mutex.c 23.0 KB
Newer Older
I
Ingo Molnar 已提交
1
/*
2
 * kernel/locking/mutex.c
I
Ingo Molnar 已提交
3 4 5 6 7 8 9 10 11 12
 *
 * Mutexes: blocking mutual exclusion locks
 *
 * Started by Ingo Molnar:
 *
 *  Copyright (C) 2004, 2005, 2006 Red Hat, Inc., Ingo Molnar <mingo@redhat.com>
 *
 * Many thanks to Arjan van de Ven, Thomas Gleixner, Steven Rostedt and
 * David Howells for suggestions and improvements.
 *
13 14 15 16 17
 *  - Adaptive spinning for mutexes by Peter Zijlstra. (Ported to mainline
 *    from the -rt tree, where it was originally implemented for rtmutexes
 *    by Steven Rostedt, based on work by Gregory Haskins, Peter Morreale
 *    and Sven Dietrich.
 *
18
 * Also see Documentation/locking/mutex-design.txt.
I
Ingo Molnar 已提交
19 20
 */
#include <linux/mutex.h>
21
#include <linux/ww_mutex.h>
I
Ingo Molnar 已提交
22
#include <linux/sched.h>
23
#include <linux/sched/rt.h>
24
#include <linux/export.h>
I
Ingo Molnar 已提交
25 26
#include <linux/spinlock.h>
#include <linux/interrupt.h>
27
#include <linux/debug_locks.h>
28
#include <linux/osq_lock.h>
I
Ingo Molnar 已提交
29 30 31 32 33 34 35

#ifdef CONFIG_DEBUG_MUTEXES
# include "mutex-debug.h"
#else
# include "mutex.h"
#endif

36 37
void
__mutex_init(struct mutex *lock, const char *name, struct lock_class_key *key)
I
Ingo Molnar 已提交
38
{
39
	atomic_long_set(&lock->owner, 0);
I
Ingo Molnar 已提交
40 41
	spin_lock_init(&lock->wait_lock);
	INIT_LIST_HEAD(&lock->wait_list);
42
#ifdef CONFIG_MUTEX_SPIN_ON_OWNER
43
	osq_lock_init(&lock->osq);
44
#endif
I
Ingo Molnar 已提交
45

46
	debug_mutex_init(lock, name, key);
I
Ingo Molnar 已提交
47 48 49
}
EXPORT_SYMBOL(__mutex_init);

50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136
/*
 * @owner: contains: 'struct task_struct *' to the current lock owner,
 * NULL means not owned. Since task_struct pointers are aligned at
 * ARCH_MIN_TASKALIGN (which is at least sizeof(void *)), we have low
 * bits to store extra state.
 *
 * Bit0 indicates a non-empty waiter list; unlock must issue a wakeup.
 */
#define MUTEX_FLAG_WAITERS	0x01

#define MUTEX_FLAGS		0x03

static inline struct task_struct *__owner_task(unsigned long owner)
{
	return (struct task_struct *)(owner & ~MUTEX_FLAGS);
}

static inline unsigned long __owner_flags(unsigned long owner)
{
	return owner & MUTEX_FLAGS;
}

/*
 * Actual trylock that will work on any unlocked state.
 */
static inline bool __mutex_trylock(struct mutex *lock)
{
	unsigned long owner, curr = (unsigned long)current;

	owner = atomic_long_read(&lock->owner);
	for (;;) { /* must loop, can race against a flag */
		unsigned long old;

		if (__owner_task(owner))
			return false;

		old = atomic_long_cmpxchg_acquire(&lock->owner, owner,
						  curr | __owner_flags(owner));
		if (old == owner)
			return true;

		owner = old;
	}
}

#ifndef CONFIG_DEBUG_LOCK_ALLOC
/*
 * Lockdep annotations are contained to the slow paths for simplicity.
 * There is nothing that would stop spreading the lockdep annotations outwards
 * except more code.
 */

/*
 * Optimistic trylock that only works in the uncontended case. Make sure to
 * follow with a __mutex_trylock() before failing.
 */
static __always_inline bool __mutex_trylock_fast(struct mutex *lock)
{
	unsigned long curr = (unsigned long)current;

	if (!atomic_long_cmpxchg_acquire(&lock->owner, 0UL, curr))
		return true;

	return false;
}

static __always_inline bool __mutex_unlock_fast(struct mutex *lock)
{
	unsigned long curr = (unsigned long)current;

	if (atomic_long_cmpxchg_release(&lock->owner, curr, 0UL) == curr)
		return true;

	return false;
}
#endif

static inline void __mutex_set_flag(struct mutex *lock, unsigned long flag)
{
	atomic_long_or(flag, &lock->owner);
}

static inline void __mutex_clear_flag(struct mutex *lock, unsigned long flag)
{
	atomic_long_andnot(flag, &lock->owner);
}

