timer.c 47.7 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>
40
#include <linux/irq_work.h>
41
#include <linux/sched.h>
42
#include <linux/slab.h>
L
Linus Torvalds 已提交
43 44 45 46 47 48 49

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

50 51 52
#define CREATE_TRACE_POINTS
#include <trace/events/timer.h>

T
Thomas Gleixner 已提交
53 54 55 56
u64 jiffies_64 __cacheline_aligned_in_smp = INITIAL_JIFFIES;

EXPORT_SYMBOL(jiffies_64);

L
Linus Torvalds 已提交
57 58 59 60 61 62 63 64 65 66
/*
 * 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)

67
struct tvec {
L
Linus Torvalds 已提交
68
	struct list_head vec[TVN_SIZE];
69
};
L
Linus Torvalds 已提交
70

71
struct tvec_root {
L
Linus Torvalds 已提交
72
	struct list_head vec[TVR_SIZE];
73
};
L
Linus Torvalds 已提交
74

75
struct tvec_base {
76 77
	spinlock_t lock;
	struct timer_list *running_timer;
L
Linus Torvalds 已提交
78
	unsigned long timer_jiffies;
79
	unsigned long next_timer;
80 81 82 83 84
	struct tvec_root tv1;
	struct tvec tv2;
	struct tvec tv3;
	struct tvec tv4;
	struct tvec tv5;
85
} ____cacheline_aligned;
L
Linus Torvalds 已提交
86

87
struct tvec_base boot_tvec_bases;
88
EXPORT_SYMBOL(boot_tvec_bases);
89
static DEFINE_PER_CPU(struct tvec_base *, tvec_bases) = &boot_tvec_bases;
L
Linus Torvalds 已提交
90

91
/*
92
 * Note that all tvec_bases are 2 byte aligned and lower bit of
93 94 95 96 97 98 99
 * base in timer_list is guaranteed to be zero. Use the LSB to
 * indicate whether the timer is deferrable.
 *
 * A deferrable timer will work normally when the system is busy, but
 * will not cause a CPU to come out of idle just to service it; instead,
 * the timer will be serviced when the CPU eventually wakes up with a
 * subsequent non-deferrable timer.
100 101 102 103
 */
#define TBASE_DEFERRABLE_FLAG		(0x1)

/* Functions below help us manage 'deferrable' flag */
104
static inline unsigned int tbase_get_deferrable(struct tvec_base *base)
105
{
106
	return ((unsigned int)(unsigned long)base & TBASE_DEFERRABLE_FLAG);
107 108
}

109
static inline struct tvec_base *tbase_get_base(struct tvec_base *base)
110
{
111
	return ((struct tvec_base *)((unsigned long)base & ~TBASE_DEFERRABLE_FLAG));
112 113 114 115
}

static inline void timer_set_deferrable(struct timer_list *timer)
{
116
	timer->base = ((struct tvec_base *)((unsigned long)(timer->base) |
117
				       TBASE_DEFERRABLE_FLAG));
118 119 120
}

static inline void
121
timer_set_base(struct timer_list *timer, struct tvec_base *new_base)
122
{
123
	timer->base = (struct tvec_base *)((unsigned long)(new_base) |
124
				      tbase_get_deferrable(timer->base));
125 126
}

127 128
static unsigned long round_jiffies_common(unsigned long j, int cpu,
		bool force_up)
129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149
{
	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.
150
	 * But never round down if @force_up is set.
151
	 */
152
	if (rem < HZ/4 && !force_up) /* round down */
153 154 155 156 157 158 159 160 161 162 163
		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;
}
164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188

/**
 * __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
 *
 * __round_jiffies() rounds an absolute time in the future (in jiffies)
 * 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.
 *
 * The return value is the rounded version of the @j parameter.
 */
unsigned long __round_jiffies(unsigned long j, int cpu)
{
	return round_jiffies_common(j, cpu, false);
}
189 190 191 192 193 194 195
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
 *
196
 * __round_jiffies_relative() rounds a time delta  in the future (in jiffies)
197 198 199 200 201 202 203 204 205 206 207 208
 * 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.
 *
209
 * The return value is the rounded version of the @j parameter.
210 211 212
 */
unsigned long __round_jiffies_relative(unsigned long j, int cpu)
{
213 214 215 216
	unsigned long j0 = jiffies;

	/* Use j0 because jiffies might change while we run */
	return round_jiffies_common(j + j0, cpu, false) - j0;
217 218 219 220 221 222 223
}
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
 *
224
 * round_jiffies() rounds an absolute time in the future (in jiffies)
225 226 227 228 229 230 231 232
 * 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.
 *
233
 * The return value is the rounded version of the @j parameter.
234 235 236
 */
unsigned long round_jiffies(unsigned long j)
{
237
	return round_jiffies_common(j, raw_smp_processor_id(), false);
238 239 240 241 242 243 244
}
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
 *
245
 * round_jiffies_relative() rounds a time delta  in the future (in jiffies)
246 247 248 249 250 251 252 253
 * 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.
 *
254
 * The return value is the rounded version of the @j parameter.
255 256 257 258 259 260 261
 */
unsigned long round_jiffies_relative(unsigned long j)
{
	return __round_jiffies_relative(j, raw_smp_processor_id());
}
EXPORT_SYMBOL_GPL(round_jiffies_relative);

262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326
/**
 * __round_jiffies_up - function to round jiffies up to a full second
 * @j: the time in (absolute) jiffies that should be rounded
 * @cpu: the processor number on which the timeout will happen
 *
 * This is the same as __round_jiffies() except that it will never
 * round down.  This is useful for timeouts for which the exact time
 * of firing does not matter too much, as long as they don't fire too
 * early.
 */
unsigned long __round_jiffies_up(unsigned long j, int cpu)
{
	return round_jiffies_common(j, cpu, true);
}
EXPORT_SYMBOL_GPL(__round_jiffies_up);

/**
 * __round_jiffies_up_relative - function to round jiffies up to a full second
 * @j: the time in (relative) jiffies that should be rounded
 * @cpu: the processor number on which the timeout will happen
 *
 * This is the same as __round_jiffies_relative() except that it will never
 * round down.  This is useful for timeouts for which the exact time
 * of firing does not matter too much, as long as they don't fire too
 * early.
 */
unsigned long __round_jiffies_up_relative(unsigned long j, int cpu)
{
	unsigned long j0 = jiffies;

