rtmutex.c 26.9 KB
Newer Older
I
Ingo Molnar 已提交
1 2 3 4 5 6 7 8 9
/*
 * RT-Mutexes: simple blocking mutual exclusion locks with PI support
 *
 * started by Ingo Molnar and Thomas Gleixner.
 *
 *  Copyright (C) 2004-2006 Red Hat, Inc., Ingo Molnar <mingo@redhat.com>
 *  Copyright (C) 2005-2006 Timesys Corp., Thomas Gleixner <tglx@timesys.com>
 *  Copyright (C) 2005 Kihon Technologies Inc., Steven Rostedt
 *  Copyright (C) 2006 Esben Nielsen
10 11
 *
 *  See Documentation/rt-mutex-design.txt for details.
I
Ingo Molnar 已提交
12 13
 */
#include <linux/spinlock.h>
14
#include <linux/export.h>
I
Ingo Molnar 已提交
15
#include <linux/sched.h>
16
#include <linux/sched/rt.h>
I
Ingo Molnar 已提交
17 18 19 20 21 22 23
#include <linux/timer.h>

#include "rtmutex_common.h"

/*
 * lock->owner state tracking:
 *
24 25
 * lock->owner holds the task_struct pointer of the owner. Bit 0
 * is used to keep track of the "lock has waiters" state.
I
Ingo Molnar 已提交
26
 *
27 28 29 30 31 32
 * owner	bit0
 * NULL		0	lock is free (fast acquire possible)
 * NULL		1	lock is free and has waiters and the top waiter
 *				is going to take the lock*
 * taskpointer	0	lock is held (fast release possible)
 * taskpointer	1	lock is held and has waiters**
I
Ingo Molnar 已提交
33 34
 *
 * The fast atomic compare exchange based acquire and release is only
35 36 37 38 39 40
 * possible when bit 0 of lock->owner is 0.
 *
 * (*) It also can be a transitional state when grabbing the lock
 * with ->wait_lock is held. To prevent any fast path cmpxchg to the lock,
 * we need to set the bit0 before looking at the lock, and the owner may be
 * NULL in this small time, hence this can be a transitional state.
I
Ingo Molnar 已提交
41
 *
42 43 44 45
 * (**) There is a small time when bit 0 is set but there are no
 * waiters. This can happen when grabbing the lock in the slow path.
 * To prevent a cmpxchg of the owner releasing the lock, we need to
 * set this bit before looking at the lock.
I
Ingo Molnar 已提交
46 47
 */

48
static void
49
rt_mutex_set_owner(struct rt_mutex *lock, struct task_struct *owner)
I
Ingo Molnar 已提交
50
{
51
	unsigned long val = (unsigned long)owner;
I
Ingo Molnar 已提交
52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70

	if (rt_mutex_has_waiters(lock))
		val |= RT_MUTEX_HAS_WAITERS;

	lock->owner = (struct task_struct *)val;
}

static inline void clear_rt_mutex_waiters(struct rt_mutex *lock)
{
	lock->owner = (struct task_struct *)
			((unsigned long)lock->owner & ~RT_MUTEX_HAS_WAITERS);
}

static void fixup_rt_mutex_waiters(struct rt_mutex *lock)
{
	if (!rt_mutex_has_waiters(lock))
		clear_rt_mutex_waiters(lock);
}

71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93
/*
 * We can speed up the acquire/release, if the architecture
 * supports cmpxchg and if there's no debugging state to be set up
 */
#if defined(__HAVE_ARCH_CMPXCHG) && !defined(CONFIG_DEBUG_RT_MUTEXES)
# define rt_mutex_cmpxchg(l,c,n)	(cmpxchg(&l->owner, c, n) == c)
static inline void mark_rt_mutex_waiters(struct rt_mutex *lock)
{
	unsigned long owner, *p = (unsigned long *) &lock->owner;

	do {
		owner = *p;
	} while (cmpxchg(p, owner, owner | RT_MUTEX_HAS_WAITERS) != owner);
}
#else
# define rt_mutex_cmpxchg(l,c,n)	(0)
static inline void mark_rt_mutex_waiters(struct rt_mutex *lock)
{
	lock->owner = (struct task_struct *)
			((unsigned long)lock->owner | RT_MUTEX_HAS_WAITERS);
}
#endif

I
Ingo Molnar 已提交
94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113
/*
 * Calculate task priority from the waiter list priority
 *
 * Return task->normal_prio when the waiter list is empty or when
 * the waiter is not allowed to do priority boosting
 */
int rt_mutex_getprio(struct task_struct *task)
{
	if (likely(!task_has_pi_waiters(task)))
		return task->normal_prio;

	return min(task_top_pi_waiter(task)->pi_list_entry.prio,
		   task->normal_prio);
}

/*
 * Adjust the priority of a task, after its pi_waiters got modified.
 *
 * This can be both boosting and unboosting. task->pi_lock must be held.
 */
114
static void __rt_mutex_adjust_prio(struct task_struct *task)
I
Ingo Molnar 已提交
115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134
{
	int prio = rt_mutex_getprio(task);

	if (task->prio != prio)
		rt_mutex_setprio(task, prio);
}

/*
 * Adjust task priority (undo boosting). Called from the exit path of
 * rt_mutex_slowunlock() and rt_mutex_slowlock().
 *
 * (Note: We do this outside of the protection of lock->wait_lock to
 * allow the lock to be taken while or before we readjust the priority
 * of task. We do not use the spin_xx_mutex() variants here as we are
 * outside of the debug path.)
 */
static void rt_mutex_adjust_prio(struct task_struct *task)
{
	unsigned long flags;