P
Peter Zijlstra 已提交
137
#ifndef CONFIG_DEBUG_LOCK_ALLOC
I
Ingo Molnar 已提交
138 139 140 141 142 143
/*
 * We split the mutex lock/unlock logic into separate fastpath and
 * slowpath functions, to reduce the register pressure on the fastpath.
 * We also put the fastpath first in the kernel image, to make sure the
 * branch is predicted by the CPU as default-untaken.
 */
144
static void __sched __mutex_lock_slowpath(struct mutex *lock);
I
Ingo Molnar 已提交
145

146
/**
I
Ingo Molnar 已提交
147 148 149 150 151 152 153 154 155
 * mutex_lock - acquire the mutex
 * @lock: the mutex to be acquired
 *
 * Lock the mutex exclusively for this task. If the mutex is not
 * available right now, it will sleep until it can get it.
 *
 * The mutex must later on be released by the same task that
 * acquired it. Recursive locking is not allowed. The task
 * may not exit without first unlocking the mutex. Also, kernel
156
 * memory where the mutex resides must not be freed with
I
Ingo Molnar 已提交
157 158 159 160 161 162 163 164 165 166
 * the mutex still locked. The mutex must first be initialized
 * (or statically defined) before it can be locked. memset()-ing
 * the mutex to 0 is not allowed.
 *
 * ( The CONFIG_DEBUG_MUTEXES .config option turns on debugging
 *   checks that will enforce the restrictions and will also do
 *   deadlock debugging. )
 *
 * This function is similar to (but not equivalent to) down().
 */
167
void __sched mutex_lock(struct mutex *lock)
I
Ingo Molnar 已提交
168
{
169
	might_sleep();
I
Ingo Molnar 已提交
170

171 172 173
	if (!__mutex_trylock_fast(lock))
		__mutex_lock_slowpath(lock);
}
I
Ingo Molnar 已提交
174
EXPORT_SYMBOL(mutex_lock);
P
Peter Zijlstra 已提交
175
#endif
I
Ingo Molnar 已提交
176

177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217
static __always_inline void ww_mutex_lock_acquired(struct ww_mutex *ww,
						   struct ww_acquire_ctx *ww_ctx)
{
#ifdef CONFIG_DEBUG_MUTEXES
	/*
	 * If this WARN_ON triggers, you used ww_mutex_lock to acquire,
	 * but released with a normal mutex_unlock in this call.
	 *
	 * This should never happen, always use ww_mutex_unlock.
	 */
	DEBUG_LOCKS_WARN_ON(ww->ctx);

	/*
	 * Not quite done after calling ww_acquire_done() ?
	 */
	DEBUG_LOCKS_WARN_ON(ww_ctx->done_acquire);

	if (ww_ctx->contending_lock) {
		/*
		 * After -EDEADLK you tried to
		 * acquire a different ww_mutex? Bad!
		 */
		DEBUG_LOCKS_WARN_ON(ww_ctx->contending_lock != ww);

		/*
		 * You called ww_mutex_lock after receiving -EDEADLK,
		 * but 'forgot' to unlock everything else first?
		 */
		DEBUG_LOCKS_WARN_ON(ww_ctx->acquired > 0);
		ww_ctx->contending_lock = NULL;
	}

	/*
	 * Naughty, using a different class will lead to undefined behavior!
	 */
	DEBUG_LOCKS_WARN_ON(ww_ctx->ww_class != ww->ww_class);
#endif
	ww_ctx->acquired++;
}

/*
218
 * After acquiring lock with fastpath or when we lost out in contested
219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243
 * slowpath, set ctx and wake up any waiters so they can recheck.
 */
static __always_inline void
ww_mutex_set_context_fastpath(struct ww_mutex *lock,
			       struct ww_acquire_ctx *ctx)
{
	unsigned long flags;
	struct mutex_waiter *cur;

	ww_mutex_lock_acquired(lock, ctx);

	lock->ctx = ctx;