	/* Use j0 because jiffies might change while we run */
	return round_jiffies_common(j + j0, cpu, true) - j0;
}
EXPORT_SYMBOL_GPL(__round_jiffies_up_relative);

/**
 * round_jiffies_up - function to round jiffies up to a full second
 * @j: the time in (absolute) jiffies that should be rounded
 *
 * This is the same as round_jiffies() except that it will never
 * round down.  This is useful for timeouts for which the exact time
 * of firing does not matter too much, as long as they don't fire too
 * early.
 */
unsigned long round_jiffies_up(unsigned long j)
{
	return round_jiffies_common(j, raw_smp_processor_id(), true);
}
EXPORT_SYMBOL_GPL(round_jiffies_up);

/**
 * round_jiffies_up_relative - function to round jiffies up to a full second
 * @j: the time in (relative) jiffies that should be rounded
 *
 * This is the same as round_jiffies_relative() except that it will never
 * round down.  This is useful for timeouts for which the exact time
 * of firing does not matter too much, as long as they don't fire too
 * early.
 */
unsigned long round_jiffies_up_relative(unsigned long j)
{
	return __round_jiffies_up_relative(j, raw_smp_processor_id());
}
EXPORT_SYMBOL_GPL(round_jiffies_up_relative);

327 328
/**
 * set_timer_slack - set the allowed slack for a timer
329
 * @timer: the timer to be modified
330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345
 * @slack_hz: the amount of time (in jiffies) allowed for rounding
 *
 * Set the amount of time, in jiffies, that a certain timer has
 * in terms of slack. By setting this value, the timer subsystem
 * will schedule the actual timer somewhere between
 * the time mod_timer() asks for, and that time plus the slack.
 *
 * By setting the slack to -1, a percentage of the delay is used
 * instead.
 */
void set_timer_slack(struct timer_list *timer, int slack_hz)
{
	timer->slack = slack_hz;
}
EXPORT_SYMBOL_GPL(set_timer_slack);

346

347
static inline void set_running_timer(struct tvec_base *base,
L
Linus Torvalds 已提交
348 349 350
					struct timer_list *timer)
{
#ifdef CONFIG_SMP
351
	base->running_timer = timer;
L
Linus Torvalds 已提交
352 353 354
#endif
}

355
static void internal_add_timer(struct tvec_base *base, struct timer_list *timer)
L
Linus Torvalds 已提交
356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396
{
	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);
}

397 398 399 400 401 402 403 404 405 406
#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;
}
407 408 409 410 411

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

412 413
	if (likely(!timer->start_site))
		return;
414 415 416 417 418 419 420 421 422
	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) {}
423 424
#endif

425 426 427 428 429 430 431
#ifdef CONFIG_DEBUG_OBJECTS_TIMERS

static struct debug_obj_descr timer_debug_descr;

/*
 * fixup_init is called when:
 * - an active object is initialized
432
 */
433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526
static int timer_fixup_init(void *addr, enum debug_obj_state state)
{
	struct timer_list *timer = addr;

	switch (state) {
	case ODEBUG_STATE_ACTIVE:
		del_timer_sync(timer);
		debug_object_init(timer, &timer_debug_descr);
		return 1;
	default:
		return 0;
	}
}

/*
 * fixup_activate is called when:
 * - an active object is activated
 * - an unknown object is activated (might be a statically initialized object)
 */
static int timer_fixup_activate(void *addr, enum debug_obj_state state)
{
	struct timer_list *timer = addr;

	switch (state) {

	case ODEBUG_STATE_NOTAVAILABLE:
		/*
		 * This is not really a fixup. The timer was
		 * statically initialized. We just make sure that it
		 * is tracked in the object tracker.
		 */
		if (timer->entry.next == NULL &&
		    timer->entry.prev == TIMER_ENTRY_STATIC) {
			debug_object_init(timer, &timer_debug_descr);
			debug_object_activate(timer, &timer_debug_descr);
			return 0;
		} else {
			WARN_ON_ONCE(1);
		}
		return 0;

	case ODEBUG_STATE_ACTIVE:
		WARN_ON(1);

	default:
		return 0;
	}
}

/*
 * fixup_free is called when:
 * - an active object is freed
 */
static int timer_fixup_free(void *addr, enum debug_obj_state state)
{
	struct timer_list *timer = addr;

	switch (state) {
	case ODEBUG_STATE_ACTIVE:
		del_timer_sync(timer);
		debug_object_free(timer, &timer_debug_descr);
		return 1;
	default:
		return 0;
	}
}

static struct debug_obj_descr timer_debug_descr = {
	.name		= "timer_list",
	.fixup_init	= timer_fixup_init,
	.fixup_activate	= timer_fixup_activate,
	.fixup_free	= timer_fixup_free,
};

static inline void debug_timer_init(struct timer_list *timer)
{
	debug_object_init(timer, &timer_debug_descr);
}

static inline void debug_timer_activate(struct timer_list *timer)
{
	debug_object_activate(timer, &timer_debug_descr);
}

static inline void debug_timer_deactivate(struct timer_list *timer)
{
	debug_object_deactivate(timer, &timer_debug_descr);
}

static inline void debug_timer_free(struct timer_list *timer)
{
	debug_object_free(timer, &timer_debug_descr);
}

527 528 529
static void __init_timer(struct timer_list *timer,
			 const char *name,
			 struct lock_class_key *key);
530

531 532 533
void init_timer_on_stack_key(struct timer_list *timer,
			     const char *name,
			     struct lock_class_key *key)
534 535
{
	debug_object_init_on_stack(timer, &timer_debug_descr);
536
	__init_timer(timer, name, key);
537
}
538
EXPORT_SYMBOL_GPL(init_timer_on_stack_key);
539 540 541 542 543 544 545 546 547 548 549 550 551

void destroy_timer_on_stack(struct timer_list *timer)
{
	debug_object_free(timer, &timer_debug_descr);
}
EXPORT_SYMBOL_GPL(destroy_timer_on_stack);

#else
static inline void debug_timer_init(struct timer_list *timer) { }
static inline void debug_timer_activate(struct timer_list *timer) { }
static inline void debug_timer_deactivate(struct timer_list *timer) { }
#endif

552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570
static inline void debug_init(struct timer_list *timer)
{
	debug_timer_init(timer);
	trace_timer_init(timer);
}

static inline void
debug_activate(struct timer_list *timer, unsigned long expires)
{
	debug_timer_activate(timer);
	trace_timer_start(timer, expires);
}

static inline void debug_deactivate(struct timer_list *timer)
{
	debug_timer_deactivate(timer);
	trace_timer_cancel(timer);
}

