timer.c 49.1 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35
/*
 *  linux/kernel/timer.c
 *
 *  Kernel internal timers, kernel timekeeping, basic process system calls
 *
 *  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>
#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
36
#include <linux/delay.h>
37
#include <linux/clockchips.h>
L
Linus Torvalds 已提交
38 39 40 41 42 43 44

#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
45 46 47 48
u64 jiffies_64 __cacheline_aligned_in_smp = INITIAL_JIFFIES;

EXPORT_SYMBOL(jiffies_64);

L
Linus Torvalds 已提交
49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67
/*
 * 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)

typedef struct tvec_s {
	struct list_head vec[TVN_SIZE];
} tvec_t;

typedef struct tvec_root_s {
	struct list_head vec[TVR_SIZE];
} tvec_root_t;

struct tvec_t_base_s {
68 69
	spinlock_t lock;
	struct timer_list *running_timer;
L
Linus Torvalds 已提交
70 71 72 73 74 75 76 77 78
	unsigned long timer_jiffies;
	tvec_root_t tv1;
	tvec_t tv2;
	tvec_t tv3;
	tvec_t tv4;
	tvec_t tv5;
} ____cacheline_aligned_in_smp;

typedef struct tvec_t_base_s tvec_base_t;
A
Andrew Morton 已提交
79

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

84 85 86 87 88
/**
 * __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
 *
89
 * __round_jiffies() rounds an absolute time in the future (in jiffies)
90 91 92 93 94 95 96 97 98 99 100 101
 * 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.
 *
102
 * The return value is the rounded version of the @j parameter.
103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145
 */
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
 *
146
 * __round_jiffies_relative() rounds a time delta  in the future (in jiffies)
147 148 149 150 151 152 153 154 155 156 157 158
 * 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.
 *
159
 * The return value is the rounded version of the @j parameter.
160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176
 */
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
 *
177
 * round_jiffies() rounds an absolute time in the future (in jiffies)
178 179 180 181 182 183 184 185
 * 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.
 *
186
 * The return value is the rounded version of the @j parameter.
187 188 189 190 191 192 193 194 195 196 197
 */
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
 *
198
 * round_jiffies_relative() rounds a time delta  in the future (in jiffies)
199 200 201 202 203 204 205 206
 * 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.
 *
207
 * The return value is the rounded version of the @j parameter.
208 209 210 211 212 213 214 215
 */
unsigned long round_jiffies_relative(unsigned long j)
{
	return __round_jiffies_relative(j, raw_smp_processor_id());
}
EXPORT_SYMBOL_GPL(round_jiffies_relative);


L
Linus Torvalds 已提交
216 217 218 219
static inline void set_running_timer(tvec_base_t *base,
					struct timer_list *timer)
{
#ifdef CONFIG_SMP
220
	base->running_timer = timer;
L
Linus Torvalds 已提交
221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265
#endif
}

static void internal_add_timer(tvec_base_t *base, struct timer_list *timer)
{
	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);
}

266
/**
267 268 269 270 271 272 273 274 275
 * 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.
 */
void fastcall init_timer(struct timer_list *timer)
{
	timer->entry.next = NULL;
276
	timer->base = __raw_get_cpu_var(tvec_bases);
277 278 279 280 281 282 283 284 285 286 287 288 289 290 291
}
EXPORT_SYMBOL(init_timer);

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

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

/*
292
 * We are using hashed locking: holding per_cpu(tvec_bases).lock
293 294 295 296 297 298 299 300 301 302
 * 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.
 */
303
static tvec_base_t *lock_timer_base(struct timer_list *timer,
304
					unsigned long *flags)
305
	__acquires(timer->base->lock)
306
{
307
	tvec_base_t *base;
308 309 310 311 312 313 314 315 316 317 318 319 320 321

	for (;;) {
		base = timer->base;
		if (likely(base != NULL)) {
			spin_lock_irqsave(&base->lock, *flags);
			if (likely(base == timer->base))
				return base;
			/* The timer has migrated to another CPU */
			spin_unlock_irqrestore(&base->lock, *flags);
		}
		cpu_relax();
	}
}

L
Linus Torvalds 已提交
322 323
int __mod_timer(struct timer_list *timer, unsigned long expires)
{
324
	tvec_base_t *base, *new_base;
L
Linus Torvalds 已提交
325 326 327 328 329
	unsigned long flags;
	int ret = 0;

	BUG_ON(!timer->function);

330 331 332 333 334 335 336
	base = lock_timer_base(timer, &flags);

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

337
	new_base = __get_cpu_var(tvec_bases);
L
Linus Torvalds 已提交
338

339
	if (base != new_base) {
L
Linus Torvalds 已提交
340
		/*
341 342 343 344 345
		 * 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
346
		 */
347
		if (likely(base->running_timer != timer)) {
348 349 350
			/* See the comment in lock_timer_base() */
			timer->base = NULL;
			spin_unlock(&base->lock);
351 352 353
			base = new_base;
			spin_lock(&base->lock);
			timer->base = base;
L
Linus Torvalds 已提交
354 355 356 357
		}
	}

	timer->expires = expires;
358 359
	internal_add_timer(base, timer);
	spin_unlock_irqrestore(&base->lock, flags);
L
Linus Torvalds 已提交
360 361 362 363 364 365

	return ret;
}

EXPORT_SYMBOL(__mod_timer);

366
/**
L
Linus Torvalds 已提交
367 368 369 370 371 372 373 374
 * 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)
{
375
	tvec_base_t *base = per_cpu(tvec_bases, cpu);
L
Linus Torvalds 已提交
376
  	unsigned long flags;
377

L
Linus Torvalds 已提交
378
  	BUG_ON(timer_pending(timer) || !timer->function);
379 380
	spin_lock_irqsave(&base->lock, flags);
	timer->base = base;
L
Linus Torvalds 已提交
381
	internal_add_timer(base, timer);
382
	spin_unlock_irqrestore(&base->lock, flags);
L
Linus Torvalds 已提交
383 384 385
}


