rcutree_plugin.h 31.9 KB
Newer Older
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
/*
 * Read-Copy Update mechanism for mutual exclusion (tree-based version)
 * Internal non-public definitions that provide either classic
 * or preemptable semantics.
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation; either version 2 of the License, or
 * (at your option) any later version.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, write to the Free Software
 * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
 *
 * Copyright Red Hat, 2009
 * Copyright IBM Corporation, 2009
 *
 * Author: Ingo Molnar <mingo@elte.hu>
 *	   Paul E. McKenney <paulmck@linux.vnet.ibm.com>
 */

27
#include <linux/delay.h>
28

29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67
/*
 * Check the RCU kernel configuration parameters and print informative
 * messages about anything out of the ordinary.  If you like #ifdef, you
 * will love this function.
 */
static void __init rcu_bootup_announce_oddness(void)
{
#ifdef CONFIG_RCU_TRACE
	printk(KERN_INFO "\tRCU debugfs-based tracing is enabled.\n");
#endif
#if (defined(CONFIG_64BIT) && CONFIG_RCU_FANOUT != 64) || (!defined(CONFIG_64BIT) && CONFIG_RCU_FANOUT != 32)
	printk(KERN_INFO "\tCONFIG_RCU_FANOUT set to non-default value of %d\n",
	       CONFIG_RCU_FANOUT);
#endif
#ifdef CONFIG_RCU_FANOUT_EXACT
	printk(KERN_INFO "\tHierarchical RCU autobalancing is disabled.\n");
#endif
#ifdef CONFIG_RCU_FAST_NO_HZ
	printk(KERN_INFO
	       "\tRCU dyntick-idle grace-period acceleration is enabled.\n");
#endif
#ifdef CONFIG_PROVE_RCU
	printk(KERN_INFO "\tRCU lockdep checking is enabled.\n");
#endif
#ifdef CONFIG_RCU_TORTURE_TEST_RUNNABLE
	printk(KERN_INFO "\tRCU torture testing starts during boot.\n");
#endif
#ifndef CONFIG_RCU_CPU_STALL_DETECTOR
	printk(KERN_INFO
	       "\tRCU-based detection of stalled CPUs is disabled.\n");
#endif
#ifndef CONFIG_RCU_CPU_STALL_VERBOSE
	printk(KERN_INFO "\tVerbose stalled-CPUs detection is disabled.\n");
#endif
#if NUM_RCU_LVL_4 != 0
	printk(KERN_INFO "\tExperimental four-level hierarchy is enabled.\n");
#endif
}

68 69 70 71 72
#ifdef CONFIG_TREE_PREEMPT_RCU

struct rcu_state rcu_preempt_state = RCU_STATE_INITIALIZER(rcu_preempt_state);
DEFINE_PER_CPU(struct rcu_data, rcu_preempt_data);

73 74
static int rcu_preempted_readers_exp(struct rcu_node *rnp);

75 76 77
/*
 * Tell them what RCU they are running.
 */
78
static void __init rcu_bootup_announce(void)
79
{
80 81
	printk(KERN_INFO "Preemptable hierarchical RCU implementation.\n");
	rcu_bootup_announce_oddness();
82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102
}

/*
 * Return the number of RCU-preempt batches processed thus far
 * for debug and statistics.
 */
long rcu_batches_completed_preempt(void)
{
	return rcu_preempt_state.completed;
}
EXPORT_SYMBOL_GPL(rcu_batches_completed_preempt);

/*
 * Return the number of RCU batches processed thus far for debug & stats.
 */
long rcu_batches_completed(void)
{
	return rcu_batches_completed_preempt();
}
EXPORT_SYMBOL_GPL(rcu_batches_completed);

103 104 105 106 107 108 109 110 111
/*
 * Force a quiescent state for preemptible RCU.
 */
void rcu_force_quiescent_state(void)
{
	force_quiescent_state(&rcu_preempt_state, 0);
}
EXPORT_SYMBOL_GPL(rcu_force_quiescent_state);