571 572 573
static void __init_timer(struct timer_list *timer,
			 const char *name,
			 struct lock_class_key *key)
574 575
{
	timer->entry.next = NULL;
576
	timer->base = __raw_get_cpu_var(tvec_bases);
577
	timer->slack = -1;
578 579 580 581 582
#ifdef CONFIG_TIMER_STATS
	timer->start_site = NULL;
	timer->start_pid = -1;
	memset(timer->start_comm, 0, TASK_COMM_LEN);
#endif
583
	lockdep_init_map(&timer->lockdep_map, name, key, 0);
584
}
585

586 587 588 589 590 591 592 593 594 595 596 597 598
void setup_deferrable_timer_on_stack_key(struct timer_list *timer,
					 const char *name,
					 struct lock_class_key *key,
					 void (*function)(unsigned long),
					 unsigned long data)
{
	timer->function = function;
	timer->data = data;
	init_timer_on_stack_key(timer, name, key);
	timer_set_deferrable(timer);
}
EXPORT_SYMBOL_GPL(setup_deferrable_timer_on_stack_key);

599
/**
R
Randy Dunlap 已提交
600
 * init_timer_key - initialize a timer
601
 * @timer: the timer to be initialized
R
Randy Dunlap 已提交
602 603 604
 * @name: name of the timer
 * @key: lockdep class key of the fake lock used for tracking timer
 *       sync lock dependencies
605
 *
R
Randy Dunlap 已提交
606
 * init_timer_key() must be done to a timer prior calling *any* of the
607 608
 * other timer functions.
 */
609 610 611
void init_timer_key(struct timer_list *timer,
		    const char *name,
		    struct lock_class_key *key)
612
{
613
	debug_init(timer);
614
	__init_timer(timer, name, key);
615
}
616
EXPORT_SYMBOL(init_timer_key);
617

618 619 620
void init_timer_deferrable_key(struct timer_list *timer,
			       const char *name,
			       struct lock_class_key *key)
621
{
622
	init_timer_key(timer, name, key);
623 624
	timer_set_deferrable(timer);
}
625
EXPORT_SYMBOL(init_timer_deferrable_key);
626

627
static inline void detach_timer(struct timer_list *timer,
628
				int clear_pending)
629 630 631
{
	struct list_head *entry = &timer->entry;

632
	debug_deactivate(timer);
633

634 635 636 637 638 639 640
	__list_del(entry->prev, entry->next);
	if (clear_pending)
		entry->next = NULL;
	entry->prev = LIST_POISON2;
}

/*
641
 * We are using hashed locking: holding per_cpu(tvec_bases).lock
642 643 644 645 646 647 648 649 650 651
 * 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.
 */
652
static struct tvec_base *lock_timer_base(struct timer_list *timer,
653
					unsigned long *flags)
654
	__acquires(timer->base->lock)
655
{
656
	struct tvec_base *base;
657 658

	for (;;) {
659
		struct tvec_base *prelock_base = timer->base;
660
		base = tbase_get_base(prelock_base);
661 662
		if (likely(base != NULL)) {
			spin_lock_irqsave(&base->lock, *flags);
663
			if (likely(prelock_base == timer->base))
664 665 666 667 668 669 670 671
				return base;
			/* The timer has migrated to another CPU */
			spin_unlock_irqrestore(&base->lock, *flags);
		}
		cpu_relax();
	}
}

I
Ingo Molnar 已提交
672
static inline int
673 674
__mod_timer(struct timer_list *timer, unsigned long expires,
						bool pending_only, int pinned)
L
Linus Torvalds 已提交
675
{
676
	struct tvec_base *base, *new_base;
L
Linus Torvalds 已提交
677
	unsigned long flags;
678
	int ret = 0 , cpu;
L
Linus Torvalds 已提交
679

680
	timer_stats_timer_set_start_info(timer);
L
Linus Torvalds 已提交
681 682
	BUG_ON(!timer->function);

683 684 685 686
	base = lock_timer_base(timer, &flags);

	if (timer_pending(timer)) {
		detach_timer(timer, 0);
687 688 689
		if (timer->expires == base->next_timer &&
		    !tbase_get_deferrable(timer->base))
			base->next_timer = base->timer_jiffies;
690
		ret = 1;
I
Ingo Molnar 已提交
691 692 693
	} else {
		if (pending_only)
			goto out_unlock;
694 695
	}

696
	debug_activate(timer, expires);
697

698 699 700
	cpu = smp_processor_id();

#if defined(CONFIG_NO_HZ) && defined(CONFIG_SMP)
701 702
	if (!pinned && get_sysctl_timer_migration() && idle_cpu(cpu))
		cpu = get_nohz_timer_target();
703 704 705
#endif
	new_base = per_cpu(tvec_bases, cpu);

706
	if (base != new_base) {
L
Linus Torvalds 已提交
707
		/*
708 709 710 711 712
		 * 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
713
		 */
714
		if (likely(base->running_timer != timer)) {
715
			/* See the comment in lock_timer_base() */
716
			timer_set_base(timer, NULL);
717
			spin_unlock(&base->lock);
718 719
			base = new_base;
			spin_lock(&base->lock);
720
			timer_set_base(timer, base);
L
Linus Torvalds 已提交
721 722 723 724
		}
	}

	timer->expires = expires;
725 726 727
	if (time_before(timer->expires, base->next_timer) &&
	    !tbase_get_deferrable(timer->base))
		base->next_timer = timer->expires;
728
	internal_add_timer(base, timer);
I
Ingo Molnar 已提交
729 730

out_unlock:
731
	spin_unlock_irqrestore(&base->lock, flags);
L
Linus Torvalds 已提交
732 733 734 735

	return ret;
}

736
/**
I
Ingo Molnar 已提交
737 738 739
 * mod_timer_pending - modify a pending timer's timeout
 * @timer: the pending timer to be modified
 * @expires: new timeout in jiffies
L
Linus Torvalds 已提交
740
 *
I
Ingo Molnar 已提交
741 742 743 744
 * mod_timer_pending() is the same for pending timers as mod_timer(),
 * but will not re-activate and modify already deleted timers.
 *
 * It is useful for unserialized use of timers.
L
Linus Torvalds 已提交
745
 */
I
Ingo Molnar 已提交
746
int mod_timer_pending(struct timer_list *timer, unsigned long expires)
L
Linus Torvalds 已提交
747
{
748
	return __mod_timer(timer, expires, true, TIMER_NOT_PINNED);
L
Linus Torvalds 已提交
749
}
I
Ingo Molnar 已提交
750
EXPORT_SYMBOL(mod_timer_pending);
L
Linus Torvalds 已提交
751