	/*
	 * The lock->ctx update should be visible on all cores before
	 * the atomic read is done, otherwise contended waiters might be
	 * missed. The contended waiters will either see ww_ctx == NULL
	 * and keep spinning, or it will acquire wait_lock, add itself
	 * to waiter list and sleep.
	 */
	smp_mb(); /* ^^^ */

	/*
	 * Check if lock is contended, if not there is nobody to wake up
	 */
244
	if (likely(!(atomic_long_read(&lock->base.owner) & MUTEX_FLAG_WAITERS)))
245 246 247 248 249 250 251 252 253 254 255 256 257 258
		return;

	/*
	 * Uh oh, we raced in fastpath, wake up everyone in this case,
	 * so they can see the new lock->ctx.
	 */
	spin_lock_mutex(&lock->base.wait_lock, flags);
	list_for_each_entry(cur, &lock->base.wait_list, list) {
		debug_mutex_wake_waiter(&lock->base, cur);
		wake_up_process(cur->task);
	}
	spin_unlock_mutex(&lock->base.wait_lock, flags);
}

259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282
/*
 * After acquiring lock in the slowpath set ctx and wake up any
 * waiters so they can recheck.
 *
 * Callers must hold the mutex wait_lock.
 */
static __always_inline void
ww_mutex_set_context_slowpath(struct ww_mutex *lock,
			      struct ww_acquire_ctx *ctx)
{
	struct mutex_waiter *cur;

	ww_mutex_lock_acquired(lock, ctx);
	lock->ctx = ctx;

	/*
	 * Give any possible sleeping processes the chance to wake up,
	 * so they can recheck if they have to back off.
	 */
	list_for_each_entry(cur, &lock->base.wait_list, list) {
		debug_mutex_wake_waiter(&lock->base, cur);
		wake_up_process(cur->task);
	}
}
283

284 285 286 287 288 289
#ifdef CONFIG_MUTEX_SPIN_ON_OWNER
/*
 * Look out! "owner" is an entirely speculative pointer
 * access and not reliable.
 */
static noinline
290
bool mutex_spin_on_owner(struct mutex *lock, struct task_struct *owner)
291
{
292
	bool ret = true;
293

294
	rcu_read_lock();
295
	while (__mutex_owner(lock) == owner) {
296 297
		/*
		 * Ensure we emit the owner->on_cpu, dereference _after_
298 299
		 * checking lock->owner still matches owner. If that fails,
		 * owner might point to freed memory. If it still matches,
300 301 302 303 304 305 306 307
		 * the rcu_read_lock() ensures the memory stays valid.
		 */
		barrier();

		if (!owner->on_cpu || need_resched()) {
			ret = false;
			break;
		}
308

309
		cpu_relax_lowlatency();
310 311 312
	}
	rcu_read_unlock();

313
	return ret;
314
}
315 316 317 318 319 320

/*
 * Initial check for entering the mutex spinning loop
 */
static inline int mutex_can_spin_on_owner(struct mutex *lock)
{
321
	struct task_struct *owner;
322 323
	int retval = 1;

324 325 326
	if (need_resched())
		return 0;

327
	rcu_read_lock();
328
	owner = __mutex_owner(lock);
329 330
	if (owner)
		retval = owner->on_cpu;
331
	rcu_read_unlock();
332

333
	/*
334 335 336
	 * If lock->owner is not set, the mutex has been released. Return true
	 * such that we'll trylock in the spin path, which is a faster option
	 * than the blocking slow path.
337 338 339
	 */
	return retval;
}
340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364

/*
 * Optimistic spinning.
 *
 * We try to spin for acquisition when we find that the lock owner
 * is currently running on a (different) CPU and while we don't
 * need to reschedule. The rationale is that if the lock owner is
 * running, it is likely to release the lock soon.
 *
 * The mutex spinners are queued up using MCS lock so that only one
 * spinner can compete for the mutex. However, if mutex spinning isn't
 * going to happen, there is no point in going through the lock/unlock
 * overhead.
 *
 * Returns true when the lock was taken, otherwise false, indicating
 * that we need to jump to the slowpath and sleep.
 */
static bool mutex_optimistic_spin(struct mutex *lock,
				  struct ww_acquire_ctx *ww_ctx, const bool use_ww_ctx)
{
	struct task_struct *task = current;

	if (!mutex_can_spin_on_owner(lock))
		goto done;