386
/**
L
Linus Torvalds 已提交
387 388
 * mod_timer - modify a timer's timeout
 * @timer: the timer to be modified
389
 * @expires: new timeout in jiffies
L
Linus Torvalds 已提交
390
 *
391
 * mod_timer() is a more efficient way to update the expire field of an
L
Linus Torvalds 已提交
392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422
 * 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);

	/*
	 * 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);

423
/**
L
Linus Torvalds 已提交
424 425 426 427 428 429 430 431 432 433 434 435
 * 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)
{
436
	tvec_base_t *base;
L
Linus Torvalds 已提交
437
	unsigned long flags;
438
	int ret = 0;
L
Linus Torvalds 已提交
439

440 441 442 443 444 445
	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
446 447 448
		spin_unlock_irqrestore(&base->lock, flags);
	}

449
	return ret;
L
Linus Torvalds 已提交
450 451 452 453 454
}

EXPORT_SYMBOL(del_timer);

#ifdef CONFIG_SMP
455 456 457 458
/**
 * try_to_del_timer_sync - Try to deactivate a timer
 * @timer: timer do del
 *
459 460 461 462 463 464 465
 * 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)
{
466
	tvec_base_t *base;
467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485
	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;
}

486
/**
L
Linus Torvalds 已提交
487 488 489 490 491 492 493
 * 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.
 *
494
 * Synchronization rules: Callers must prevent restarting of the timer,
L
Linus Torvalds 已提交
495 496
 * otherwise this function is meaningless. It must not be called from
 * interrupt contexts. The caller must not hold locks which would prevent
497 498 499
 * 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
500 501 502 503 504
 *
 * The function returns whether it has deactivated a pending timer or not.
 */
int del_timer_sync(struct timer_list *timer)
{
505 506 507 508
	for (;;) {
		int ret = try_to_del_timer_sync(timer);
		if (ret >= 0)
			return ret;
509
		cpu_relax();
510
	}
L
Linus Torvalds 已提交
511 512
}

513
EXPORT_SYMBOL(del_timer_sync);
L
Linus Torvalds 已提交
514 515 516 517 518
#endif

static int cascade(tvec_base_t *base, tvec_t *tv, int index)
{
	/* cascade all the timers from tv up one level */
519 520 521 522
	struct timer_list *timer, *tmp;
	struct list_head tv_list;

	list_replace_init(tv->vec + index, &tv_list);
L
Linus Torvalds 已提交
523 524

	/*
525 526
	 * We are removing _all_ timers from the list, so we
	 * don't have to detach them individually.
L
Linus Torvalds 已提交
527
	 */
528 529 530
	list_for_each_entry_safe(timer, tmp, &tv_list, entry) {
		BUG_ON(timer->base != base);
		internal_add_timer(base, timer);
L
Linus Torvalds 已提交
531 532 533 534 535
	}

	return index;
}

536 537 538
#define INDEX(N) ((base->timer_jiffies >> (TVR_BITS + (N) * TVN_BITS)) & TVN_MASK)

/**
L
Linus Torvalds 已提交
539 540 541 542 543 544 545 546 547 548
 * __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.
 */
static inline void __run_timers(tvec_base_t *base)
{
	struct timer_list *timer;

549
	spin_lock_irq(&base->lock);
L
Linus Torvalds 已提交
550
	while (time_after_eq(jiffies, base->timer_jiffies)) {
551
		struct list_head work_list;
L
Linus Torvalds 已提交
552 553
		struct list_head *head = &work_list;
 		int index = base->timer_jiffies & TVR_MASK;
554

L
Linus Torvalds 已提交
555 556 557 558 559 560 561 562
		/*
		 * 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));
563 564
		++base->timer_jiffies;
		list_replace_init(base->tv1.vec + index, &work_list);
565
		while (!list_empty(head)) {
L
Linus Torvalds 已提交
566 567 568 569 570 571 572 573
			void (*fn)(unsigned long);
			unsigned long data;

			timer = list_entry(head->next,struct timer_list,entry);
 			fn = timer->function;
 			data = timer->data;

			set_running_timer(base, timer);
574
			detach_timer(timer, 1);
575
			spin_unlock_irq(&base->lock);
L
Linus Torvalds 已提交
576
			{
577
				int preempt_count = preempt_count();
L
Linus Torvalds 已提交
578 579
				fn(data);
				if (preempt_count != preempt_count()) {
580 581 582 583 584
					printk(KERN_WARNING "huh, entered %p "
					       "with preempt_count %08x, exited"
					       " with %08x?\n",
					       fn, preempt_count,
					       preempt_count());
L
Linus Torvalds 已提交
585 586 587
					BUG();
				}
			}
588
			spin_lock_irq(&base->lock);
L
Linus Torvalds 已提交
589 590 591
		}
	}
	set_running_timer(base, NULL);
592
	spin_unlock_irq(&base->lock);
L
Linus Torvalds 已提交
593 594
}

595
#if defined(CONFIG_NO_IDLE_HZ) || defined(CONFIG_NO_HZ)
L
Linus Torvalds 已提交
596 597 598 599 600
/*
 * 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.
 */
601
static unsigned long __next_timer_interrupt(tvec_base_t *base)
L
Linus Torvalds 已提交
602
{
603 604 605
	unsigned long timer_jiffies = base->timer_jiffies;
	unsigned long expires = timer_jiffies + (LONG_MAX >> 1);
	int index, slot, array, found = 0;
L
Linus Torvalds 已提交
606 607 608 609
	struct timer_list *nte;
	tvec_t *varray[4];

	/* Look for timer events in tv1. */
610
	index = slot = timer_jiffies & TVR_MASK;
L
Linus Torvalds 已提交
611
	do {
612 613
		list_for_each_entry(nte, base->tv1.vec + slot, entry) {
			found = 1;
L
Linus Torvalds 已提交
614
			expires = nte->expires;
615 616 617 618
			/* Look at the cascade bucket(s)? */
			if (!index || slot < index)
				goto cascade;
			return expires;
L
Linus Torvalds 已提交
619
		}
620 621 622 623 624 625 626 627
		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
628 629 630 631 632 633