112 113 114 115 116
/*
 * Record a preemptable-RCU quiescent state for the specified CPU.  Note
 * that this just means that the task currently running on the CPU is
 * not in a quiescent state.  There might be any number of tasks blocked
 * while in an RCU read-side critical section.
117 118 119 120
 *
 * Unlike the other rcu_*_qs() functions, callers to this function
 * must disable irqs in order to protect the assignment to
 * ->rcu_read_unlock_special.
121
 */
122
static void rcu_preempt_qs(int cpu)
123 124
{
	struct rcu_data *rdp = &per_cpu(rcu_preempt_data, cpu);
125

126
	rdp->passed_quiesc_completed = rdp->gpnum - 1;
127 128
	barrier();
	rdp->passed_quiesc = 1;
129
	current->rcu_read_unlock_special &= ~RCU_READ_UNLOCK_NEED_QS;
130 131 132
}

/*
133 134 135 136 137 138 139 140 141 142 143
 * We have entered the scheduler, and the current task might soon be
 * context-switched away from.  If this task is in an RCU read-side
 * critical section, we will no longer be able to rely on the CPU to
 * record that fact, so we enqueue the task on the appropriate entry
 * of the blocked_tasks[] array.  The task will dequeue itself when
 * it exits the outermost enclosing RCU read-side critical section.
 * Therefore, the current grace period cannot be permitted to complete
 * until the blocked_tasks[] entry indexed by the low-order bit of
 * rnp->gpnum empties.
 *
 * Caller must disable preemption.
144
 */
145
static void rcu_preempt_note_context_switch(int cpu)
146 147
{
	struct task_struct *t = current;
148
	unsigned long flags;
149 150 151 152 153 154 155 156
	int phase;
	struct rcu_data *rdp;
	struct rcu_node *rnp;

	if (t->rcu_read_lock_nesting &&
	    (t->rcu_read_unlock_special & RCU_READ_UNLOCK_BLOCKED) == 0) {

		/* Possibly blocking in an RCU read-side critical section. */
157
		rdp = per_cpu_ptr(rcu_preempt_state.rda, cpu);
158
		rnp = rdp->mynode;
P
Paul E. McKenney 已提交
159
		raw_spin_lock_irqsave(&rnp->lock, flags);
160
		t->rcu_read_unlock_special |= RCU_READ_UNLOCK_BLOCKED;
161
		t->rcu_blocked_node = rnp;
162 163 164 165 166 167 168 169 170 171

		/*
		 * If this CPU has already checked in, then this task
		 * will hold up the next grace period rather than the
		 * current grace period.  Queue the task accordingly.
		 * If the task is queued for the current grace period
		 * (i.e., this CPU has not yet passed through a quiescent
		 * state for the current grace period), then as long
		 * as that task remains queued, the current grace period
		 * cannot end.
172 173 174
		 *
		 * But first, note that the current CPU must still be
		 * on line!
175
		 */
176
		WARN_ON_ONCE((rdp->grpmask & rnp->qsmaskinit) == 0);
177 178
		WARN_ON_ONCE(!list_empty(&t->rcu_node_entry));
		phase = (rnp->gpnum + !(rnp->qsmask & rdp->grpmask)) & 0x1;
179
		list_add(&t->rcu_node_entry, &rnp->blocked_tasks[phase]);
P
Paul E. McKenney 已提交
180
		raw_spin_unlock_irqrestore(&rnp->lock, flags);
181 182 183 184 185 186 187 188 189 190 191
	}

	/*
	 * Either we were not in an RCU read-side critical section to
	 * begin with, or we have now recorded that critical section
	 * globally.  Either way, we can now note a quiescent state
	 * for this CPU.  Again, if we were in an RCU read-side critical
	 * section, and if that critical section was blocking the current
	 * grace period, then the fact that the task has been enqueued
	 * means that we continue to block the current grace period.
	 */
192
	local_irq_save(flags);
193
	rcu_preempt_qs(cpu);
194
	local_irq_restore(flags);
195 196 197 198 199 200 201 202 203 204 205 206 207 208
}