365 366 367 368 369
	/*
	 * In order to avoid a stampede of mutex spinners trying to
	 * acquire the mutex all at once, the spinners need to take a
	 * MCS (queued) lock first before spinning on the owner field.
	 */
370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387
	if (!osq_lock(&lock->osq))
		goto done;

	while (true) {
		struct task_struct *owner;

		if (use_ww_ctx && ww_ctx->acquired > 0) {
			struct ww_mutex *ww;

			ww = container_of(lock, struct ww_mutex, base);
			/*
			 * If ww->ctx is set the contents are undefined, only
			 * by acquiring wait_lock there is a guarantee that
			 * they are not invalid when reading.
			 *
			 * As such, when deadlock detection needs to be
			 * performed the optimistic spinning cannot be done.
			 */
388
			if (READ_ONCE(ww->ctx))
389 390 391 392 393 394 395
				break;
		}

		/*
		 * If there's an owner, wait for it to either
		 * release the lock or go to sleep.
		 */
396
		owner = __mutex_owner(lock);
397 398 399 400
		if (owner && !mutex_spin_on_owner(lock, owner))
			break;

		/* Try to acquire the mutex if it is unlocked. */
401
		if (__mutex_trylock(lock)) {
402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421
			osq_unlock(&lock->osq);
			return true;
		}

		/*
		 * The cpu_relax() call is a compiler barrier which forces
		 * everything in this loop to be re-loaded. We don't need
		 * memory barriers as we'll eventually observe the right
		 * values at the cost of a few extra spins.
		 */
		cpu_relax_lowlatency();
	}

	osq_unlock(&lock->osq);
done:
	/*
	 * If we fell out of the spin path because of need_resched(),
	 * reschedule now, before we try-lock the mutex. This avoids getting
	 * scheduled out right after we obtained the mutex.
	 */
422 423 424 425 426 427
	if (need_resched()) {
		/*
		 * We _should_ have TASK_RUNNING here, but just in case
		 * we do not, make it so, otherwise we might get stuck.
		 */
		__set_current_state(TASK_RUNNING);
428
		schedule_preempt_disabled();
429
	}
430 431 432 433 434 435 436 437 438

	return false;
}
#else
static bool mutex_optimistic_spin(struct mutex *lock,
				  struct ww_acquire_ctx *ww_ctx, const bool use_ww_ctx)
{
	return false;
}
439 440
#endif

441
static noinline void __sched __mutex_unlock_slowpath(struct mutex *lock, unsigned long ip);
I
Ingo Molnar 已提交
442

443
/**
I
Ingo Molnar 已提交
444 445 446 447 448 449 450 451 452 453
 * mutex_unlock - release the mutex
 * @lock: the mutex to be released
 *
 * Unlock a mutex that has been locked by this task previously.
 *
 * This function must not be used in interrupt context. Unlocking
 * of a not locked mutex is not allowed.
 *
 * This function is similar to (but not equivalent to) up().
 */
454
void __sched mutex_unlock(struct mutex *lock)
I
Ingo Molnar 已提交
455
{
456 457 458
#ifndef CONFIG_DEBUG_LOCK_ALLOC
	if (__mutex_unlock_fast(lock))
		return;
459
#endif
460
	__mutex_unlock_slowpath(lock, _RET_IP_);
I
Ingo Molnar 已提交
461 462 463
}
EXPORT_SYMBOL(mutex_unlock);

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
/**
 * ww_mutex_unlock - release the w/w mutex
 * @lock: the mutex to be released
 *
 * Unlock a mutex that has been locked by this task previously with any of the
 * ww_mutex_lock* functions (with or without an acquire context). It is
 * forbidden to release the locks after releasing the acquire context.
 *
 * This function must not be used in interrupt context. Unlocking
 * of a unlocked mutex is not allowed.
 */
void __sched ww_mutex_unlock(struct ww_mutex *lock)
{
	/*
	 * The unlocking fastpath is the 0->1 transition from 'locked'
	 * into 'unlocked' state:
	 */
	if (lock->ctx) {
#ifdef CONFIG_DEBUG_MUTEXES
		DEBUG_LOCKS_WARN_ON(!lock->ctx->acquired);
#endif
		if (lock->ctx->acquired > 0)
			lock->ctx->acquired--;
		lock->ctx = NULL;
	}