752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767
/*
 * Decide where to put the timer while taking the slack into account
 *
 * Algorithm:
 *   1) calculate the maximum (absolute) time
 *   2) calculate the highest bit where the expires and new max are different
 *   3) use this bit to make a mask
 *   4) use the bitmask to round down the maximum time, so that all last
 *      bits are zeros
 */
static inline
unsigned long apply_slack(struct timer_list *timer, unsigned long expires)
{
	unsigned long expires_limit, mask;
	int bit;

768
	expires_limit = expires;
769

770
	if (timer->slack >= 0) {
771
		expires_limit = expires + timer->slack;
772
	} else {
773
		unsigned long now = jiffies;
774

775 776 777 778
		/* No slack, if already expired else auto slack 0.4% */
		if (time_after(expires, now))
			expires_limit = expires + (expires - now)/256;
	}
779 780 781 782 783 784 785 786 787 788 789 790 791
	mask = expires ^ expires_limit;
	if (mask == 0)
		return expires;

	bit = find_last_bit(&mask, BITS_PER_LONG);

	mask = (1 << bit) - 1;

	expires_limit = expires_limit & ~(mask);

	return expires_limit;
}

792
/**
L
Linus Torvalds 已提交
793 794
 * mod_timer - modify a timer's timeout
 * @timer: the timer to be modified
795
 * @expires: new timeout in jiffies
L
Linus Torvalds 已提交
796
 *
797
 * mod_timer() is a more efficient way to update the expire field of an
L
Linus Torvalds 已提交
798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818
 * 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)
{
	/*
	 * This is a common optimization triggered by the
	 * networking code - if the timer is re-modified
	 * to be the same thing then just return:
	 */
819
	if (timer_pending(timer) && timer->expires == expires)
L
Linus Torvalds 已提交
820 821
		return 1;

822 823
	expires = apply_slack(timer, expires);

824
	return __mod_timer(timer, expires, false, TIMER_NOT_PINNED);
L
Linus Torvalds 已提交
825 826 827
}
EXPORT_SYMBOL(mod_timer);

828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849
/**
 * mod_timer_pinned - modify a timer's timeout
 * @timer: the timer to be modified
 * @expires: new timeout in jiffies
 *
 * mod_timer_pinned() is a way to update the expire field of an
 * active timer (if the timer is inactive it will be activated)
 * and not allow the timer to be migrated to a different CPU.
 *
 * mod_timer_pinned(timer, expires) is equivalent to:
 *
 *     del_timer(timer); timer->expires = expires; add_timer(timer);
 */
int mod_timer_pinned(struct timer_list *timer, unsigned long expires)
{
	if (timer->expires == expires && timer_pending(timer))
		return 1;

	return __mod_timer(timer, expires, false, TIMER_PINNED);
}
EXPORT_SYMBOL(mod_timer_pinned);

I
Ingo Molnar 已提交
850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886
/**
 * add_timer - start a timer
 * @timer: the timer to be added
 *
 * The kernel will do a ->function(->data) callback from the
 * timer interrupt at the ->expires point in the future. The
 * current time is 'jiffies'.
 *
 * The timer's ->expires, ->function (and if the handler uses it, ->data)
 * fields must be set prior calling this function.
 *
 * Timers with an ->expires field in the past will be executed in the next
 * timer tick.
 */
void add_timer(struct timer_list *timer)
{
	BUG_ON(timer_pending(timer));
	mod_timer(timer, timer->expires);
}
EXPORT_SYMBOL(add_timer);

/**
 * 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)
{
	struct tvec_base *base = per_cpu(tvec_bases, cpu);
	unsigned long flags;

	timer_stats_timer_set_start_info(timer);
	BUG_ON(timer_pending(timer) || !timer->function);
	spin_lock_irqsave(&base->lock, flags);
	timer_set_base(timer, base);
887
	debug_activate(timer, timer->expires);
888 889 890
	if (time_before(timer->expires, base->next_timer) &&
	    !tbase_get_deferrable(timer->base))
		base->next_timer = timer->expires;
I
Ingo Molnar 已提交
891 892 893 894 895 896 897 898 899 900 901 902
	internal_add_timer(base, timer);
	/*
	 * 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);
	spin_unlock_irqrestore(&base->lock, flags);
}
A
Andi Kleen 已提交
903
EXPORT_SYMBOL_GPL(add_timer_on);
I
Ingo Molnar 已提交
904

905
/**
L
Linus Torvalds 已提交
906 907 908 909 910 911 912 913 914 915 916 917
 * 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)
{
918
	struct tvec_base *base;
L
Linus Torvalds 已提交
919
	unsigned long flags;
920
	int ret = 0;
L
Linus Torvalds 已提交
921

922
	timer_stats_timer_clear_start_info(timer);
923 924 925 926
	if (timer_pending(timer)) {
		base = lock_timer_base(timer, &flags);
		if (timer_pending(timer)) {
			detach_timer(timer, 1);
927 928 929
			if (timer->expires == base->next_timer &&
			    !tbase_get_deferrable(timer->base))
				base->next_timer = base->timer_jiffies;
930 931
			ret = 1;
		}
L
Linus Torvalds 已提交
932 933 934
		spin_unlock_irqrestore(&base->lock, flags);
	}

935
	return ret;
L
Linus Torvalds 已提交
936 937 938 939
}
EXPORT_SYMBOL(del_timer);

#ifdef CONFIG_SMP
940 941 942 943
/**
 * try_to_del_timer_sync - Try to deactivate a timer
 * @timer: timer do del
 *
944 945 946 947 948 949 950
 * 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)
{
951
	struct tvec_base *base;
952 953 954 955 956 957 958 959
	unsigned long flags;
	int ret = -1;

	base = lock_timer_base(timer, &flags);

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

960
	timer_stats_timer_clear_start_info(timer);
961 962 963
	ret = 0;
	if (timer_pending(timer)) {
		detach_timer(timer, 1);
964 965 966
		if (timer->expires == base->next_timer &&
		    !tbase_get_deferrable(timer->base))
			base->next_timer = base->timer_jiffies;
967 968 969 970 971 972 973
		ret = 1;
	}
out:
	spin_unlock_irqrestore(&base->lock, flags);

	return ret;
}
974 975
EXPORT_SYMBOL(try_to_del_timer_sync);