/*
 * Tree-preemptable RCU implementation for rcu_read_lock().
 * Just increment ->rcu_read_lock_nesting, shared state will be updated
 * if we block.
 */
void __rcu_read_lock(void)
{
	ACCESS_ONCE(current->rcu_read_lock_nesting)++;
	barrier();  /* needed if we ever invoke rcu_read_lock in rcutree.c */
}
EXPORT_SYMBOL_GPL(__rcu_read_lock);

209 210 211 212 213 214 215
/*
 * Check for preempted RCU readers blocking the current grace period
 * for the specified rcu_node structure.  If the caller needs a reliable
 * answer, it must hold the rcu_node's ->lock.
 */
static int rcu_preempted_readers(struct rcu_node *rnp)
{
216 217 218 219
	int phase = rnp->gpnum & 0x1;

	return !list_empty(&rnp->blocked_tasks[phase]) ||
	       !list_empty(&rnp->blocked_tasks[phase + 2]);
220 221
}

222 223 224 225 226 227 228
/*
 * Record a quiescent state for all tasks that were previously queued
 * on the specified rcu_node structure and that were blocking the current
 * RCU grace period.  The caller must hold the specified rnp->lock with
 * irqs disabled, and this lock is released upon return, but irqs remain
 * disabled.
 */
P
Paul E. McKenney 已提交
229
static void rcu_report_unblock_qs_rnp(struct rcu_node *rnp, unsigned long flags)
230 231 232 233 234 235
	__releases(rnp->lock)
{
	unsigned long mask;
	struct rcu_node *rnp_p;

	if (rnp->qsmask != 0 || rcu_preempted_readers(rnp)) {
P
Paul E. McKenney 已提交
236
		raw_spin_unlock_irqrestore(&rnp->lock, flags);
237 238 239 240 241 242 243 244 245 246
		return;  /* Still need more quiescent states! */
	}

	rnp_p = rnp->parent;
	if (rnp_p == NULL) {
		/*
		 * Either there is only one rcu_node in the tree,
		 * or tasks were kicked up to root rcu_node due to
		 * CPUs going offline.
		 */
P
Paul E. McKenney 已提交
247
		rcu_report_qs_rsp(&rcu_preempt_state, flags);
248 249 250 251 252
		return;
	}

	/* Report up the rest of the hierarchy. */
	mask = rnp->grpmask;
P
Paul E. McKenney 已提交
253 254
	raw_spin_unlock(&rnp->lock);	/* irqs remain disabled. */
	raw_spin_lock(&rnp_p->lock);	/* irqs already disabled. */
P
Paul E. McKenney 已提交
255
	rcu_report_qs_rnp(mask, &rcu_preempt_state, rnp_p, flags);
256 257 258 259 260 261 262
}

/*
 * Handle special cases during rcu_read_unlock(), such as needing to
 * notify RCU core processing or task having blocked during the RCU
 * read-side critical section.
 */
263 264 265
static void rcu_read_unlock_special(struct task_struct *t)
{
	int empty;
266
	int empty_exp;
267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282
	unsigned long flags;
	struct rcu_node *rnp;
	int special;

	/* NMI handlers cannot block and cannot safely manipulate state. */
	if (in_nmi())
		return;

	local_irq_save(flags);

	/*
	 * If RCU core is waiting for this CPU to exit critical section,
	 * let it know that we have done so.
	 */
	special = t->rcu_read_unlock_special;
	if (special & RCU_READ_UNLOCK_NEED_QS) {
283
		rcu_preempt_qs(smp_processor_id());
284 285 286 287 288 289 290 291 292 293 294 295
	}

	/* Hardware IRQ handlers cannot block. */
	if (in_irq()) {
		local_irq_restore(flags);
		return;
	}

	/* Clean up if blocked during RCU read-side critical section. */
	if (special & RCU_READ_UNLOCK_BLOCKED) {
		t->rcu_read_unlock_special &= ~RCU_READ_UNLOCK_BLOCKED;