490
	mutex_unlock(&lock->base);
491 492 493 494
}
EXPORT_SYMBOL(ww_mutex_unlock);

static inline int __sched
495
__ww_mutex_lock_check_stamp(struct mutex *lock, struct ww_acquire_ctx *ctx)
496 497
{
	struct ww_mutex *ww = container_of(lock, struct ww_mutex, base);
498
	struct ww_acquire_ctx *hold_ctx = READ_ONCE(ww->ctx);
499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514

	if (!hold_ctx)
		return 0;

	if (ctx->stamp - hold_ctx->stamp <= LONG_MAX &&
	    (ctx->stamp != hold_ctx->stamp || ctx > hold_ctx)) {
#ifdef CONFIG_DEBUG_MUTEXES
		DEBUG_LOCKS_WARN_ON(ctx->contending_lock);
		ctx->contending_lock = ww;
#endif
		return -EDEADLK;
	}

	return 0;
}

I
Ingo Molnar 已提交
515 516 517
/*
 * Lock a mutex (possibly interruptible), slowpath:
 */
518
static __always_inline int __sched
P
Peter Zijlstra 已提交
519
__mutex_lock_common(struct mutex *lock, long state, unsigned int subclass,
520
		    struct lockdep_map *nest_lock, unsigned long ip,
521
		    struct ww_acquire_ctx *ww_ctx, const bool use_ww_ctx)
I
Ingo Molnar 已提交
522 523 524
{
	struct task_struct *task = current;
	struct mutex_waiter waiter;
525
	unsigned long flags;
526
	int ret;
I
Ingo Molnar 已提交
527

528 529 530 531 532 533
	if (use_ww_ctx) {
		struct ww_mutex *ww = container_of(lock, struct ww_mutex, base);
		if (unlikely(ww_ctx == READ_ONCE(ww->ctx)))
			return -EALREADY;
	}

P
Peter Zijlstra 已提交
534
	preempt_disable();
535
	mutex_acquire_nest(&lock->dep_map, subclass, 0, nest_lock, ip);
536

537
	if (__mutex_trylock(lock) || mutex_optimistic_spin(lock, ww_ctx, use_ww_ctx)) {
538
		/* got the lock, yay! */
539 540 541 542 543 544 545
		lock_acquired(&lock->dep_map, ip);
		if (use_ww_ctx) {
			struct ww_mutex *ww;
			ww = container_of(lock, struct ww_mutex, base);

			ww_mutex_set_context_fastpath(ww, ww_ctx);
		}
546 547
		preempt_enable();
		return 0;
548
	}
549

550
	spin_lock_mutex(&lock->wait_lock, flags);
551
	/*
552
	 * After waiting to acquire the wait_lock, try again.
553
	 */
554
	if (__mutex_trylock(lock))
555 556
		goto skip_wait;

557
	debug_mutex_lock_common(lock, &waiter);
558
	debug_mutex_add_waiter(lock, &waiter, task);
I
Ingo Molnar 已提交
559 560 561 562 563

	/* add waiting tasks to the end of the waitqueue (FIFO): */
	list_add_tail(&waiter.list, &lock->wait_list);
	waiter.task = task;

564 565 566
	if (list_first_entry(&lock->wait_list, struct mutex_waiter, list) == &waiter)
		__mutex_set_flag(lock, MUTEX_FLAG_WAITERS);

P
Peter Zijlstra 已提交
567
	lock_contended(&lock->dep_map, ip);
568

I
Ingo Molnar 已提交
569
	for (;;) {
570
		if (__mutex_trylock(lock))
I
Ingo Molnar 已提交
571 572 573 574 575 576
			break;

		/*
		 * got a signal? (This code gets eliminated in the
		 * TASK_UNINTERRUPTIBLE case.)
		 */
577
		if (unlikely(signal_pending_state(state, task))) {
578 579 580
			ret = -EINTR;
			goto err;
		}
I
Ingo Molnar 已提交
581

582
		if (use_ww_ctx && ww_ctx->acquired > 0) {
583
			ret = __ww_mutex_lock_check_stamp(lock, ww_ctx);
584 585
			if (ret)
				goto err;
I
Ingo Molnar 已提交
586
		}
587

I
Ingo Molnar 已提交
588 589
		__set_task_state(task, state);

L
Lucas De Marchi 已提交
590
		/* didn't get the lock, go to sleep: */
591
		spin_unlock_mutex(&lock->wait_lock, flags);
592
		schedule_preempt_disabled();
593
		spin_lock_mutex(&lock->wait_lock, flags);
I
Ingo Molnar 已提交
594
	}
595 596
	__set_task_state(task, TASK_RUNNING);

597
	mutex_remove_waiter(lock, &waiter, task);
598
	if (likely(list_empty(&lock->wait_list)))
599 600
		__mutex_clear_flag(lock, MUTEX_FLAG_WAITERS);