976
/**
L
Linus Torvalds 已提交
977 978 979 980 981 982 983
 * 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.
 *
984
 * Synchronization rules: Callers must prevent restarting of the timer,
L
Linus Torvalds 已提交
985 986
 * otherwise this function is meaningless. It must not be called from
 * interrupt contexts. The caller must not hold locks which would prevent
987 988 989
 * 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
990 991 992 993 994
 *
 * The function returns whether it has deactivated a pending timer or not.
 */
int del_timer_sync(struct timer_list *timer)
{
995 996 997 998 999 1000 1001 1002 1003
#ifdef CONFIG_LOCKDEP
	unsigned long flags;

	local_irq_save(flags);
	lock_map_acquire(&timer->lockdep_map);
	lock_map_release(&timer->lockdep_map);
	local_irq_restore(flags);
#endif

1004 1005 1006 1007
	for (;;) {
		int ret = try_to_del_timer_sync(timer);
		if (ret >= 0)
			return ret;
1008
		cpu_relax();
1009
	}
L
Linus Torvalds 已提交
1010
}
1011
EXPORT_SYMBOL(del_timer_sync);
L
Linus Torvalds 已提交
1012 1013
#endif

1014
static int cascade(struct tvec_base *base, struct tvec *tv, int index)
L
Linus Torvalds 已提交
1015 1016
{
	/* cascade all the timers from tv up one level */
1017 1018 1019 1020
	struct timer_list *timer, *tmp;
	struct list_head tv_list;

	list_replace_init(tv->vec + index, &tv_list);
L
Linus Torvalds 已提交
1021 1022

	/*
1023 1024
	 * We are removing _all_ timers from the list, so we
	 * don't have to detach them individually.
L
Linus Torvalds 已提交
1025
	 */
1026
	list_for_each_entry_safe(timer, tmp, &tv_list, entry) {
1027
		BUG_ON(tbase_get_base(timer->base) != base);
1028
		internal_add_timer(base, timer);
L
Linus Torvalds 已提交
1029 1030 1031 1032 1033
	}

	return index;
}

1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062
static void call_timer_fn(struct timer_list *timer, void (*fn)(unsigned long),
			  unsigned long data)
{
	int preempt_count = preempt_count();

#ifdef CONFIG_LOCKDEP
	/*
	 * It is permissible to free the timer from inside the
	 * function that is called from it, this we need to take into
	 * account for lockdep too. To avoid bogus "held lock freed"
	 * warnings as well as problems when looking into
	 * timer->lockdep_map, make a copy and use that here.
	 */
	struct lockdep_map lockdep_map = timer->lockdep_map;
#endif
	/*
	 * Couple the lock chain with the lock chain at
	 * del_timer_sync() by acquiring the lock_map around the fn()
	 * call here and in del_timer_sync().
	 */
	lock_map_acquire(&lockdep_map);

	trace_timer_expire_entry(timer);
	fn(data);
	trace_timer_expire_exit(timer);

	lock_map_release(&lockdep_map);

	if (preempt_count != preempt_count()) {
1063 1064 1065 1066 1067 1068 1069 1070 1071
		WARN_ONCE(1, "timer: %pF preempt leak: %08x -> %08x\n",
			  fn, preempt_count, preempt_count());
		/*
		 * Restore the preempt count. That gives us a decent
		 * chance to survive and extract information. If the
		 * callback kept a lock held, bad luck, but not worse
		 * than the BUG() we had.
		 */
		preempt_count() = preempt_count;
1072 1073 1074
	}
}

1075 1076 1077
#define INDEX(N) ((base->timer_jiffies >> (TVR_BITS + (N) * TVN_BITS)) & TVN_MASK)

/**
L
Linus Torvalds 已提交
1078 1079 1080 1081 1082 1083
 * __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.
 */
1084
static inline void __run_timers(struct tvec_base *base)
L
Linus Torvalds 已提交
1085 1086 1087
{
	struct timer_list *timer;

1088
	spin_lock_irq(&base->lock);
L
Linus Torvalds 已提交
1089
	while (time_after_eq(jiffies, base->timer_jiffies)) {
1090
		struct list_head work_list;
L
Linus Torvalds 已提交
1091
		struct list_head *head = &work_list;
1092
		int index = base->timer_jiffies & TVR_MASK;
1093

L
Linus Torvalds 已提交
1094 1095 1096 1097 1098 1099 1100 1101
		/*
		 * 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));
1102 1103
		++base->timer_jiffies;
		list_replace_init(base->tv1.vec + index, &work_list);
1104
		while (!list_empty(head)) {
L
Linus Torvalds 已提交
1105 1106 1107
			void (*fn)(unsigned long);
			unsigned long data;

1108
			timer = list_first_entry(head, struct timer_list,entry);
1109 1110
			fn = timer->function;
			data = timer->data;
L
Linus Torvalds 已提交
1111

1112 1113
			timer_stats_account_timer(timer);

L
Linus Torvalds 已提交
1114
			set_running_timer(base, timer);
1115
			detach_timer(timer, 1);
1116

1117
			spin_unlock_irq(&base->lock);
1118
			call_timer_fn(timer, fn, data);
1119
			spin_lock_irq(&base->lock);
L
Linus Torvalds 已提交
1120 1121 1122
		}
	}
	set_running_timer(base, NULL);
1123
	spin_unlock_irq(&base->lock);
L
Linus Torvalds 已提交
1124 1125
}

1126
#ifdef CONFIG_NO_HZ
L
Linus Torvalds 已提交
1127 1128
/*
 * Find out when the next timer event is due to happen. This
R
Randy Dunlap 已提交
1129 1130
 * is used on S/390 to stop all activity when a CPU is idle.
 * This function needs to be called with interrupts disabled.
L
Linus Torvalds 已提交
1131
 */
1132
static unsigned long __next_timer_interrupt(struct tvec_base *base)
L
Linus Torvalds 已提交
1133
{
1134
	unsigned long timer_jiffies = base->timer_jiffies;
1135
	unsigned long expires = timer_jiffies + NEXT_TIMER_MAX_DELTA;
1136
	int index, slot, array, found = 0;
L
Linus Torvalds 已提交
1137
	struct timer_list *nte;
1138
	struct tvec *varray[4];
L
Linus Torvalds 已提交
1139 1140