	/* Check tv2-tv5. */
	varray[0] = &base->tv2;
	varray[1] = &base->tv3;
	varray[2] = &base->tv4;
	varray[3] = &base->tv5;
634 635 636 637 638

	for (array = 0; array < 4; array++) {
		tvec_t *varp = varray[array];

		index = slot = timer_jiffies & TVN_MASK;
L
Linus Torvalds 已提交
639
		do {
640 641
			list_for_each_entry(nte, varp->vec + slot, entry) {
				found = 1;
L
Linus Torvalds 已提交
642 643
				if (time_before(nte->expires, expires))
					expires = nte->expires;
644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660
			}
			/*
			 * 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
661
	}
662 663
	return expires;
}
664

665 666 667 668 669 670 671 672 673 674 675 676
/*
 * 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;

	if (hr_delta.tv64 == KTIME_MAX)
		return expires;
677

678 679
	if (hr_delta.tv64 <= TICK_NSEC)
		return now;
680

681 682 683 684
	tsdelta = ktime_to_timespec(hr_delta);
	now += timespec_to_jiffies(&tsdelta);
	if (time_before(now, expires))
		return now;
L
Linus Torvalds 已提交
685 686
	return expires;
}
687 688 689 690

/**
 * next_timer_interrupt - return the jiffy of the next pending timer
 */
691
unsigned long get_next_timer_interrupt(unsigned long now)
692 693
{
	tvec_base_t *base = __get_cpu_var(tvec_bases);
694
	unsigned long expires;
695 696 697 698 699 700 701 702 703 704

	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);
}
705 706 707 708 709 710 711 712

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

L
Linus Torvalds 已提交
713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729
#endif

/******************************************************************/

/* 
 * The current time 
 * wall_to_monotonic is what we need to add to xtime (or xtime corrected 
 * for sub jiffie times) to get to monotonic time.  Monotonic is pegged
 * at zero at system boot time, so wall_to_monotonic will be negative,
 * however, we will ALWAYS keep the tv_nsec part positive so we can use
 * the usual normalization.
 */
struct timespec xtime __attribute__ ((aligned (16)));
struct timespec wall_to_monotonic __attribute__ ((aligned (16)));

EXPORT_SYMBOL(xtime);

730

731 732 733
/* XXX - all of this timekeeping code should be later moved to time.c */
#include <linux/clocksource.h>
static struct clocksource *clock; /* pointer to current clocksource */
734 735 736 737 738 739 740 741 742 743 744 745 746 747 748

#ifdef CONFIG_GENERIC_TIME
/**
 * __get_nsec_offset - Returns nanoseconds since last call to periodic_hook
 *
 * private function, must hold xtime_lock lock when being
 * called. Returns the number of nanoseconds since the
 * last call to update_wall_time() (adjusted by NTP scaling)
 */
static inline s64 __get_nsec_offset(void)
{
	cycle_t cycle_now, cycle_delta;
	s64 ns_offset;

	/* read clocksource: */
749
	cycle_now = clocksource_read(clock);
750 751

	/* calculate the delta since the last update_wall_time: */
752
	cycle_delta = (cycle_now - clock->cycle_last) & clock->mask;
753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783

	/* convert to nanoseconds: */
	ns_offset = cyc2ns(clock, cycle_delta);

	return ns_offset;
}

/**
 * __get_realtime_clock_ts - Returns the time of day in a timespec
 * @ts:		pointer to the timespec to be set
 *
 * Returns the time of day in a timespec. Used by
 * do_gettimeofday() and get_realtime_clock_ts().
 */
static inline void __get_realtime_clock_ts(struct timespec *ts)
{
	unsigned long seq;
	s64 nsecs;

	do {
		seq = read_seqbegin(&xtime_lock);

		*ts = xtime;
		nsecs = __get_nsec_offset();

	} while (read_seqretry(&xtime_lock, seq));

	timespec_add_ns(ts, nsecs);
}

/**
784
 * getnstimeofday - Returns the time of day in a timespec
785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836
 * @ts:		pointer to the timespec to be set
 *
 * Returns the time of day in a timespec.
 */
void getnstimeofday(struct timespec *ts)
{
	__get_realtime_clock_ts(ts);
}

EXPORT_SYMBOL(getnstimeofday);

/**
 * do_gettimeofday - Returns the time of day in a timeval
 * @tv:		pointer to the timeval to be set
 *
 * NOTE: Users should be converted to using get_realtime_clock_ts()
 */
void do_gettimeofday(struct timeval *tv)
{
	struct timespec now;

	__get_realtime_clock_ts(&now);
	tv->tv_sec = now.tv_sec;
	tv->tv_usec = now.tv_nsec/1000;
}

EXPORT_SYMBOL(do_gettimeofday);
/**
 * do_settimeofday - Sets the time of day
 * @tv:		pointer to the timespec variable containing the new time
 *
 * Sets the time of day to the new time and update NTP and notify hrtimers
 */
int do_settimeofday(struct timespec *tv)
{
	unsigned long flags;
	time_t wtm_sec, sec = tv->tv_sec;
	long wtm_nsec, nsec = tv->tv_nsec;

	if ((unsigned long)tv->tv_nsec >= NSEC_PER_SEC)
		return -EINVAL;

	write_seqlock_irqsave(&xtime_lock, flags);

	nsec -= __get_nsec_offset();

	wtm_sec  = wall_to_monotonic.tv_sec + (xtime.tv_sec - sec);
	wtm_nsec = wall_to_monotonic.tv_nsec + (xtime.tv_nsec - nsec);

	set_normalized_timespec(&xtime, sec, nsec);
	set_normalized_timespec(&wall_to_monotonic, wtm_sec, wtm_nsec);

837
	clock->error = 0;
838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854
	ntp_clear();

	write_sequnlock_irqrestore(&xtime_lock, flags);

	/* signal hrtimers about time change */
	clock_was_set();

	return 0;
}

EXPORT_SYMBOL(do_settimeofday);

/**
 * change_clocksource - Swaps clocksources if a new one is available
 *
 * Accumulates current time interval and initializes new clocksource
 */
855
static void change_clocksource(void)
856 857 858 859
{
	struct clocksource *new;
	cycle_t now;
	u64 nsec;
860