135
	raw_spin_lock_irqsave(&task->pi_lock, flags);
I
Ingo Molnar 已提交
136
	__rt_mutex_adjust_prio(task);
137
	raw_spin_unlock_irqrestore(&task->pi_lock, flags);
I
Ingo Molnar 已提交
138 139 140 141 142 143 144 145 146 147
}

/*
 * Max number of times we'll walk the boosting chain:
 */
int max_lock_depth = 1024;

/*
 * Adjust the priority chain. Also used for deadlock detection.
 * Decreases task's usage by one - may thus free the task.
148 149 150 151 152 153 154 155 156 157 158 159 160
 *
 * @task: the task owning the mutex (owner) for which a chain walk is probably
 *	  needed
 * @deadlock_detect: do we have to carry out deadlock detection?
 * @orig_lock: the mutex (can be NULL if we are walking the chain to recheck
 * 	       things for a task that has just got its priority adjusted, and
 *	       is waiting on a mutex)
 * @orig_waiter: rt_mutex_waiter struct for the task that has just donated
 *		 its priority to the mutex owner (can be NULL in the case
 *		 depicted above or if the top waiter is gone away and we are
 *		 actually deboosting the owner)
 * @top_task: the current top waiter
 *
I
Ingo Molnar 已提交
161 162
 * Returns 0 or -EDEADLK.
 */
163 164 165 166 167
static int rt_mutex_adjust_prio_chain(struct task_struct *task,
				      int deadlock_detect,
				      struct rt_mutex *orig_lock,
				      struct rt_mutex_waiter *orig_waiter,
				      struct task_struct *top_task)
I
Ingo Molnar 已提交
168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194
{
	struct rt_mutex *lock;
	struct rt_mutex_waiter *waiter, *top_waiter = orig_waiter;
	int detect_deadlock, ret = 0, depth = 0;
	unsigned long flags;

	detect_deadlock = debug_rt_mutex_detect_deadlock(orig_waiter,
							 deadlock_detect);

	/*
	 * The (de)boosting is a step by step approach with a lot of
	 * pitfalls. We want this to be preemptible and we want hold a
	 * maximum of two locks per step. So we have to check
	 * carefully whether things change under us.
	 */
 again:
	if (++depth > max_lock_depth) {
		static int prev_max;

		/*
		 * Print this only once. If the admin changes the limit,
		 * print a new message when reaching the limit again.
		 */
		if (prev_max != max_lock_depth) {
			prev_max = max_lock_depth;
			printk(KERN_WARNING "Maximum lock depth %d reached "
			       "task: %s (%d)\n", max_lock_depth,
195
			       top_task->comm, task_pid_nr(top_task));
I
Ingo Molnar 已提交
196 197 198 199 200 201 202 203 204
		}
		put_task_struct(task);

		return deadlock_detect ? -EDEADLK : 0;
	}
 retry:
	/*
	 * Task can not go away as we did a get_task() before !
	 */
205
	raw_spin_lock_irqsave(&task->pi_lock, flags);
I
Ingo Molnar 已提交
206 207 208 209 210 211 212

	waiter = task->pi_blocked_on;
	/*
	 * Check whether the end of the boosting chain has been
	 * reached or the state of the chain has changed while we
	 * dropped the locks.
	 */
213
	if (!waiter)
I
Ingo Molnar 已提交
214 215
		goto out_unlock_pi;

216 217
	/*
	 * Check the orig_waiter state. After we dropped the locks,
218
	 * the previous owner of the lock might have released the lock.
219
	 */
220
	if (orig_waiter && !rt_mutex_owner(orig_lock))
221 222 223 224 225 226 227
		goto out_unlock_pi;

	/*
	 * Drop out, when the task has no waiters. Note,
	 * top_waiter can be NULL, when we are in the deboosting
	 * mode!
	 */
I
Ingo Molnar 已提交
228 229 230 231 232 233 234 235 236 237 238 239
	if (top_waiter && (!task_has_pi_waiters(task) ||
			   top_waiter != task_top_pi_waiter(task)))
		goto out_unlock_pi;

	/*
	 * When deadlock detection is off then we check, if further
	 * priority adjustment is necessary.
	 */
	if (!detect_deadlock && waiter->list_entry.prio == task->prio)
		goto out_unlock_pi;

	lock = waiter->lock;
240
	if (!raw_spin_trylock(&lock->wait_lock)) {
241
		raw_spin_unlock_irqrestore(&task->pi_lock, flags);
I
Ingo Molnar 已提交
242 243 244 245 246
		cpu_relax();
		goto retry;
	}

	/* Deadlock detection */
247
	if (lock == orig_lock || rt_mutex_owner(lock) == top_task) {
I
Ingo Molnar 已提交
248
		debug_rt_mutex_deadlock(deadlock_detect, orig_waiter, lock);
249
		raw_spin_unlock(&lock->wait_lock);
I
Ingo Molnar 已提交
250 251 252 253 254 255 256 257 258 259 260 261
		ret = deadlock_detect ? -EDEADLK : 0;
		goto out_unlock_pi;
	}

	top_waiter = rt_mutex_top_waiter(lock);