601
	debug_mutex_free_waiter(&waiter);
I
Ingo Molnar 已提交
602

603 604
skip_wait:
	/* got the lock - cleanup and rejoice! */
P
Peter Zijlstra 已提交
605
	lock_acquired(&lock->dep_map, ip);
I
Ingo Molnar 已提交
606

607
	if (use_ww_ctx) {
608
		struct ww_mutex *ww = container_of(lock, struct ww_mutex, base);
609
		ww_mutex_set_context_slowpath(ww, ww_ctx);
610 611
	}

612
	spin_unlock_mutex(&lock->wait_lock, flags);
P
Peter Zijlstra 已提交
613
	preempt_enable();
I
Ingo Molnar 已提交
614
	return 0;
615 616

err:
617
	mutex_remove_waiter(lock, &waiter, task);
618 619 620 621 622
	spin_unlock_mutex(&lock->wait_lock, flags);
	debug_mutex_free_waiter(&waiter);
	mutex_release(&lock->dep_map, 1, ip);
	preempt_enable();
	return ret;
I
Ingo Molnar 已提交
623 624
}

625 626 627 628 629
#ifdef CONFIG_DEBUG_LOCK_ALLOC
void __sched
mutex_lock_nested(struct mutex *lock, unsigned int subclass)
{
	might_sleep();
630
	__mutex_lock_common(lock, TASK_UNINTERRUPTIBLE,
631
			    subclass, NULL, _RET_IP_, NULL, 0);
632 633 634
}

EXPORT_SYMBOL_GPL(mutex_lock_nested);
635

636 637 638 639
void __sched
_mutex_lock_nest_lock(struct mutex *lock, struct lockdep_map *nest)
{
	might_sleep();
640
	__mutex_lock_common(lock, TASK_UNINTERRUPTIBLE,
641
			    0, nest, _RET_IP_, NULL, 0);
642 643 644
}
EXPORT_SYMBOL_GPL(_mutex_lock_nest_lock);

L
Liam R. Howlett 已提交
645 646 647 648
int __sched
mutex_lock_killable_nested(struct mutex *lock, unsigned int subclass)
{
	might_sleep();
649
	return __mutex_lock_common(lock, TASK_KILLABLE,
650
				   subclass, NULL, _RET_IP_, NULL, 0);
L
Liam R. Howlett 已提交
651 652 653
}
EXPORT_SYMBOL_GPL(mutex_lock_killable_nested);

654 655 656 657
int __sched
mutex_lock_interruptible_nested(struct mutex *lock, unsigned int subclass)
{
	might_sleep();
658
	return __mutex_lock_common(lock, TASK_INTERRUPTIBLE,
659
				   subclass, NULL, _RET_IP_, NULL, 0);
660 661
}
EXPORT_SYMBOL_GPL(mutex_lock_interruptible_nested);
662

663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687
static inline int
ww_mutex_deadlock_injection(struct ww_mutex *lock, struct ww_acquire_ctx *ctx)
{
#ifdef CONFIG_DEBUG_WW_MUTEX_SLOWPATH
	unsigned tmp;

	if (ctx->deadlock_inject_countdown-- == 0) {
		tmp = ctx->deadlock_inject_interval;
		if (tmp > UINT_MAX/4)
			tmp = UINT_MAX;
		else
			tmp = tmp*2 + tmp + tmp/2;

		ctx->deadlock_inject_interval = tmp;
		ctx->deadlock_inject_countdown = tmp;
		ctx->contending_lock = lock;

		ww_mutex_unlock(lock);

		return -EDEADLK;
	}
#endif

	return 0;
}
688 689 690 691

int __sched
__ww_mutex_lock(struct ww_mutex *lock, struct ww_acquire_ctx *ctx)
{
692 693
	int ret;

694
	might_sleep();
695
	ret =  __mutex_lock_common(&lock->base, TASK_UNINTERRUPTIBLE,
696
				   0, &ctx->dep_map, _RET_IP_, ctx, 1);
697
	if (!ret && ctx->acquired > 1)
698 699 700
		return ww_mutex_deadlock_injection(lock, ctx);

	return ret;
701 702 703 704 705 706
}
EXPORT_SYMBOL_GPL(__ww_mutex_lock);

int __sched
__ww_mutex_lock_interruptible(struct ww_mutex *lock, struct ww_acquire_ctx *ctx)
{
707 708
	int ret;