861
	new = clocksource_get_next();
862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878

	if (clock == new)
		return;

	now = clocksource_read(new);
	nsec =  __get_nsec_offset();
	timespec_add_ns(&xtime, nsec);

	clock = new;
	clock->cycle_last = now;

	clock->error = 0;
	clock->xtime_nsec = 0;
	clocksource_calculate_interval(clock, NTP_INTERVAL_LENGTH);

	printk(KERN_INFO "Time: %s clocksource has been installed.\n",
	       clock->name);
879 880
}
#else
881
static inline void change_clocksource(void) { }
882 883 884 885 886 887 888 889 890 891 892 893 894
#endif

/**
 * timeofday_is_continuous - check to see if timekeeping is free running
 */
int timekeeping_is_continuous(void)
{
	unsigned long seq;
	int ret;

	do {
		seq = read_seqbegin(&xtime_lock);

895
		ret = clock->flags & CLOCK_SOURCE_VALID_FOR_HRES;
896 897 898 899 900 901

	} while (read_seqretry(&xtime_lock, seq));

	return ret;
}

902 903 904 905 906 907 908 909 910 911 912 913 914 915
/**
 * read_persistent_clock -  Return time in seconds from the persistent clock.
 *
 * Weak dummy function for arches that do not yet support it.
 * Returns seconds from epoch using the battery backed persistent clock.
 * Returns zero if unsupported.
 *
 *  XXX - Do be sure to remove it once all arches implement it.
 */
unsigned long __attribute__((weak)) read_persistent_clock(void)
{
	return 0;
}

L
Linus Torvalds 已提交
916
/*
917
 * timekeeping_init - Initializes the clocksource and common timekeeping values
L
Linus Torvalds 已提交
918
 */
919
void __init timekeeping_init(void)
L
Linus Torvalds 已提交
920
{
921
	unsigned long flags;
922
	unsigned long sec = read_persistent_clock();
923 924

	write_seqlock_irqsave(&xtime_lock, flags);
925 926 927

	ntp_clear();

928
	clock = clocksource_get_next();
J
john stultz 已提交
929
	clocksource_calculate_interval(clock, NTP_INTERVAL_LENGTH);
930
	clock->cycle_last = clocksource_read(clock);
931

932 933 934 935 936
	xtime.tv_sec = sec;
	xtime.tv_nsec = 0;
	set_normalized_timespec(&wall_to_monotonic,
		-xtime.tv_sec, -xtime.tv_nsec);

937 938 939 940
	write_sequnlock_irqrestore(&xtime_lock, flags);
}


941
/* flag for if timekeeping is suspended */
942
static int timekeeping_suspended;
943 944 945
/* time in seconds when suspend began */
static unsigned long timekeeping_suspend_time;

946
/**
947 948 949 950
 * timekeeping_resume - Resumes the generic timekeeping subsystem.
 * @dev:	unused
 *
 * This is for the generic clocksource timekeeping.
A
Atsushi Nemoto 已提交
951
 * xtime/wall_to_monotonic/jiffies/etc are
952 953 954 955 956
 * still managed by arch specific suspend/resume code.
 */
static int timekeeping_resume(struct sys_device *dev)
{
	unsigned long flags;
957
	unsigned long now = read_persistent_clock();
958 959

	write_seqlock_irqsave(&xtime_lock, flags);
960 961 962 963 964 965 966 967

	if (now && (now > timekeeping_suspend_time)) {
		unsigned long sleep_length = now - timekeeping_suspend_time;

		xtime.tv_sec += sleep_length;
		wall_to_monotonic.tv_sec -= sleep_length;
	}
	/* re-base the last cycle value */
968
	clock->cycle_last = clocksource_read(clock);
969 970 971
	clock->error = 0;
	timekeeping_suspended = 0;
	write_sequnlock_irqrestore(&xtime_lock, flags);
972 973

	touch_softlockup_watchdog();
974 975
	/* Resume hrtimers */
	clock_was_set();
976

977 978 979 980 981 982 983 984 985
	return 0;
}

static int timekeeping_suspend(struct sys_device *dev, pm_message_t state)
{
	unsigned long flags;

	write_seqlock_irqsave(&xtime_lock, flags);
	timekeeping_suspended = 1;
986
	timekeeping_suspend_time = read_persistent_clock();
987 988 989 990 991 992 993
	write_sequnlock_irqrestore(&xtime_lock, flags);
	return 0;
}

/* sysfs resume/suspend bits for timekeeping */
static struct sysdev_class timekeeping_sysclass = {
	.resume		= timekeeping_resume,
994
	.suspend	= timekeeping_suspend,
995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012
	set_kset_name("timekeeping"),
};

static struct sys_device device_timer = {
	.id		= 0,
	.cls		= &timekeeping_sysclass,
};

static int __init timekeeping_init_device(void)
{
	int error = sysdev_class_register(&timekeeping_sysclass);
	if (!error)
		error = sysdev_register(&device_timer);
	return error;
}

device_initcall(timekeeping_init_device);

1013
/*
1014
 * If the error is already larger, we look ahead even further
1015 1016
 * to compensate for late or lost adjustments.
 */
D
Daniel Walker 已提交
1017 1018
static __always_inline int clocksource_bigadjust(s64 error, s64 *interval,
						 s64 *offset)
1019
{
1020 1021 1022
	s64 tick_error, i;
	u32 look_ahead, adj;
	s32 error2, mult;
1023 1024

	/*
1025 1026 1027 1028 1029 1030 1031
	 * Use the current error value to determine how much to look ahead.
	 * The larger the error the slower we adjust for it to avoid problems
	 * with losing too many ticks, otherwise we would overadjust and
	 * produce an even larger error.  The smaller the adjustment the
	 * faster we try to adjust for it, as lost ticks can do less harm
	 * here.  This is tuned so that an error of about 1 msec is adusted
	 * within about 1 sec (or 2^20 nsec in 2^SHIFT_HZ ticks).
1032
	 */
1033 1034 1035 1036
	error2 = clock->error >> (TICK_LENGTH_SHIFT + 22 - 2 * SHIFT_HZ);
	error2 = abs(error2);
	for (look_ahead = 0; error2 > 0; look_ahead++)
		error2 >>= 2;
1037 1038