	/* Requeue the waiter */
	plist_del(&waiter->list_entry, &lock->wait_list);
	waiter->list_entry.prio = task->prio;
	plist_add(&waiter->list_entry, &lock->wait_list);

	/* Release the task */
262
	raw_spin_unlock_irqrestore(&task->pi_lock, flags);
263 264 265 266 267 268 269 270 271 272 273
	if (!rt_mutex_owner(lock)) {
		/*
		 * If the requeue above changed the top waiter, then we need
		 * to wake the new top waiter up to try to get the lock.
		 */

		if (top_waiter != rt_mutex_top_waiter(lock))
			wake_up_process(rt_mutex_top_waiter(lock)->task);
		raw_spin_unlock(&lock->wait_lock);
		goto out_put_task;
	}
I
Ingo Molnar 已提交
274 275 276 277
	put_task_struct(task);

	/* Grab the next task */
	task = rt_mutex_owner(lock);
278
	get_task_struct(task);
279
	raw_spin_lock_irqsave(&task->pi_lock, flags);
I
Ingo Molnar 已提交
280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296

	if (waiter == rt_mutex_top_waiter(lock)) {
		/* Boost the owner */
		plist_del(&top_waiter->pi_list_entry, &task->pi_waiters);
		waiter->pi_list_entry.prio = waiter->list_entry.prio;
		plist_add(&waiter->pi_list_entry, &task->pi_waiters);
		__rt_mutex_adjust_prio(task);

	} else if (top_waiter == waiter) {
		/* Deboost the owner */
		plist_del(&waiter->pi_list_entry, &task->pi_waiters);
		waiter = rt_mutex_top_waiter(lock);
		waiter->pi_list_entry.prio = waiter->list_entry.prio;
		plist_add(&waiter->pi_list_entry, &task->pi_waiters);
		__rt_mutex_adjust_prio(task);
	}

297
	raw_spin_unlock_irqrestore(&task->pi_lock, flags);
I
Ingo Molnar 已提交
298 299

	top_waiter = rt_mutex_top_waiter(lock);
300
	raw_spin_unlock(&lock->wait_lock);
I
Ingo Molnar 已提交
301 302 303 304 305 306 307

	if (!detect_deadlock && waiter != top_waiter)
		goto out_put_task;

	goto again;

 out_unlock_pi:
308
	raw_spin_unlock_irqrestore(&task->pi_lock, flags);
I
Ingo Molnar 已提交
309 310
 out_put_task:
	put_task_struct(task);
311

I
Ingo Molnar 已提交
312 313 314 315 316 317 318
	return ret;
}

/*
 * Try to take an rt-mutex
 *
 * Must be called with lock->wait_lock held.
319 320 321 322
 *
 * @lock:   the lock to be acquired.
 * @task:   the task which wants to acquire the lock
 * @waiter: the waiter that is queued to the lock's wait list. (could be NULL)
I
Ingo Molnar 已提交
323
 */
324 325
static int try_to_take_rt_mutex(struct rt_mutex *lock, struct task_struct *task,
		struct rt_mutex_waiter *waiter)
I
Ingo Molnar 已提交
326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347
{
	/*
	 * We have to be careful here if the atomic speedups are
	 * enabled, such that, when
	 *  - no other waiter is on the lock
	 *  - the lock has been released since we did the cmpxchg
	 * the lock can be released or taken while we are doing the
	 * checks and marking the lock with RT_MUTEX_HAS_WAITERS.
	 *
	 * The atomic acquire/release aware variant of
	 * mark_rt_mutex_waiters uses a cmpxchg loop. After setting
	 * the WAITERS bit, the atomic release / acquire can not
	 * happen anymore and lock->wait_lock protects us from the
	 * non-atomic case.
	 *
	 * Note, that this might set lock->owner =
	 * RT_MUTEX_HAS_WAITERS in the case the lock is not contended
	 * any more. This is fixed up when we take the ownership.
	 * This is the transitional state explained at the top of this file.
	 */
	mark_rt_mutex_waiters(lock);

348
	if (rt_mutex_owner(lock))
I
Ingo Molnar 已提交
349 350
		return 0;

351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387
	/*
	 * It will get the lock because of one of these conditions:
	 * 1) there is no waiter
	 * 2) higher priority than waiters
	 * 3) it is top waiter
	 */
	if (rt_mutex_has_waiters(lock)) {
		if (task->prio >= rt_mutex_top_waiter(lock)->list_entry.prio) {
			if (!waiter || waiter != rt_mutex_top_waiter(lock))
				return 0;
		}
	}

	if (waiter || rt_mutex_has_waiters(lock)) {
		unsigned long flags;
		struct rt_mutex_waiter *top;

		raw_spin_lock_irqsave(&task->pi_lock, flags);

		/* remove the queued waiter. */
		if (waiter) {
			plist_del(&waiter->list_entry, &lock->wait_list);
			task->pi_blocked_on = NULL;
		}

		/*
		 * We have to enqueue the top waiter(if it exists) into
		 * task->pi_waiters list.
		 */
		if (rt_mutex_has_waiters(lock)) {
			top = rt_mutex_top_waiter(lock);
			top->pi_list_entry.prio = top->list_entry.prio;
			plist_add(&top->pi_list_entry, &task->pi_waiters);
		}
		raw_spin_unlock_irqrestore(&task->pi_lock, flags);
	}

