交
1349 1350 1351
		break;
#ifdef CONFIG_HOTPLUG_CPU
	case CPU_DEAD:
1352
	case CPU_DEAD_FROZEN:
L
Linus Torvalds 已提交
1353 1354 1355 1356 1357 1358 1359 1360 1361
		migrate_timers(cpu);
		break;
#endif
	default:
		break;
	}
	return NOTIFY_OK;
}

1362
static struct notifier_block __cpuinitdata timers_nb = {
L
Linus Torvalds 已提交
1363 1364 1365 1366 1367 1368
	.notifier_call	= timer_cpu_notify,
};


void __init init_timers(void)
{
1369
	int err = timer_cpu_notify(&timers_nb, (unsigned long)CPU_UP_PREPARE,
L
Linus Torvalds 已提交
1370
				(void *)(long)smp_processor_id());
1371

1372 1373
	init_timer_stats();

1374
	BUG_ON(err == NOTIFY_BAD);
L
Linus Torvalds 已提交
1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386
	register_cpu_notifier(&timers_nb);
	open_softirq(TIMER_SOFTIRQ, run_timer_softirq, NULL);
}

/**
 * msleep - sleep safely even with waitqueue interruptions
 * @msecs: Time in milliseconds to sleep for
 */
void msleep(unsigned int msecs)
{
	unsigned long timeout = msecs_to_jiffies(msecs) + 1;

1387 1388
	while (timeout)
		timeout = schedule_timeout_uninterruptible(timeout);
L
Linus Torvalds 已提交
1389 1390 1391 1392 1393
}

EXPORT_SYMBOL(msleep);

/**
1394
 * msleep_interruptible - sleep waiting for signals
L
Linus Torvalds 已提交
1395 1396 1397 1398 1399 1400
 * @msecs: Time in milliseconds to sleep for
 */
unsigned long msleep_interruptible(unsigned int msecs)
{
	unsigned long timeout = msecs_to_jiffies(msecs) + 1;

1401 1402
	while (timeout && !signal_pending(current))
		timeout = schedule_timeout_interruptible(timeout);
L
Linus Torvalds 已提交
1403 1404 1405 1406
	return jiffies_to_msecs(timeout);
}

EXPORT_SYMBOL(msleep_interruptible);