	/*
1039 1040
	 * Now calculate the error in (1 << look_ahead) ticks, but first
	 * remove the single look ahead already included in the error.
1041
	 */
D
Daniel Walker 已提交
1042 1043
	tick_error = current_tick_length() >>
		(TICK_LENGTH_SHIFT - clock->shift + 1);
1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054
	tick_error -= clock->xtime_interval >> 1;
	error = ((error - tick_error) >> look_ahead) + tick_error;

	/* Finally calculate the adjustment shift value.  */
	i = *interval;
	mult = 1;
	if (error < 0) {
		error = -error;
		*interval = -*interval;
		*offset = -*offset;
		mult = -1;
1055
	}
1056 1057
	for (adj = 0; error > i; adj++)
		error >>= 1;
1058 1059 1060

	*interval <<= adj;
	*offset <<= adj;
1061
	return mult << adj;
1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075
}

/*
 * Adjust the multiplier to reduce the error value,
 * this is optimized for the most common adjustments of -1,0,1,
 * for other values we can do a bit more work.
 */
static void clocksource_adjust(struct clocksource *clock, s64 offset)
{
	s64 error, interval = clock->cycle_interval;
	int adj;

	error = clock->error >> (TICK_LENGTH_SHIFT - clock->shift - 1);
	if (error > interval) {
1076 1077 1078 1079 1080
		error >>= 2;
		if (likely(error <= interval))
			adj = 1;
		else
			adj = clocksource_bigadjust(error, &interval, &offset);
1081
	} else if (error < -interval) {
1082 1083 1084 1085 1086 1087 1088
		error >>= 2;
		if (likely(error >= -interval)) {
			adj = -1;
			interval = -interval;
			offset = -offset;
		} else
			adj = clocksource_bigadjust(error, &interval, &offset);
1089 1090 1091 1092 1093 1094
	} else
		return;

	clock->mult += adj;
	clock->xtime_interval += interval;
	clock->xtime_nsec -= offset;
D
Daniel Walker 已提交
1095 1096
	clock->error -= (interval - offset) <<
			(TICK_LENGTH_SHIFT - clock->shift);
1097 1098
}

1099
/**
1100 1101 1102 1103 1104 1105
 * update_wall_time - Uses the current clocksource to increment the wall time
 *
 * Called from the timer interrupt, must hold a write on xtime_lock.
 */
static void update_wall_time(void)
{
1106
	cycle_t offset;
1107

1108 1109 1110
	/* Make sure we're fully resumed: */
	if (unlikely(timekeeping_suspended))
		return;
1111

1112 1113 1114 1115 1116
#ifdef CONFIG_GENERIC_TIME
	offset = (clocksource_read(clock) - clock->cycle_last) & clock->mask;
#else
	offset = clock->cycle_interval;
#endif
1117
	clock->xtime_nsec += (s64)xtime.tv_nsec << clock->shift;
1118 1119 1120 1121

	/* normally this loop will run just once, however in the
	 * case of lost or late ticks, it will accumulate correctly.
	 */
1122
	while (offset >= clock->cycle_interval) {
1123
		/* accumulate one interval */
1124 1125 1126 1127 1128 1129 1130 1131 1132
		clock->xtime_nsec += clock->xtime_interval;
		clock->cycle_last += clock->cycle_interval;
		offset -= clock->cycle_interval;

		if (clock->xtime_nsec >= (u64)NSEC_PER_SEC << clock->shift) {
			clock->xtime_nsec -= (u64)NSEC_PER_SEC << clock->shift;
			xtime.tv_sec++;
			second_overflow();
		}
1133

1134
		/* interpolator bits */
1135
		time_interpolator_update(clock->xtime_interval
1136 1137 1138
						>> clock->shift);

		/* accumulate error between NTP and clock interval */
1139 1140 1141
		clock->error += current_tick_length();
		clock->error -= clock->xtime_interval << (TICK_LENGTH_SHIFT - clock->shift);
	}
1142

1143 1144
	/* correct the clock when NTP error is too big */
	clocksource_adjust(clock, offset);
1145 1146

	/* store full nanoseconds into xtime */
1147
	xtime.tv_nsec = (s64)clock->xtime_nsec >> clock->shift;
1148
	clock->xtime_nsec -= (s64)xtime.tv_nsec << clock->shift;
1149 1150

	/* check to see if there is a new clocksource to use */
1151
	change_clocksource();
L
Linus Torvalds 已提交
1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179
}

/*
 * Called from the timer interrupt handler to charge one tick to the current 
 * 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. */
	if (user_tick)
		account_user_time(p, jiffies_to_cputime(1));
	else
		account_system_time(p, HARDIRQ_OFFSET, jiffies_to_cputime(1));
	run_local_timers();
	if (rcu_pending(cpu))
		rcu_check_callbacks(cpu, user_tick);
	scheduler_tick();
 	run_posix_cpu_timers(p);
}

/*
 * Nr of active tasks - counted in fixed-point numbers
 */
static unsigned long count_active_tasks(void)
{
1180
	return nr_active() * FIXED_1;
L
Linus Torvalds 已提交
1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203
}

/*
 * 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
1204 1205 1206 1207 1208 1209 1210 1211 1212
	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
1213 1214 1215 1216 1217 1218 1219
	}
}

/*
 * This read-write spinlock protects us from races in SMP while
 * playing with xtime and avenrun.
 */
1220
__attribute__((weak)) __cacheline_aligned_in_smp DEFINE_SEQLOCK(xtime_lock);
L
Linus Torvalds 已提交
1221 1222 1223 1224 1225 1226 1227 1228

EXPORT_SYMBOL(xtime_lock);

/*
 * This function runs timers and the timer-tq in bottom half context.
 */
static void run_timer_softirq(struct softirq_action *h)
{
1229
	tvec_base_t *base = __get_cpu_var(tvec_bases);
L
Linus Torvalds 已提交
1230