I
Ingo Molnar 已提交
388
	/* We got the lock. */
389
	debug_rt_mutex_lock(lock);
I
Ingo Molnar 已提交
390

391
	rt_mutex_set_owner(lock, task);
I
Ingo Molnar 已提交
392

393
	rt_mutex_deadlock_account_lock(lock, task);
I
Ingo Molnar 已提交
394 395 396 397 398 399 400 401 402 403 404 405 406

	return 1;
}

/*
 * Task blocks on lock.
 *
 * Prepare waiter and propagate pi chain
 *
 * This must be called with lock->wait_lock held.
 */
static int task_blocks_on_rt_mutex(struct rt_mutex *lock,
				   struct rt_mutex_waiter *waiter,
D
Darren Hart 已提交
407
				   struct task_struct *task,
408
				   int detect_deadlock)
I
Ingo Molnar 已提交
409
{
410
	struct task_struct *owner = rt_mutex_owner(lock);
I
Ingo Molnar 已提交
411 412
	struct rt_mutex_waiter *top_waiter = waiter;
	unsigned long flags;
413
	int chain_walk = 0, res;
I
Ingo Molnar 已提交
414

415
	raw_spin_lock_irqsave(&task->pi_lock, flags);
D
Darren Hart 已提交
416 417
	__rt_mutex_adjust_prio(task);
	waiter->task = task;
I
Ingo Molnar 已提交
418
	waiter->lock = lock;
D
Darren Hart 已提交
419 420
	plist_node_init(&waiter->list_entry, task->prio);
	plist_node_init(&waiter->pi_list_entry, task->prio);
I
Ingo Molnar 已提交
421 422 423 424 425 426

	/* Get the top priority waiter on the lock */
	if (rt_mutex_has_waiters(lock))
		top_waiter = rt_mutex_top_waiter(lock);
	plist_add(&waiter->list_entry, &lock->wait_list);

D
Darren Hart 已提交
427
	task->pi_blocked_on = waiter;
I
Ingo Molnar 已提交
428

429
	raw_spin_unlock_irqrestore(&task->pi_lock, flags);
I
Ingo Molnar 已提交
430

431 432 433
	if (!owner)
		return 0;

I
Ingo Molnar 已提交
434
	if (waiter == rt_mutex_top_waiter(lock)) {
435
		raw_spin_lock_irqsave(&owner->pi_lock, flags);
I
Ingo Molnar 已提交
436 437 438 439
		plist_del(&top_waiter->pi_list_entry, &owner->pi_waiters);
		plist_add(&waiter->pi_list_entry, &owner->pi_waiters);

		__rt_mutex_adjust_prio(owner);
440 441
		if (owner->pi_blocked_on)
			chain_walk = 1;
442
		raw_spin_unlock_irqrestore(&owner->pi_lock, flags);
I
Ingo Molnar 已提交
443
	}
444 445 446 447
	else if (debug_rt_mutex_detect_deadlock(waiter, detect_deadlock))
		chain_walk = 1;

	if (!chain_walk)
I
Ingo Molnar 已提交
448 449
		return 0;

450 451 452 453 454 455 456
	/*
	 * The owner can't disappear while holding a lock,
	 * so the owner struct is protected by wait_lock.
	 * Gets dropped in rt_mutex_adjust_prio_chain()!
	 */
	get_task_struct(owner);

457
	raw_spin_unlock(&lock->wait_lock);
I
Ingo Molnar 已提交
458

459
	res = rt_mutex_adjust_prio_chain(owner, detect_deadlock, lock, waiter,
D
Darren Hart 已提交
460
					 task);
I
Ingo Molnar 已提交
461

462
	raw_spin_lock(&lock->wait_lock);
I
Ingo Molnar 已提交
463 464 465 466 467 468 469

	return res;
}

/*
 * Wake up the next waiter on the lock.
 *
470
 * Remove the top waiter from the current tasks waiter list and wake it up.
I
Ingo Molnar 已提交
471 472 473 474 475 476 477 478
 *
 * Called with lock->wait_lock held.
 */
static void wakeup_next_waiter(struct rt_mutex *lock)
{
	struct rt_mutex_waiter *waiter;
	unsigned long flags;

479
	raw_spin_lock_irqsave(&current->pi_lock, flags);
I
Ingo Molnar 已提交
480 481 482 483 484 485 486 487 488 489 490

	waiter = rt_mutex_top_waiter(lock);

	/*
	 * Remove it from current->pi_waiters. We do not adjust a
	 * possible priority boost right now. We execute wakeup in the
	 * boosted mode and go back to normal after releasing
	 * lock->wait_lock.
	 */
	plist_del(&waiter->pi_list_entry, &current->pi_waiters);