296 297 298 299 300 301
		/*
		 * Remove this task from the list it blocked on.  The
		 * task can migrate while we acquire the lock, but at
		 * most one time.  So at most two passes through loop.
		 */
		for (;;) {
302
			rnp = t->rcu_blocked_node;
P
Paul E. McKenney 已提交
303
			raw_spin_lock(&rnp->lock);  /* irqs already disabled. */
304
			if (rnp == t->rcu_blocked_node)
305
				break;
P
Paul E. McKenney 已提交
306
			raw_spin_unlock(&rnp->lock); /* irqs remain disabled. */
307
		}
308
		empty = !rcu_preempted_readers(rnp);
309 310
		empty_exp = !rcu_preempted_readers_exp(rnp);
		smp_mb(); /* ensure expedited fastpath sees end of RCU c-s. */
311
		list_del_init(&t->rcu_node_entry);
312
		t->rcu_blocked_node = NULL;
313 314 315 316

		/*
		 * If this was the last task on the current list, and if
		 * we aren't waiting on any CPUs, report the quiescent state.
P
Paul E. McKenney 已提交
317
		 * Note that rcu_report_unblock_qs_rnp() releases rnp->lock.
318
		 */
319
		if (empty)
P
Paul E. McKenney 已提交
320
			raw_spin_unlock_irqrestore(&rnp->lock, flags);
321
		else
P
Paul E. McKenney 已提交
322
			rcu_report_unblock_qs_rnp(rnp, flags);
323 324 325 326 327 328 329

		/*
		 * If this was the last task on the expedited lists,
		 * then we need to report up the rcu_node hierarchy.
		 */
		if (!empty_exp && !rcu_preempted_readers_exp(rnp))
			rcu_report_exp_rnp(&rcu_preempt_state, rnp);
330 331
	} else {
		local_irq_restore(flags);
332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349
	}
}

/*
 * Tree-preemptable RCU implementation for rcu_read_unlock().
 * Decrement ->rcu_read_lock_nesting.  If the result is zero (outermost
 * rcu_read_unlock()) and ->rcu_read_unlock_special is non-zero, then
 * invoke rcu_read_unlock_special() to clean up after a context switch
 * in an RCU read-side critical section and other special cases.
 */
void __rcu_read_unlock(void)
{
	struct task_struct *t = current;

	barrier();  /* needed if we ever invoke rcu_read_unlock in rcutree.c */
	if (--ACCESS_ONCE(t->rcu_read_lock_nesting) == 0 &&
	    unlikely(ACCESS_ONCE(t->rcu_read_unlock_special)))
		rcu_read_unlock_special(t);
350 351 352
#ifdef CONFIG_PROVE_LOCKING
	WARN_ON_ONCE(ACCESS_ONCE(t->rcu_read_lock_nesting) < 0);
#endif /* #ifdef CONFIG_PROVE_LOCKING */
353 354 355 356 357
}
EXPORT_SYMBOL_GPL(__rcu_read_unlock);

#ifdef CONFIG_RCU_CPU_STALL_DETECTOR

358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401
#ifdef CONFIG_RCU_CPU_STALL_VERBOSE

/*
 * Dump detailed information for all tasks blocking the current RCU
 * grace period on the specified rcu_node structure.
 */
static void rcu_print_detail_task_stall_rnp(struct rcu_node *rnp)
{
	unsigned long flags;
	struct list_head *lp;
	int phase;
	struct task_struct *t;

	if (rcu_preempted_readers(rnp)) {
		raw_spin_lock_irqsave(&rnp->lock, flags);
		phase = rnp->gpnum & 0x1;
		lp = &rnp->blocked_tasks[phase];
		list_for_each_entry(t, lp, rcu_node_entry)
			sched_show_task(t);
		raw_spin_unlock_irqrestore(&rnp->lock, flags);
	}
}

/*
 * Dump detailed information for all tasks blocking the current RCU
 * grace period.
 */
static void rcu_print_detail_task_stall(struct rcu_state *rsp)
{
	struct rcu_node *rnp = rcu_get_root(rsp);

	rcu_print_detail_task_stall_rnp(rnp);
	rcu_for_each_leaf_node(rsp, rnp)
		rcu_print_detail_task_stall_rnp(rnp);
}