1231
 	hrtimer_run_queues();
L
Linus Torvalds 已提交
1232 1233 1234 1235 1236 1237 1238 1239 1240 1241
	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)
{
	raise_softirq(TIMER_SOFTIRQ);
1242
	softlockup_tick();
L
Linus Torvalds 已提交
1243 1244 1245 1246 1247 1248
}

/*
 * Called by the timer interrupt. xtime_lock must already be taken
 * by the timer IRQ!
 */
1249
static inline void update_times(unsigned long ticks)
L
Linus Torvalds 已提交
1250
{
1251
	update_wall_time();
L
Linus Torvalds 已提交
1252 1253 1254 1255 1256 1257 1258 1259 1260
	calc_load(ticks);
}
  
/*
 * 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...
 */

1261
void do_timer(unsigned long ticks)
L
Linus Torvalds 已提交
1262
{
1263 1264
	jiffies_64 += ticks;
	update_times(ticks);
L
Linus Torvalds 已提交
1265 1266 1267 1268 1269 1270 1271 1272 1273 1274
}

#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)
{
1275
	return alarm_setitimer(seconds);
L
Linus Torvalds 已提交
1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301
}

#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)
{
	return current->tgid;
}

/*
1302 1303 1304 1305
 * 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
1306 1307 1308 1309 1310
 */
asmlinkage long sys_getppid(void)
{
	int pid;

1311 1312 1313
	rcu_read_lock();
	pid = rcu_dereference(current->real_parent)->tgid;
	rcu_read_unlock();
L
Linus Torvalds 已提交
1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345

	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)
{
1346
	wake_up_process((struct task_struct *)__data);
L
Linus Torvalds 已提交
1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399
}

/**
 * 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.
 */
fastcall signed long __sched schedule_timeout(signed long timeout)
{
	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.
		 */
1400
		if (timeout < 0) {
L
Linus Torvalds 已提交
1401
			printk(KERN_ERR "schedule_timeout: wrong timeout "
1402 1403
				"value %lx\n", timeout);
			dump_stack();
L
Linus Torvalds 已提交
1404 1405 1406 1407 1408 1409 1410
			current->state = TASK_RUNNING;
			goto out;
		}
	}

	expire = timeout + jiffies;

1411 1412
	setup_timer(&timer, process_timeout, (unsigned long)current);
	__mod_timer(&timer, expire);
L
Linus Torvalds 已提交
1413 1414 1415 1416 1417 1418 1419 1420 1421 1422
	schedule();
	del_singleshot_timer_sync(&timer);

	timeout = expire - jiffies;

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

1423 1424 1425 1426
/*
 * We can use __set_current_state() here because schedule_timeout() calls
 * schedule() unconditionally.
 */
1427 1428
signed long __sched schedule_timeout_interruptible(signed long timeout)
{
A
Andrew Morton 已提交
1429 1430
	__set_current_state(TASK_INTERRUPTIBLE);
	return schedule_timeout(timeout);
1431 1432 1433 1434 1435
}
EXPORT_SYMBOL(schedule_timeout_interruptible);

signed long __sched schedule_timeout_uninterruptible(signed long timeout)
{
A
Andrew Morton 已提交
1436 1437
	__set_current_state(TASK_UNINTERRUPTIBLE);
	return schedule_timeout(timeout);
1438 1439 1440
}
EXPORT_SYMBOL(schedule_timeout_uninterruptible);

L
Linus Torvalds 已提交
1441 1442 1443 1444 1445 1446
/* Thread ID - the internal kernel "pid" */
asmlinkage long sys_gettid(void)
{
	return current->pid;
}

1447
/**
1448
 * do_sysinfo - fill in sysinfo struct
1449
 * @info: pointer to buffer to fill
L
Linus Torvalds 已提交
1450
 */ 
1451
int do_sysinfo(struct sysinfo *info)
L
Linus Torvalds 已提交
1452 1453 1454 1455 1456
{
	unsigned long mem_total, sav_total;
	unsigned int mem_unit, bitcount;
	unsigned long seq;

1457
	memset(info, 0, sizeof(struct sysinfo));
L
Linus Torvalds 已提交
1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476

	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;
		if (tp.tv_nsec - NSEC_PER_SEC >= 0) {
			tp.tv_nsec = tp.tv_nsec - NSEC_PER_SEC;
			tp.tv_sec++;
		}
1477
		info->uptime = tp.tv_sec + (tp.tv_nsec ? 1 : 0);
L
Linus Torvalds 已提交
1478

1479 1480 1481
		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
1482

1483
		info->procs = nr_threads;
L
Linus Torvalds 已提交
1484 1485
	} while (read_seqretry(&xtime_lock, seq));

1486 1487
	si_meminfo(info);
	si_swapinfo(info);
L
Linus Torvalds 已提交
1488 1489 1490 1491 1492 1493 1494 1495 1496 1497

	/*
	 * 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>
	 */

1498 1499
	mem_total = info->totalram + info->totalswap;
	if (mem_total < info->totalram || mem_total < info->totalswap)
L
Linus Torvalds 已提交
1500 1501
		goto out;
	bitcount = 0;
1502
	mem_unit = info->mem_unit;
L
Linus Torvalds 已提交
1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513
	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
1514
	 * info->mem_unit and set it to 1.  This leaves things compatible
L
Linus Torvalds 已提交
1515 1516 1517 1518
	 * with 2.2.x, and also retains compatibility with earlier 2.4.x
	 * kernels...
	 */

1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537
	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
1538 1539 1540 1541 1542 1543 1544

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

	return 0;
}

1545 1546 1547 1548 1549 1550 1551
/*
 * 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];

1552
static int __devinit init_timers_cpu(int cpu)
L
Linus Torvalds 已提交
1553 1554 1555
{
	int j;
	tvec_base_t *base;
A
Andrew Morton 已提交
1556
	static char __devinitdata tvec_base_done[NR_CPUS];
1557