491
	rt_mutex_set_owner(lock, NULL);
I
Ingo Molnar 已提交
492

493
	raw_spin_unlock_irqrestore(&current->pi_lock, flags);
I
Ingo Molnar 已提交
494

495
	wake_up_process(waiter->task);
I
Ingo Molnar 已提交
496 497 498
}

/*
499
 * Remove a waiter from a lock and give up
I
Ingo Molnar 已提交
500
 *
501 502
 * Must be called with lock->wait_lock held and
 * have just failed to try_to_take_rt_mutex().
I
Ingo Molnar 已提交
503
 */
504 505
static void remove_waiter(struct rt_mutex *lock,
			  struct rt_mutex_waiter *waiter)
I
Ingo Molnar 已提交
506 507
{
	int first = (waiter == rt_mutex_top_waiter(lock));
508
	struct task_struct *owner = rt_mutex_owner(lock);
I
Ingo Molnar 已提交
509
	unsigned long flags;
510
	int chain_walk = 0;
I
Ingo Molnar 已提交
511

512
	raw_spin_lock_irqsave(&current->pi_lock, flags);
I
Ingo Molnar 已提交
513 514
	plist_del(&waiter->list_entry, &lock->wait_list);
	current->pi_blocked_on = NULL;
515
	raw_spin_unlock_irqrestore(&current->pi_lock, flags);
I
Ingo Molnar 已提交
516

517 518 519 520
	if (!owner)
		return;

	if (first) {
I
Ingo Molnar 已提交
521

522
		raw_spin_lock_irqsave(&owner->pi_lock, flags);
I
Ingo Molnar 已提交
523 524 525 526 527 528 529 530 531 532 533

		plist_del(&waiter->pi_list_entry, &owner->pi_waiters);

		if (rt_mutex_has_waiters(lock)) {
			struct rt_mutex_waiter *next;

			next = rt_mutex_top_waiter(lock);
			plist_add(&next->pi_list_entry, &owner->pi_waiters);
		}
		__rt_mutex_adjust_prio(owner);

534 535 536
		if (owner->pi_blocked_on)
			chain_walk = 1;

537
		raw_spin_unlock_irqrestore(&owner->pi_lock, flags);
I
Ingo Molnar 已提交
538 539 540 541
	}

	WARN_ON(!plist_node_empty(&waiter->pi_list_entry));

542
	if (!chain_walk)
I
Ingo Molnar 已提交
543 544
		return;

545 546 547
	/* gets dropped in rt_mutex_adjust_prio_chain()! */
	get_task_struct(owner);

548
	raw_spin_unlock(&lock->wait_lock);
I
Ingo Molnar 已提交
549

550
	rt_mutex_adjust_prio_chain(owner, 0, lock, NULL, current);
I
Ingo Molnar 已提交
551

552
	raw_spin_lock(&lock->wait_lock);
I
Ingo Molnar 已提交
553 554
}

555 556 557 558 559 560 561 562 563 564
/*
 * Recheck the pi chain, in case we got a priority setting
 *
 * Called from sched_setscheduler
 */
void rt_mutex_adjust_pi(struct task_struct *task)
{
	struct rt_mutex_waiter *waiter;
	unsigned long flags;

565
	raw_spin_lock_irqsave(&task->pi_lock, flags);
566 567 568

	waiter = task->pi_blocked_on;
	if (!waiter || waiter->list_entry.prio == task->prio) {
569
		raw_spin_unlock_irqrestore(&task->pi_lock, flags);
570 571 572
		return;
	}

573
	raw_spin_unlock_irqrestore(&task->pi_lock, flags);
574

575 576
	/* gets dropped in rt_mutex_adjust_prio_chain()! */
	get_task_struct(task);
577
	rt_mutex_adjust_prio_chain(task, 0, NULL, NULL, task);
578 579
}

D
Darren Hart 已提交
580 581 582 583 584 585 586 587 588
/**
 * __rt_mutex_slowlock() - Perform the wait-wake-try-to-take loop
 * @lock:		 the rt_mutex to take
 * @state:		 the state the task should block in (TASK_INTERRUPTIBLE
 * 			 or TASK_UNINTERRUPTIBLE)
 * @timeout:		 the pre-initialized and started timer, or NULL for none
 * @waiter:		 the pre-initialized rt_mutex_waiter
 *
 * lock->wait_lock must be held by the caller.
I
Ingo Molnar 已提交
589 590
 */
static int __sched
D
Darren Hart 已提交
591 592
__rt_mutex_slowlock(struct rt_mutex *lock, int state,
		    struct hrtimer_sleeper *timeout,
593
		    struct rt_mutex_waiter *waiter)
I
Ingo Molnar 已提交
594 595 596 597 598
{
	int ret = 0;

	for (;;) {
		/* Try to acquire the lock: */
599
		if (try_to_take_rt_mutex(lock, current, waiter))
I
Ingo Molnar 已提交
600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615
			break;

		/*
		 * TASK_INTERRUPTIBLE checks for signals and
		 * timeout. Ignored otherwise.
		 */
		if (unlikely(state == TASK_INTERRUPTIBLE)) {
			/* Signal pending? */
			if (signal_pending(current))
				ret = -EINTR;
			if (timeout && !timeout->task)
				ret = -ETIMEDOUT;
			if (ret)
				break;
		}

616
		raw_spin_unlock(&lock->wait_lock);
I
Ingo Molnar 已提交
617

D
Darren Hart 已提交
618
		debug_rt_mutex_print_deadlock(waiter);
I
Ingo Molnar 已提交
619

620
		schedule_rt_mutex(lock);
I
Ingo Molnar 已提交
621

622
		raw_spin_lock(&lock->wait_lock);
I
Ingo Molnar 已提交
623 624 625
		set_current_state(state);
	}