	/* Look for timer events in tv1. */
1141
	index = slot = timer_jiffies & TVR_MASK;
L
Linus Torvalds 已提交
1142
	do {
1143
		list_for_each_entry(nte, base->tv1.vec + slot, entry) {
1144 1145
			if (tbase_get_deferrable(nte->base))
				continue;
1146

1147
			found = 1;
L
Linus Torvalds 已提交
1148
			expires = nte->expires;
1149 1150 1151 1152
			/* Look at the cascade bucket(s)? */
			if (!index || slot < index)
				goto cascade;
			return expires;
L
Linus Torvalds 已提交
1153
		}
1154 1155 1156 1157 1158 1159 1160 1161
		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
1162 1163 1164 1165 1166 1167

	/* Check tv2-tv5. */
	varray[0] = &base->tv2;
	varray[1] = &base->tv3;
	varray[2] = &base->tv4;
	varray[3] = &base->tv5;
1168 1169

	for (array = 0; array < 4; array++) {
1170
		struct tvec *varp = varray[array];
1171 1172

		index = slot = timer_jiffies & TVN_MASK;
L
Linus Torvalds 已提交
1173
		do {
1174
			list_for_each_entry(nte, varp->vec + slot, entry) {
1175 1176 1177
				if (tbase_get_deferrable(nte->base))
					continue;

1178
				found = 1;
L
Linus Torvalds 已提交
1179 1180
				if (time_before(nte->expires, expires))
					expires = nte->expires;
1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197
			}
			/*
			 * 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
1198
	}
1199 1200
	return expires;
}
1201

1202 1203 1204 1205 1206 1207 1208 1209 1210
/*
 * 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;
1211
	unsigned long delta;
1212 1213 1214

	if (hr_delta.tv64 == KTIME_MAX)
		return expires;
1215

1216 1217 1218 1219 1220
	/*
	 * Expired timer available, let it expire in the next tick
	 */
	if (hr_delta.tv64 <= 0)
		return now + 1;
1221

1222
	tsdelta = ktime_to_timespec(hr_delta);
1223
	delta = timespec_to_jiffies(&tsdelta);
1224 1225 1226 1227 1228 1229 1230 1231

	/*
	 * 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;

1232 1233 1234 1235 1236 1237 1238 1239 1240
	/*
	 * 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;
1241 1242
	if (time_before(now, expires))
		return now;
L
Linus Torvalds 已提交
1243 1244
	return expires;
}
1245 1246

/**
1247
 * get_next_timer_interrupt - return the jiffy of the next pending timer
1248
 * @now: current time (in jiffies)
1249
 */
1250
unsigned long get_next_timer_interrupt(unsigned long now)
1251
{
1252
	struct tvec_base *base = __get_cpu_var(tvec_bases);
1253
	unsigned long expires;
1254

1255 1256 1257 1258 1259 1260
	/*
	 * Pretend that there is no timer pending if the cpu is offline.
	 * Possible pending timers will be migrated later to an active cpu.
	 */
	if (cpu_is_offline(smp_processor_id()))
		return now + NEXT_TIMER_MAX_DELTA;
1261
	spin_lock(&base->lock);
1262 1263 1264
	if (time_before_eq(base->next_timer, base->timer_jiffies))
		base->next_timer = __next_timer_interrupt(base);
	expires = base->next_timer;
1265 1266 1267 1268 1269 1270 1271
	spin_unlock(&base->lock);

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

	return cmp_next_hrtimer_event(now, expires);
}
L
Linus Torvalds 已提交
1272 1273 1274
#endif

/*
D
Daniel Walker 已提交
1275
 * Called from the timer interrupt handler to charge one tick to the current
L
Linus Torvalds 已提交
1276 1277 1278 1279 1280 1281 1282 1283
 * 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. */
1284
	account_process_tick(p, user_tick);
L
Linus Torvalds 已提交
1285
	run_local_timers();
1286
	rcu_check_callbacks(cpu, user_tick);
P
Peter Zijlstra 已提交
1287
	printk_tick();
1288 1289 1290 1291
#ifdef CONFIG_IRQ_WORK
	if (in_irq())
		irq_work_run();
#endif
L
Linus Torvalds 已提交
1292
	scheduler_tick();
1293
	run_posix_cpu_timers(p);
L
Linus Torvalds 已提交
1294 1295 1296 1297 1298 1299 1300
}

/*
 * This function runs timers and the timer-tq in bottom half context.
 */
static void run_timer_softirq(struct softirq_action *h)
{
1301
	struct tvec_base *base = __get_cpu_var(tvec_bases);
L
Linus Torvalds 已提交
1302

1303
	hrtimer_run_pending();
1304

L
Linus Torvalds 已提交
1305 1306 1307 1308 1309 1310 1311 1312 1313
	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)
{
1314
	hrtimer_run_queues();
L
Linus Torvalds 已提交
1315 1316 1317 1318 1319 1320 1321 1322 1323
	raise_softirq(TIMER_SOFTIRQ);
}

/*
 * 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...
 */

1324
void do_timer(unsigned long ticks)
L
Linus Torvalds 已提交
1325
{
1326
	jiffies_64 += ticks;
1327
	update_wall_time();
1328
	calc_global_load(ticks);
L
Linus Torvalds 已提交
1329 1330 1331 1332 1333 1334 1335 1336
}

#ifdef __ARCH_WANT_SYS_ALARM

/*
 * For backwards compatibility?  This can be done in libc so Alpha
 * and all newer ports shouldn't need it.
 */
1337
SYSCALL_DEFINE1(alarm, unsigned int, seconds)
L
Linus Torvalds 已提交
1338
{
1339
	return alarm_setitimer(seconds);
L
Linus Torvalds 已提交
1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359
}

#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.
 */
1360
SYSCALL_DEFINE0(getpid)
L
Linus Torvalds 已提交
1361
{
1362
	return task_tgid_vnr(current);
L
Linus Torvalds 已提交
1363 1364 1365
}

/*
1366 1367 1368 1369
 * 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
1370
 */
1371
SYSCALL_DEFINE0(getppid)
L
Linus Torvalds 已提交
1372 1373 1374
{
	int pid;

1375
	rcu_read_lock();
1376
	pid = task_tgid_vnr(current->real_parent);
1377
	rcu_read_unlock();
L
Linus Torvalds 已提交
1378 1379 1380 1381

	return pid;
}

1382
SYSCALL_DEFINE0(getuid)
L
Linus Torvalds 已提交
1383 1384
{
	/* Only we change this so SMP safe */
1385
	return current_uid();
L
Linus Torvalds 已提交
1386 1387
}