709
	might_sleep();
710
	ret = __mutex_lock_common(&lock->base, TASK_INTERRUPTIBLE,
711
				  0, &ctx->dep_map, _RET_IP_, ctx, 1);
712

713
	if (!ret && ctx->acquired > 1)
714 715 716
		return ww_mutex_deadlock_injection(lock, ctx);

	return ret;
717 718 719
}
EXPORT_SYMBOL_GPL(__ww_mutex_lock_interruptible);

720 721
#endif

I
Ingo Molnar 已提交
722 723 724
/*
 * Release the lock, slowpath:
 */
725
static noinline void __sched __mutex_unlock_slowpath(struct mutex *lock, unsigned long ip)
I
Ingo Molnar 已提交
726
{
727
	unsigned long owner, flags;
728
	WAKE_Q(wake_q);
I
Ingo Molnar 已提交
729

730 731
	mutex_release(&lock->dep_map, 1, ip);

I
Ingo Molnar 已提交
732
	/*
733 734
	 * Release the lock before (potentially) taking the spinlock
	 * such that other contenders can get on with things ASAP.
I
Ingo Molnar 已提交
735
	 */
736 737 738
	owner = atomic_long_fetch_and_release(MUTEX_FLAGS, &lock->owner);
	if (!__owner_flags(owner))
		return;
I
Ingo Molnar 已提交
739

740 741 742
	spin_lock_mutex(&lock->wait_lock, flags);
	debug_mutex_unlock(lock);

I
Ingo Molnar 已提交
743 744 745 746 747 748 749
	if (!list_empty(&lock->wait_list)) {
		/* get the first entry from the wait-list: */
		struct mutex_waiter *waiter =
				list_entry(lock->wait_list.next,
					   struct mutex_waiter, list);

		debug_mutex_wake_waiter(lock, waiter);
750
		wake_q_add(&wake_q, waiter->task);
I
Ingo Molnar 已提交
751 752
	}

753
	spin_unlock_mutex(&lock->wait_lock, flags);
754
	wake_up_q(&wake_q);
I
Ingo Molnar 已提交
755 756
}

P
Peter Zijlstra 已提交
757
#ifndef CONFIG_DEBUG_LOCK_ALLOC
I
Ingo Molnar 已提交
758 759 760 761
/*
 * Here come the less common (and hence less performance-critical) APIs:
 * mutex_lock_interruptible() and mutex_trylock().
 */
762
static noinline int __sched
763
__mutex_lock_killable_slowpath(struct mutex *lock);
L
Liam R. Howlett 已提交
764

765
static noinline int __sched
766
__mutex_lock_interruptible_slowpath(struct mutex *lock);
I
Ingo Molnar 已提交
767

768 769
/**
 * mutex_lock_interruptible - acquire the mutex, interruptible
I
Ingo Molnar 已提交
770 771 772 773 774 775 776 777 778
 * @lock: the mutex to be acquired
 *
 * Lock the mutex like mutex_lock(), and return 0 if the mutex has
 * been acquired or sleep until the mutex becomes available. If a
 * signal arrives while waiting for the lock then this function
 * returns -EINTR.
 *
 * This function is similar to (but not equivalent to) down_interruptible().
 */
779
int __sched mutex_lock_interruptible(struct mutex *lock)
I
Ingo Molnar 已提交
780
{
781
	might_sleep();
782 783

	if (__mutex_trylock_fast(lock))
784
		return 0;
785 786

	return __mutex_lock_interruptible_slowpath(lock);
I
Ingo Molnar 已提交
787 788 789 790
}

EXPORT_SYMBOL(mutex_lock_interruptible);

791
int __sched mutex_lock_killable(struct mutex *lock)
L
Liam R. Howlett 已提交
792 793
{
	might_sleep();
794 795

	if (__mutex_trylock_fast(lock))
796
		return 0;
797 798

	return __mutex_lock_killable_slowpath(lock);
L
Liam R. Howlett 已提交
799 800 801
}
EXPORT_SYMBOL(mutex_lock_killable);

802 803
static noinline void __sched
__mutex_lock_slowpath(struct mutex *lock)
P
Peter Zijlstra 已提交
804
{
805
	__mutex_lock_common(lock, TASK_UNINTERRUPTIBLE, 0,
806
			    NULL, _RET_IP_, NULL, 0);
P
Peter Zijlstra 已提交
807 808
}