D
Darren Hart 已提交
626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641
	return ret;
}

/*
 * Slow path lock function:
 */
static int __sched
rt_mutex_slowlock(struct rt_mutex *lock, int state,
		  struct hrtimer_sleeper *timeout,
		  int detect_deadlock)
{
	struct rt_mutex_waiter waiter;
	int ret = 0;

	debug_rt_mutex_init_waiter(&waiter);

642
	raw_spin_lock(&lock->wait_lock);
D
Darren Hart 已提交
643 644

	/* Try to acquire the lock again: */
645
	if (try_to_take_rt_mutex(lock, current, NULL)) {
646
		raw_spin_unlock(&lock->wait_lock);
D
Darren Hart 已提交
647 648 649 650 651 652 653 654 655 656 657 658
		return 0;
	}

	set_current_state(state);

	/* Setup the timer, when timeout != NULL */
	if (unlikely(timeout)) {
		hrtimer_start_expires(&timeout->timer, HRTIMER_MODE_ABS);
		if (!hrtimer_active(&timeout->timer))
			timeout->task = NULL;
	}

659 660 661 662
	ret = task_blocks_on_rt_mutex(lock, &waiter, current, detect_deadlock);

	if (likely(!ret))
		ret = __rt_mutex_slowlock(lock, state, timeout, &waiter);
D
Darren Hart 已提交
663

I
Ingo Molnar 已提交
664 665
	set_current_state(TASK_RUNNING);

666
	if (unlikely(ret))
667
		remove_waiter(lock, &waiter);
I
Ingo Molnar 已提交
668 669 670 671 672 673 674

	/*
	 * try_to_take_rt_mutex() sets the waiter bit
	 * unconditionally. We might have to fix that up.
	 */
	fixup_rt_mutex_waiters(lock);

675
	raw_spin_unlock(&lock->wait_lock);
I
Ingo Molnar 已提交
676 677 678 679 680 681 682 683 684 685 686 687 688 689

	/* Remove pending timer: */
	if (unlikely(timeout))
		hrtimer_cancel(&timeout->timer);

	debug_rt_mutex_free_waiter(&waiter);

	return ret;
}

/*
 * Slow path try-lock function:
 */
static inline int
690
rt_mutex_slowtrylock(struct rt_mutex *lock)
I
Ingo Molnar 已提交
691 692 693
{
	int ret = 0;

694
	raw_spin_lock(&lock->wait_lock);
I
Ingo Molnar 已提交
695 696 697

	if (likely(rt_mutex_owner(lock) != current)) {

698
		ret = try_to_take_rt_mutex(lock, current, NULL);
I
Ingo Molnar 已提交
699 700 701 702 703 704 705
		/*
		 * try_to_take_rt_mutex() sets the lock waiters
		 * bit unconditionally. Clean this up.
		 */
		fixup_rt_mutex_waiters(lock);
	}

706
	raw_spin_unlock(&lock->wait_lock);
I
Ingo Molnar 已提交
707 708 709 710 711 712 713 714 715 716

	return ret;
}

/*
 * Slow path to release a rt-mutex:
 */
static void __sched
rt_mutex_slowunlock(struct rt_mutex *lock)
{
717
	raw_spin_lock(&lock->wait_lock);
I
Ingo Molnar 已提交
718 719 720 721 722 723 724

	debug_rt_mutex_unlock(lock);

	rt_mutex_deadlock_account_unlock(current);

	if (!rt_mutex_has_waiters(lock)) {
		lock->owner = NULL;
725
		raw_spin_unlock(&lock->wait_lock);
I
Ingo Molnar 已提交
726 727 728 729 730
		return;
	}

	wakeup_next_waiter(lock);

731
	raw_spin_unlock(&lock->wait_lock);
I
Ingo Molnar 已提交
732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747

	/* Undo pi boosting if necessary: */
	rt_mutex_adjust_prio(current);
}

/*
 * debug aware fast / slowpath lock,trylock,unlock
 *
 * The atomic acquire/release ops are compiled away, when either the
 * architecture does not support cmpxchg or when debugging is enabled.
 */
static inline int
rt_mutex_fastlock(struct rt_mutex *lock, int state,
		  int detect_deadlock,
		  int (*slowfn)(struct rt_mutex *lock, int state,
				struct hrtimer_sleeper *timeout,
748
				int detect_deadlock))
I
Ingo Molnar 已提交
749 750 751 752 753
{
	if (!detect_deadlock && likely(rt_mutex_cmpxchg(lock, NULL, current))) {
		rt_mutex_deadlock_account_lock(lock, current);
		return 0;
	} else
754
		return slowfn(lock, state, NULL, detect_deadlock);
I
Ingo Molnar 已提交
755 756 757 758 759 760 761
}

static inline int
rt_mutex_timed_fastlock(struct rt_mutex *lock, int state,
			struct hrtimer_sleeper *timeout, int detect_deadlock,
			int (*slowfn)(struct rt_mutex *lock, int state,
				      struct hrtimer_sleeper *timeout,
762
				      int detect_deadlock))
I
Ingo Molnar 已提交
763 764 765 766 767
{
	if (!detect_deadlock && likely(rt_mutex_cmpxchg(lock, NULL, current))) {
		rt_mutex_deadlock_account_lock(lock, current);
		return 0;
	} else
768
		return slowfn(lock, state, timeout, detect_deadlock);
I
Ingo Molnar 已提交
769 770 771 772
}