1388
SYSCALL_DEFINE0(geteuid)
L
Linus Torvalds 已提交
1389 1390
{
	/* Only we change this so SMP safe */
1391
	return current_euid();
L
Linus Torvalds 已提交
1392 1393
}

1394
SYSCALL_DEFINE0(getgid)
L
Linus Torvalds 已提交
1395 1396
{
	/* Only we change this so SMP safe */
1397
	return current_gid();
L
Linus Torvalds 已提交
1398 1399
}

1400
SYSCALL_DEFINE0(getegid)
L
Linus Torvalds 已提交
1401 1402
{
	/* Only we change this so SMP safe */
1403
	return  current_egid();
L
Linus Torvalds 已提交
1404 1405 1406 1407 1408 1409
}

#endif

static void process_timeout(unsigned long __data)
{
1410
	wake_up_process((struct task_struct *)__data);
L
Linus Torvalds 已提交
1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438
}

/**
 * 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.
 */
1439
signed long __sched schedule_timeout(signed long timeout)
L
Linus Torvalds 已提交
1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463
{
	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.
		 */
1464
		if (timeout < 0) {
L
Linus Torvalds 已提交
1465
			printk(KERN_ERR "schedule_timeout: wrong timeout "
1466 1467
				"value %lx\n", timeout);
			dump_stack();
L
Linus Torvalds 已提交
1468 1469 1470 1471 1472 1473 1474
			current->state = TASK_RUNNING;
			goto out;
		}
	}

	expire = timeout + jiffies;

1475
	setup_timer_on_stack(&timer, process_timeout, (unsigned long)current);
1476
	__mod_timer(&timer, expire, false, TIMER_NOT_PINNED);
L
Linus Torvalds 已提交
1477 1478 1479
	schedule();
	del_singleshot_timer_sync(&timer);

1480 1481 1482
	/* Remove the timer from the object tracker */
	destroy_timer_on_stack(&timer);

L
Linus Torvalds 已提交
1483 1484 1485 1486 1487 1488 1489
	timeout = expire - jiffies;

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

1490 1491 1492 1493
/*
 * We can use __set_current_state() here because schedule_timeout() calls
 * schedule() unconditionally.
 */
1494 1495
signed long __sched schedule_timeout_interruptible(signed long timeout)
{
A
Andrew Morton 已提交
1496 1497
	__set_current_state(TASK_INTERRUPTIBLE);
	return schedule_timeout(timeout);
1498 1499 1500
}
EXPORT_SYMBOL(schedule_timeout_interruptible);

M
Matthew Wilcox 已提交
1501 1502 1503 1504 1505 1506 1507
signed long __sched schedule_timeout_killable(signed long timeout)
{
	__set_current_state(TASK_KILLABLE);
	return schedule_timeout(timeout);
}
EXPORT_SYMBOL(schedule_timeout_killable);

1508 1509
signed long __sched schedule_timeout_uninterruptible(signed long timeout)
{
A
Andrew Morton 已提交
1510 1511
	__set_current_state(TASK_UNINTERRUPTIBLE);
	return schedule_timeout(timeout);
1512 1513 1514
}
EXPORT_SYMBOL(schedule_timeout_uninterruptible);

L
Linus Torvalds 已提交
1515
/* Thread ID - the internal kernel "pid" */
1516
SYSCALL_DEFINE0(gettid)
L
Linus Torvalds 已提交
1517
{
1518
	return task_pid_vnr(current);
L
Linus Torvalds 已提交
1519 1520
}

1521
/**
1522
 * do_sysinfo - fill in sysinfo struct
1523
 * @info: pointer to buffer to fill
1524
 */
1525
int do_sysinfo(struct sysinfo *info)
L
Linus Torvalds 已提交
1526 1527 1528
{
	unsigned long mem_total, sav_total;
	unsigned int mem_unit, bitcount;
1529
	struct timespec tp;
L
Linus Torvalds 已提交
1530

1531
	memset(info, 0, sizeof(struct sysinfo));
L
Linus Torvalds 已提交
1532

1533 1534 1535
	ktime_get_ts(&tp);
	monotonic_to_bootbased(&tp);
	info->uptime = tp.tv_sec + (tp.tv_nsec ? 1 : 0);
L
Linus Torvalds 已提交
1536

1537
	get_avenrun(info->loads, 0, SI_LOAD_SHIFT - FSHIFT);
L
Linus Torvalds 已提交
1538

1539
	info->procs = nr_threads;
L
Linus Torvalds 已提交
1540

1541 1542
	si_meminfo(info);
	si_swapinfo(info);
L
Linus Torvalds 已提交
1543 1544 1545 1546 1547 1548 1549 1550 1551 1552

	/*
	 * 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>
	 */

1553 1554
	mem_total = info->totalram + info->totalswap;
	if (mem_total < info->totalram || mem_total < info->totalswap)
L
Linus Torvalds 已提交
1555 1556
		goto out;
	bitcount = 0;
1557
	mem_unit = info->mem_unit;
L
Linus Torvalds 已提交
1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568
	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
1569
	 * info->mem_unit and set it to 1.  This leaves things compatible
L
Linus Torvalds 已提交
1570 1571 1572 1573
	 * with 2.2.x, and also retains compatibility with earlier 2.4.x
	 * kernels...
	 */

1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587
	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;
}

1588
SYSCALL_DEFINE1(sysinfo, struct sysinfo __user *, info)
1589 1590 1591 1592
{
	struct sysinfo val;

	do_sysinfo(&val);
L
Linus Torvalds 已提交
1593 1594 1595 1596 1597 1598 1599

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

	return 0;
}

A
Adrian Bunk 已提交
1600
static int __cpuinit init_timers_cpu(int cpu)
L
Linus Torvalds 已提交
1601 1602
{
	int j;
1603
	struct tvec_base *base;
A
Adrian Bunk 已提交
1604
	static char __cpuinitdata tvec_base_done[NR_CPUS];
1605

A
Andrew Morton 已提交
1606
	if (!tvec_base_done[cpu]) {
1607 1608 1609
		static char boot_done;

		if (boot_done) {
A
Andrew Morton 已提交
1610 1611 1612
			/*
			 * The APs use this path later in boot
			 */
1613 1614
			base = kmalloc_node(sizeof(*base),
						GFP_KERNEL | __GFP_ZERO,
1615 1616 1617
						cpu_to_node(cpu));
			if (!base)
				return -ENOMEM;
1618 1619 1620 1621 1622 1623 1624