A
Andrew Morton 已提交
1558
	if (!tvec_base_done[cpu]) {
1559 1560 1561
		static char boot_done;

		if (boot_done) {
A
Andrew Morton 已提交
1562 1563 1564
			/*
			 * The APs use this path later in boot
			 */
1565 1566 1567 1568 1569
			base = kmalloc_node(sizeof(*base), GFP_KERNEL,
						cpu_to_node(cpu));
			if (!base)
				return -ENOMEM;
			memset(base, 0, sizeof(*base));
A
Andrew Morton 已提交
1570
			per_cpu(tvec_bases, cpu) = base;
1571
		} else {
A
Andrew Morton 已提交
1572 1573 1574 1575 1576 1577
			/*
			 * 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.
			 */
1578
			boot_done = 1;
A
Andrew Morton 已提交
1579
			base = &boot_tvec_bases;
1580
		}
A
Andrew Morton 已提交
1581 1582 1583
		tvec_base_done[cpu] = 1;
	} else {
		base = per_cpu(tvec_bases, cpu);
1584
	}
A
Andrew Morton 已提交
1585

1586
	spin_lock_init(&base->lock);
1587 1588
	lockdep_set_class(&base->lock, base_lock_keys + cpu);

L
Linus Torvalds 已提交
1589 1590 1591 1592 1593 1594 1595 1596 1597 1598
	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;
1599
	return 0;
L
Linus Torvalds 已提交
1600 1601 1602
}

#ifdef CONFIG_HOTPLUG_CPU
1603
static void migrate_timer_list(tvec_base_t *new_base, struct list_head *head)
L
Linus Torvalds 已提交
1604 1605 1606 1607 1608
{
	struct timer_list *timer;

	while (!list_empty(head)) {
		timer = list_entry(head->next, struct timer_list, entry);
1609
		detach_timer(timer, 0);
1610
		timer->base = new_base;
L
Linus Torvalds 已提交
1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621
		internal_add_timer(new_base, timer);
	}
}

static void __devinit migrate_timers(int cpu)
{
	tvec_base_t *old_base;
	tvec_base_t *new_base;
	int i;

	BUG_ON(cpu_online(cpu));
1622 1623
	old_base = per_cpu(tvec_bases, cpu);
	new_base = get_cpu_var(tvec_bases);
L
Linus Torvalds 已提交
1624 1625

	local_irq_disable();
1626 1627 1628 1629
	spin_lock(&new_base->lock);
	spin_lock(&old_base->lock);

	BUG_ON(old_base->running_timer);
L
Linus Torvalds 已提交
1630 1631

	for (i = 0; i < TVR_SIZE; i++)
1632 1633 1634 1635 1636 1637 1638 1639
		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);
	}

1640 1641
	spin_unlock(&old_base->lock);
	spin_unlock(&new_base->lock);
L
Linus Torvalds 已提交
1642 1643 1644 1645 1646
	local_irq_enable();
	put_cpu_var(tvec_bases);
}
#endif /* CONFIG_HOTPLUG_CPU */

1647
static int __cpuinit timer_cpu_notify(struct notifier_block *self,
L
Linus Torvalds 已提交
1648 1649 1650 1651 1652
				unsigned long action, void *hcpu)
{
	long cpu = (long)hcpu;
	switch(action) {
	case CPU_UP_PREPARE:
1653 1654
		if (init_timers_cpu(cpu) < 0)
			return NOTIFY_BAD;
L
Linus Torvalds 已提交
1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666
		break;
#ifdef CONFIG_HOTPLUG_CPU
	case CPU_DEAD:
		migrate_timers(cpu);
		break;
#endif
	default:
		break;
	}
	return NOTIFY_OK;
}

1667
static struct notifier_block __cpuinitdata timers_nb = {
L
Linus Torvalds 已提交
1668 1669 1670 1671 1672 1673
	.notifier_call	= timer_cpu_notify,
};


void __init init_timers(void)
{
1674
	int err = timer_cpu_notify(&timers_nb, (unsigned long)CPU_UP_PREPARE,
L
Linus Torvalds 已提交
1675
				(void *)(long)smp_processor_id());
1676 1677

	BUG_ON(err == NOTIFY_BAD);
L
Linus Torvalds 已提交
1678 1679 1680 1681 1682 1683
	register_cpu_notifier(&timers_nb);
	open_softirq(TIMER_SOFTIRQ, run_timer_softirq, NULL);
}

#ifdef CONFIG_TIME_INTERPOLATION

1684 1685
struct time_interpolator *time_interpolator __read_mostly;
static struct time_interpolator *time_interpolator_list __read_mostly;
L
Linus Torvalds 已提交
1686 1687
static DEFINE_SPINLOCK(time_interpolator_lock);

1688
static inline cycles_t time_interpolator_get_cycles(unsigned int src)
L
Linus Torvalds 已提交
1689 1690 1691 1692 1693 1694 1695 1696 1697 1698
{
	unsigned long (*x)(void);

	switch (src)
	{
		case TIME_SOURCE_FUNCTION:
			x = time_interpolator->addr;
			return x();

		case TIME_SOURCE_MMIO64	:
1699
			return readq_relaxed((void __iomem *)time_interpolator->addr);
L
Linus Torvalds 已提交
1700 1701

		case TIME_SOURCE_MMIO32	:
1702
			return readl_relaxed((void __iomem *)time_interpolator->addr);
L
Linus Torvalds 已提交
1703 1704 1705 1706 1707

		default: return get_cycles();
	}
}

1708
static inline u64 time_interpolator_get_counter(int writelock)
L
Linus Torvalds 已提交
1709 1710 1711 1712 1713
{
	unsigned int src = time_interpolator->source;

	if (time_interpolator->jitter)
	{
1714 1715
		cycles_t lcycle;
		cycles_t now;
L
Linus Torvalds 已提交
1716 1717 1718 1719 1720 1721

		do {
			lcycle = time_interpolator->last_cycle;
			now = time_interpolator_get_cycles(src);
			if (lcycle && time_after(lcycle, now))
				return lcycle;
1722 1723 1724 1725 1726 1727 1728 1729 1730