static inline int
rt_mutex_fasttrylock(struct rt_mutex *lock,
773
		     int (*slowfn)(struct rt_mutex *lock))
I
Ingo Molnar 已提交
774 775 776 777 778
{
	if (likely(rt_mutex_cmpxchg(lock, NULL, current))) {
		rt_mutex_deadlock_account_lock(lock, current);
		return 1;
	}
779
	return slowfn(lock);
I
Ingo Molnar 已提交
780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826
}

static inline void
rt_mutex_fastunlock(struct rt_mutex *lock,
		    void (*slowfn)(struct rt_mutex *lock))
{
	if (likely(rt_mutex_cmpxchg(lock, current, NULL)))
		rt_mutex_deadlock_account_unlock(current);
	else
		slowfn(lock);
}

/**
 * rt_mutex_lock - lock a rt_mutex
 *
 * @lock: the rt_mutex to be locked
 */
void __sched rt_mutex_lock(struct rt_mutex *lock)
{
	might_sleep();

	rt_mutex_fastlock(lock, TASK_UNINTERRUPTIBLE, 0, rt_mutex_slowlock);
}
EXPORT_SYMBOL_GPL(rt_mutex_lock);

/**
 * rt_mutex_lock_interruptible - lock a rt_mutex interruptible
 *
 * @lock: 		the rt_mutex to be locked
 * @detect_deadlock:	deadlock detection on/off
 *
 * Returns:
 *  0 		on success
 * -EINTR 	when interrupted by a signal
 * -EDEADLK	when the lock would deadlock (when deadlock detection is on)
 */
int __sched rt_mutex_lock_interruptible(struct rt_mutex *lock,
						 int detect_deadlock)
{
	might_sleep();

	return rt_mutex_fastlock(lock, TASK_INTERRUPTIBLE,
				 detect_deadlock, rt_mutex_slowlock);
}
EXPORT_SYMBOL_GPL(rt_mutex_lock_interruptible);

/**
827 828 829
 * rt_mutex_timed_lock - lock a rt_mutex interruptible
 *			the timeout structure is provided
 *			by the caller
I
Ingo Molnar 已提交
830 831 832 833 834 835 836 837
 *
 * @lock: 		the rt_mutex to be locked
 * @timeout:		timeout structure or NULL (no timeout)
 * @detect_deadlock:	deadlock detection on/off
 *
 * Returns:
 *  0 		on success
 * -EINTR 	when interrupted by a signal
838
 * -ETIMEDOUT	when the timeout expired
I
Ingo Molnar 已提交
839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875
 * -EDEADLK	when the lock would deadlock (when deadlock detection is on)
 */
int
rt_mutex_timed_lock(struct rt_mutex *lock, struct hrtimer_sleeper *timeout,
		    int detect_deadlock)
{
	might_sleep();

	return rt_mutex_timed_fastlock(lock, TASK_INTERRUPTIBLE, timeout,
				       detect_deadlock, rt_mutex_slowlock);
}
EXPORT_SYMBOL_GPL(rt_mutex_timed_lock);

/**
 * rt_mutex_trylock - try to lock a rt_mutex
 *
 * @lock:	the rt_mutex to be locked
 *
 * Returns 1 on success and 0 on contention
 */
int __sched rt_mutex_trylock(struct rt_mutex *lock)
{
	return rt_mutex_fasttrylock(lock, rt_mutex_slowtrylock);
}
EXPORT_SYMBOL_GPL(rt_mutex_trylock);

/**
 * rt_mutex_unlock - unlock a rt_mutex
 *
 * @lock: the rt_mutex to be unlocked
 */
void __sched rt_mutex_unlock(struct rt_mutex *lock)
{
	rt_mutex_fastunlock(lock, rt_mutex_slowunlock);
}
EXPORT_SYMBOL_GPL(rt_mutex_unlock);

876
/**
I
Ingo Molnar 已提交
877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905
 * rt_mutex_destroy - mark a mutex unusable
 * @lock: the mutex to be destroyed
 *
 * This function marks the mutex uninitialized, and any subsequent
 * use of the mutex is forbidden. The mutex must not be locked when
 * this function is called.
 */
void rt_mutex_destroy(struct rt_mutex *lock)
{
	WARN_ON(rt_mutex_is_locked(lock));
#ifdef CONFIG_DEBUG_RT_MUTEXES
	lock->magic = NULL;
#endif
}

EXPORT_SYMBOL_GPL(rt_mutex_destroy);

/**
 * __rt_mutex_init - initialize the rt lock
 *
 * @lock: the rt lock to be initialized
 *
 * Initialize the rt lock to unlocked state.
 *
 * Initializing of a locked rt lock is not allowed
 */
void __rt_mutex_init(struct rt_mutex *lock, const char *name)
{
	lock->owner = NULL;
906
	raw_spin_lock_init(&lock->wait_lock);
907
	plist_head_init(&lock->wait_list);
I
Ingo Molnar 已提交
908 909 910 911

	debug_rt_mutex_init(lock, name);
}
EXPORT_SYMBOL_GPL(__rt_mutex_init);
912 913 914 915 916 917 918 919 920 921 922 923 924 925 926