			/* 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
1625
			per_cpu(tvec_bases, cpu) = base;
1626
		} else {
A
Andrew Morton 已提交
1627 1628 1629 1630 1631 1632
			/*
			 * 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.
			 */
1633
			boot_done = 1;
A
Andrew Morton 已提交
1634
			base = &boot_tvec_bases;
1635
		}
A
Andrew Morton 已提交
1636 1637 1638
		tvec_base_done[cpu] = 1;
	} else {
		base = per_cpu(tvec_bases, cpu);
1639
	}
A
Andrew Morton 已提交
1640

1641
	spin_lock_init(&base->lock);
1642

L
Linus Torvalds 已提交
1643 1644 1645 1646 1647 1648 1649 1650 1651 1652
	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;
1653
	base->next_timer = base->timer_jiffies;
1654
	return 0;
L
Linus Torvalds 已提交
1655 1656 1657
}

#ifdef CONFIG_HOTPLUG_CPU
1658
static void migrate_timer_list(struct tvec_base *new_base, struct list_head *head)
L
Linus Torvalds 已提交
1659 1660 1661 1662
{
	struct timer_list *timer;

	while (!list_empty(head)) {
1663
		timer = list_first_entry(head, struct timer_list, entry);
1664
		detach_timer(timer, 0);
1665
		timer_set_base(timer, new_base);
1666 1667 1668
		if (time_before(timer->expires, new_base->next_timer) &&
		    !tbase_get_deferrable(timer->base))
			new_base->next_timer = timer->expires;
L
Linus Torvalds 已提交
1669 1670 1671 1672
		internal_add_timer(new_base, timer);
	}
}

R
Randy Dunlap 已提交
1673
static void __cpuinit migrate_timers(int cpu)
L
Linus Torvalds 已提交
1674
{
1675 1676
	struct tvec_base *old_base;
	struct tvec_base *new_base;
L
Linus Torvalds 已提交
1677 1678 1679
	int i;

	BUG_ON(cpu_online(cpu));
1680 1681
	old_base = per_cpu(tvec_bases, cpu);
	new_base = get_cpu_var(tvec_bases);
1682 1683 1684 1685 1686
	/*
	 * The caller is globally serialized and nobody else
	 * takes two locks at once, deadlock is not possible.
	 */
	spin_lock_irq(&new_base->lock);
1687
	spin_lock_nested(&old_base->lock, SINGLE_DEPTH_NESTING);
1688 1689

	BUG_ON(old_base->running_timer);
L
Linus Torvalds 已提交
1690 1691

	for (i = 0; i < TVR_SIZE; i++)
1692 1693 1694 1695 1696 1697 1698 1699
		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);
	}

1700
	spin_unlock(&old_base->lock);
1701
	spin_unlock_irq(&new_base->lock);
L
Linus Torvalds 已提交
1702 1703 1704 1705
	put_cpu_var(tvec_bases);
}
#endif /* CONFIG_HOTPLUG_CPU */

1706
static int __cpuinit timer_cpu_notify(struct notifier_block *self,
L
Linus Torvalds 已提交
1707 1708 1709
				unsigned long action, void *hcpu)
{
	long cpu = (long)hcpu;
1710 1711
	int err;

L
Linus Torvalds 已提交
1712 1713
	switch(action) {
	case CPU_UP_PREPARE:
1714
	case CPU_UP_PREPARE_FROZEN:
1715 1716 1717
		err = init_timers_cpu(cpu);
		if (err < 0)
			return notifier_from_errno(err);
L
Linus Torvalds 已提交
1718 1719 1720
		break;
#ifdef CONFIG_HOTPLUG_CPU
	case CPU_DEAD:
1721
	case CPU_DEAD_FROZEN:
L
Linus Torvalds 已提交
1722 1723 1724 1725 1726 1727 1728 1729 1730
		migrate_timers(cpu);
		break;
#endif
	default:
		break;
	}
	return NOTIFY_OK;
}

1731
static struct notifier_block __cpuinitdata timers_nb = {
L
Linus Torvalds 已提交
1732 1733 1734 1735 1736 1737
	.notifier_call	= timer_cpu_notify,
};


void __init init_timers(void)
{
1738
	int err = timer_cpu_notify(&timers_nb, (unsigned long)CPU_UP_PREPARE,
L
Linus Torvalds 已提交
1739
				(void *)(long)smp_processor_id());
1740

1741 1742
	init_timer_stats();

1743
	BUG_ON(err != NOTIFY_OK);
L
Linus Torvalds 已提交
1744
	register_cpu_notifier(&timers_nb);
1745
	open_softirq(TIMER_SOFTIRQ, run_timer_softirq);
L
Linus Torvalds 已提交
1746 1747 1748 1749 1750 1751 1752 1753 1754 1755
}

/**
 * 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;

1756 1757
	while (timeout)
		timeout = schedule_timeout_uninterruptible(timeout);
L
Linus Torvalds 已提交
1758 1759 1760 1761 1762
}

EXPORT_SYMBOL(msleep);

/**
1763
 * msleep_interruptible - sleep waiting for signals
L
Linus Torvalds 已提交
1764 1765 1766 1767 1768 1769
 * @msecs: Time in milliseconds to sleep for
 */
unsigned long msleep_interruptible(unsigned int msecs)
{
	unsigned long timeout = msecs_to_jiffies(msecs) + 1;

1770 1771
	while (timeout && !signal_pending(current))
		timeout = schedule_timeout_interruptible(timeout);
L
Linus Torvalds 已提交
1772 1773 1774 1775
	return jiffies_to_msecs(timeout);
}

EXPORT_SYMBOL(msleep_interruptible);
1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797

static int __sched do_usleep_range(unsigned long min, unsigned long max)
{
	ktime_t kmin;
	unsigned long delta;

	kmin = ktime_set(0, min * NSEC_PER_USEC);
	delta = (max - min) * NSEC_PER_USEC;
	return schedule_hrtimeout_range(&kmin, delta, HRTIMER_MODE_REL);
}

/**
 * usleep_range - Drop in replacement for udelay where wakeup is flexible
 * @min: Minimum time in usecs to sleep
 * @max: Maximum time in usecs to sleep
 */
void usleep_range(unsigned long min, unsigned long max)
{
	__set_current_state(TASK_UNINTERRUPTIBLE);
	do_usleep_range(min, max);
}
EXPORT_SYMBOL(usleep_range);