			/* When holding the xtime write lock, there's no need
			 * to add the overhead of the cmpxchg.  Readers are
			 * force to retry until the write lock is released.
			 */
			if (writelock) {
				time_interpolator->last_cycle = now;
				return now;
			}
L
Linus Torvalds 已提交
1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743
			/* Keep track of the last timer value returned. The use of cmpxchg here
			 * will cause contention in an SMP environment.
			 */
		} while (unlikely(cmpxchg(&time_interpolator->last_cycle, lcycle, now) != lcycle));
		return now;
	}
	else
		return time_interpolator_get_cycles(src);
}

void time_interpolator_reset(void)
{
	time_interpolator->offset = 0;
1744
	time_interpolator->last_counter = time_interpolator_get_counter(1);
L
Linus Torvalds 已提交
1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755
}

#define GET_TI_NSECS(count,i) (((((count) - i->last_counter) & (i)->mask) * (i)->nsec_per_cyc) >> (i)->shift)

unsigned long time_interpolator_get_offset(void)
{
	/* If we do not have a time interpolator set up then just return zero */
	if (!time_interpolator)
		return 0;

	return time_interpolator->offset +
1756
		GET_TI_NSECS(time_interpolator_get_counter(0), time_interpolator);
L
Linus Torvalds 已提交
1757 1758 1759 1760 1761
}

#define INTERPOLATOR_ADJUST 65536
#define INTERPOLATOR_MAX_SKIP 10*INTERPOLATOR_ADJUST

1762
void time_interpolator_update(long delta_nsec)
L
Linus Torvalds 已提交
1763 1764 1765 1766 1767 1768 1769 1770
{
	u64 counter;
	unsigned long offset;

	/* If there is no time interpolator set up then do nothing */
	if (!time_interpolator)
		return;

A
Andrew Morton 已提交
1771 1772 1773 1774 1775 1776 1777 1778
	/*
	 * The interpolator compensates for late ticks by accumulating the late
	 * time in time_interpolator->offset. A tick earlier than expected will
	 * lead to a reset of the offset and a corresponding jump of the clock
	 * forward. Again this only works if the interpolator clock is running
	 * slightly slower than the regular clock and the tuning logic insures
	 * that.
	 */
L
Linus Torvalds 已提交
1779

1780
	counter = time_interpolator_get_counter(1);
A
Andrew Morton 已提交
1781 1782
	offset = time_interpolator->offset +
			GET_TI_NSECS(counter, time_interpolator);
L
Linus Torvalds 已提交
1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798

	if (delta_nsec < 0 || (unsigned long) delta_nsec < offset)
		time_interpolator->offset = offset - delta_nsec;
	else {
		time_interpolator->skips++;
		time_interpolator->ns_skipped += delta_nsec - offset;
		time_interpolator->offset = 0;
	}
	time_interpolator->last_counter = counter;

	/* Tuning logic for time interpolator invoked every minute or so.
	 * Decrease interpolator clock speed if no skips occurred and an offset is carried.
	 * Increase interpolator clock speed if we skip too much time.
	 */
	if (jiffies % INTERPOLATOR_ADJUST == 0)
	{
1799
		if (time_interpolator->skips == 0 && time_interpolator->offset > tick_nsec)
L
Linus Torvalds 已提交
1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822
			time_interpolator->nsec_per_cyc--;
		if (time_interpolator->ns_skipped > INTERPOLATOR_MAX_SKIP && time_interpolator->offset == 0)
			time_interpolator->nsec_per_cyc++;
		time_interpolator->skips = 0;
		time_interpolator->ns_skipped = 0;
	}
}

static inline int
is_better_time_interpolator(struct time_interpolator *new)
{
	if (!time_interpolator)
		return 1;
	return new->frequency > 2*time_interpolator->frequency ||
	    (unsigned long)new->drift < (unsigned long)time_interpolator->drift;
}

void
register_time_interpolator(struct time_interpolator *ti)
{
	unsigned long flags;

	/* Sanity check */
1823
	BUG_ON(ti->frequency == 0 || ti->mask == 0);
L
Linus Torvalds 已提交
1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877

	ti->nsec_per_cyc = ((u64)NSEC_PER_SEC << ti->shift) / ti->frequency;
	spin_lock(&time_interpolator_lock);
	write_seqlock_irqsave(&xtime_lock, flags);
	if (is_better_time_interpolator(ti)) {
		time_interpolator = ti;
		time_interpolator_reset();
	}
	write_sequnlock_irqrestore(&xtime_lock, flags);

	ti->next = time_interpolator_list;
	time_interpolator_list = ti;
	spin_unlock(&time_interpolator_lock);
}

void
unregister_time_interpolator(struct time_interpolator *ti)
{
	struct time_interpolator *curr, **prev;
	unsigned long flags;

	spin_lock(&time_interpolator_lock);
	prev = &time_interpolator_list;
	for (curr = *prev; curr; curr = curr->next) {
		if (curr == ti) {
			*prev = curr->next;
			break;
		}
		prev = &curr->next;
	}

	write_seqlock_irqsave(&xtime_lock, flags);
	if (ti == time_interpolator) {
		/* we lost the best time-interpolator: */
		time_interpolator = NULL;
		/* find the next-best interpolator */
		for (curr = time_interpolator_list; curr; curr = curr->next)
			if (is_better_time_interpolator(curr))
				time_interpolator = curr;
		time_interpolator_reset();
	}
	write_sequnlock_irqrestore(&xtime_lock, flags);
	spin_unlock(&time_interpolator_lock);
}
#endif /* CONFIG_TIME_INTERPOLATION */

/**
 * 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;

1878 1879
	while (timeout)
		timeout = schedule_timeout_uninterruptible(timeout);
L
Linus Torvalds 已提交
1880 1881 1882 1883 1884
}

EXPORT_SYMBOL(msleep);

/**
1885
 * msleep_interruptible - sleep waiting for signals
L
Linus Torvalds 已提交
1886 1887 1888 1889 1890 1891
 * @msecs: Time in milliseconds to sleep for
 */
unsigned long msleep_interruptible(unsigned int msecs)
{
	unsigned long timeout = msecs_to_jiffies(msecs) + 1;

1892 1893
	while (timeout && !signal_pending(current))
		timeout = schedule_timeout_interruptible(timeout);
L
Linus Torvalds 已提交
1894 1895 1896 1897
	return jiffies_to_msecs(timeout);
}

EXPORT_SYMBOL(msleep_interruptible);