/**
 * rt_mutex_init_proxy_locked - initialize and lock a rt_mutex on behalf of a
 *				proxy owner
 *
 * @lock: 	the rt_mutex to be locked
 * @proxy_owner:the task to set as owner
 *
 * No locking. Caller has to do serializing itself
 * Special API call for PI-futex support
 */
void rt_mutex_init_proxy_locked(struct rt_mutex *lock,
				struct task_struct *proxy_owner)
{
	__rt_mutex_init(lock, NULL);
927
	debug_rt_mutex_proxy_lock(lock, proxy_owner);
928
	rt_mutex_set_owner(lock, proxy_owner);
929 930 931 932 933 934 935 936 937 938 939 940 941 942 943
	rt_mutex_deadlock_account_lock(lock, proxy_owner);
}

/**
 * rt_mutex_proxy_unlock - release a lock on behalf of owner
 *
 * @lock: 	the rt_mutex to be locked
 *
 * No locking. Caller has to do serializing itself
 * Special API call for PI-futex support
 */
void rt_mutex_proxy_unlock(struct rt_mutex *lock,
			   struct task_struct *proxy_owner)
{
	debug_rt_mutex_proxy_unlock(lock);
944
	rt_mutex_set_owner(lock, NULL);
945 946 947
	rt_mutex_deadlock_account_unlock(proxy_owner);
}

D
Darren Hart 已提交
948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967
/**
 * rt_mutex_start_proxy_lock() - Start lock acquisition for another task
 * @lock:		the rt_mutex to take
 * @waiter:		the pre-initialized rt_mutex_waiter
 * @task:		the task to prepare
 * @detect_deadlock:	perform deadlock detection (1) or not (0)
 *
 * Returns:
 *  0 - task blocked on lock
 *  1 - acquired the lock for task, caller should wake it up
 * <0 - error
 *
 * Special API call for FUTEX_REQUEUE_PI support.
 */
int rt_mutex_start_proxy_lock(struct rt_mutex *lock,
			      struct rt_mutex_waiter *waiter,
			      struct task_struct *task, int detect_deadlock)
{
	int ret;

968
	raw_spin_lock(&lock->wait_lock);
D
Darren Hart 已提交
969

970
	if (try_to_take_rt_mutex(lock, task, NULL)) {
971
		raw_spin_unlock(&lock->wait_lock);
D
Darren Hart 已提交
972 973 974 975 976
		return 1;
	}

	ret = task_blocks_on_rt_mutex(lock, waiter, task, detect_deadlock);

977
	if (ret && !rt_mutex_owner(lock)) {
D
Darren Hart 已提交
978 979 980 981 982 983 984 985
		/*
		 * Reset the return value. We might have
		 * returned with -EDEADLK and the owner
		 * released the lock while we were walking the
		 * pi chain.  Let the waiter sort it out.
		 */
		ret = 0;
	}
986 987 988 989

	if (unlikely(ret))
		remove_waiter(lock, waiter);

990
	raw_spin_unlock(&lock->wait_lock);
D
Darren Hart 已提交
991 992 993 994 995 996

	debug_rt_mutex_print_deadlock(waiter);

	return ret;
}

997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015
/**
 * rt_mutex_next_owner - return the next owner of the lock
 *
 * @lock: the rt lock query
 *
 * Returns the next owner of the lock or NULL
 *
 * Caller has to serialize against other accessors to the lock
 * itself.
 *
 * Special API call for PI-futex support
 */
struct task_struct *rt_mutex_next_owner(struct rt_mutex *lock)
{
	if (!rt_mutex_has_waiters(lock))
		return NULL;

	return rt_mutex_top_waiter(lock)->task;
}
D
Darren Hart 已提交
1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039

/**
 * rt_mutex_finish_proxy_lock() - Complete lock acquisition
 * @lock:		the rt_mutex we were woken on
 * @to:			the timeout, null if none. hrtimer should already have
 * 			been started.
 * @waiter:		the pre-initialized rt_mutex_waiter
 * @detect_deadlock:	perform deadlock detection (1) or not (0)
 *
 * Complete the lock acquisition started our behalf by another thread.
 *
 * Returns:
 *  0 - success
 * <0 - error, one of -EINTR, -ETIMEDOUT, or -EDEADLK
 *
 * Special API call for PI-futex requeue support
 */
int rt_mutex_finish_proxy_lock(struct rt_mutex *lock,
			       struct hrtimer_sleeper *to,
			       struct rt_mutex_waiter *waiter,
			       int detect_deadlock)
{
	int ret;

1040
	raw_spin_lock(&lock->wait_lock);
D
Darren Hart 已提交
1041 1042 1043

	set_current_state(TASK_INTERRUPTIBLE);

1044
	ret = __rt_mutex_slowlock(lock, TASK_INTERRUPTIBLE, to, waiter);
D
Darren Hart 已提交
1045 1046 1047

	set_current_state(TASK_RUNNING);

1048
	if (unlikely(ret))
D
Darren Hart 已提交
1049 1050 1051 1052 1053 1054 1055 1056
		remove_waiter(lock, waiter);

	/*
	 * try_to_take_rt_mutex() sets the waiter bit unconditionally. We might
	 * have to fix that up.
	 */
	fixup_rt_mutex_waiters(lock);

1057
	raw_spin_unlock(&lock->wait_lock);
D
Darren Hart 已提交
1058 1059 1060

	return ret;
}