809
static noinline int __sched
810
__mutex_lock_killable_slowpath(struct mutex *lock)
L
Liam R. Howlett 已提交
811
{
812
	return __mutex_lock_common(lock, TASK_KILLABLE, 0,
813
				   NULL, _RET_IP_, NULL, 0);
L
Liam R. Howlett 已提交
814 815
}

816
static noinline int __sched
817
__mutex_lock_interruptible_slowpath(struct mutex *lock)
I
Ingo Molnar 已提交
818
{
819
	return __mutex_lock_common(lock, TASK_INTERRUPTIBLE, 0,
820
				   NULL, _RET_IP_, NULL, 0);
821 822 823 824 825 826
}

static noinline int __sched
__ww_mutex_lock_slowpath(struct ww_mutex *lock, struct ww_acquire_ctx *ctx)
{
	return __mutex_lock_common(&lock->base, TASK_UNINTERRUPTIBLE, 0,
827
				   NULL, _RET_IP_, ctx, 1);
I
Ingo Molnar 已提交
828
}
829 830 831 832 833 834

static noinline int __sched
__ww_mutex_lock_interruptible_slowpath(struct ww_mutex *lock,
					    struct ww_acquire_ctx *ctx)
{
	return __mutex_lock_common(&lock->base, TASK_INTERRUPTIBLE, 0,
835
				   NULL, _RET_IP_, ctx, 1);
836 837
}

P
Peter Zijlstra 已提交
838
#endif
I
Ingo Molnar 已提交
839

840 841
/**
 * mutex_trylock - try to acquire the mutex, without waiting
I
Ingo Molnar 已提交
842 843 844 845 846 847
 * @lock: the mutex to be acquired
 *
 * Try to acquire the mutex atomically. Returns 1 if the mutex
 * has been acquired successfully, and 0 on contention.
 *
 * NOTE: this function follows the spin_trylock() convention, so
848
 * it is negated from the down_trylock() return values! Be careful
I
Ingo Molnar 已提交
849 850 851 852 853
 * about this when converting semaphore users to mutexes.
 *
 * This function must not be used in interrupt context. The
 * mutex must be released by the same task that acquired it.
 */
854
int __sched mutex_trylock(struct mutex *lock)
I
Ingo Molnar 已提交
855
{
856
	bool locked = __mutex_trylock(lock);
857

858 859
	if (locked)
		mutex_acquire(&lock->dep_map, 0, 1, _RET_IP_);
860

861
	return locked;
I
Ingo Molnar 已提交
862 863
}
EXPORT_SYMBOL(mutex_trylock);
864

865 866 867 868 869 870
#ifndef CONFIG_DEBUG_LOCK_ALLOC
int __sched
__ww_mutex_lock(struct ww_mutex *lock, struct ww_acquire_ctx *ctx)
{
	might_sleep();

871
	if (__mutex_trylock_fast(&lock->base)) {
872
		ww_mutex_set_context_fastpath(lock, ctx);
873 874 875 876
		return 0;
	}

	return __ww_mutex_lock_slowpath(lock, ctx);
877 878 879 880 881 882 883 884
}
EXPORT_SYMBOL(__ww_mutex_lock);

int __sched
__ww_mutex_lock_interruptible(struct ww_mutex *lock, struct ww_acquire_ctx *ctx)
{
	might_sleep();

885
	if (__mutex_trylock_fast(&lock->base)) {
886
		ww_mutex_set_context_fastpath(lock, ctx);
887 888 889 890
		return 0;
	}

	return __ww_mutex_lock_interruptible_slowpath(lock, ctx);
891 892 893 894 895
}
EXPORT_SYMBOL(__ww_mutex_lock_interruptible);

#endif

896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918
/**
 * atomic_dec_and_mutex_lock - return holding mutex if we dec to 0
 * @cnt: the atomic which we are to dec
 * @lock: the mutex to return holding if we dec to 0
 *
 * return true and hold lock if we dec to 0, return false otherwise
 */
int atomic_dec_and_mutex_lock(atomic_t *cnt, struct mutex *lock)
{
	/* dec if we can't possibly hit 0 */
	if (atomic_add_unless(cnt, -1, 1))
		return 0;
	/* we might hit 0, so take the lock */
	mutex_lock(lock);
	if (!atomic_dec_and_test(cnt)) {
		/* when we actually did the dec, we didn't hit 0 */
		mutex_unlock(lock);
		return 0;
	}
	/* we hit 0, and we hold the lock */
	return 1;
}
EXPORT_SYMBOL(atomic_dec_and_mutex_lock);