#else /* #ifdef CONFIG_RCU_CPU_STALL_VERBOSE */

static void rcu_print_detail_task_stall(struct rcu_state *rsp)
{
}

#endif /* #else #ifdef CONFIG_RCU_CPU_STALL_VERBOSE */

402 403 404 405 406 407 408
/*
 * Scan the current list of tasks blocked within RCU read-side critical
 * sections, printing out the tid of each.
 */
static void rcu_print_task_stall(struct rcu_node *rnp)
{
	struct list_head *lp;
409
	int phase;
410 411
	struct task_struct *t;

412 413
	if (rcu_preempted_readers(rnp)) {
		phase = rnp->gpnum & 0x1;
414 415 416 417 418 419 420 421
		lp = &rnp->blocked_tasks[phase];
		list_for_each_entry(t, lp, rcu_node_entry)
			printk(" P%d", t->pid);
	}
}

#endif /* #ifdef CONFIG_RCU_CPU_STALL_DETECTOR */

422 423 424 425 426 427 428 429 430
/*
 * Check that the list of blocked tasks for the newly completed grace
 * period is in fact empty.  It is a serious bug to complete a grace
 * period that still has RCU readers blocked!  This function must be
 * invoked -before- updating this rnp's ->gpnum, and the rnp's ->lock
 * must be held by the caller.
 */
static void rcu_preempt_check_blocked_tasks(struct rcu_node *rnp)
{
431
	WARN_ON_ONCE(rcu_preempted_readers(rnp));
432
	WARN_ON_ONCE(rnp->qsmask);
433 434
}

435 436
#ifdef CONFIG_HOTPLUG_CPU

437 438 439 440 441 442
/*
 * Handle tasklist migration for case in which all CPUs covered by the
 * specified rcu_node have gone offline.  Move them up to the root
 * rcu_node.  The reason for not just moving them to the immediate
 * parent is to remove the need for rcu_read_unlock_special() to
 * make more than two attempts to acquire the target rcu_node's lock.
443 444
 * Returns true if there were tasks blocking the current RCU grace
 * period.
445
 *
446 447 448
 * Returns 1 if there was previously a task blocking the current grace
 * period on the specified rcu_node structure.
 *
449 450
 * The caller must hold rnp->lock with irqs disabled.
 */
451 452 453
static int rcu_preempt_offline_tasks(struct rcu_state *rsp,
				     struct rcu_node *rnp,
				     struct rcu_data *rdp)
454 455 456 457
{
	int i;
	struct list_head *lp;
	struct list_head *lp_root;
458
	int retval = 0;
459 460 461
	struct rcu_node *rnp_root = rcu_get_root(rsp);
	struct task_struct *tp;

462 463
	if (rnp == rnp_root) {
		WARN_ONCE(1, "Last CPU thought to be offlined?");
464
		return 0;  /* Shouldn't happen: at least one CPU online. */
465
	}
466 467
	WARN_ON_ONCE(rnp != rdp->mynode &&
		     (!list_empty(&rnp->blocked_tasks[0]) ||
468 469 470
		      !list_empty(&rnp->blocked_tasks[1]) ||
		      !list_empty(&rnp->blocked_tasks[2]) ||
		      !list_empty(&rnp->blocked_tasks[3])));
471 472 473 474 475 476 477

	/*
	 * Move tasks up to root rcu_node.  Rely on the fact that the
	 * root rcu_node can be at most one ahead of the rest of the
	 * rcu_nodes in terms of gp_num value.  This fact allows us to
	 * move the blocked_tasks[] array directly, element by element.
	 */
478 479 480 481 482
	if (rcu_preempted_readers(rnp))
		retval |= RCU_OFL_TASKS_NORM_GP;
	if (rcu_preempted_readers_exp(rnp))
		retval |= RCU_OFL_TASKS_EXP_GP;
	for (i = 0; i < 4; i++) {
483 484 485 486
		lp = &rnp->blocked_tasks[i];
		lp_root = &rnp_root->blocked_tasks[i];
		while (!list_empty(lp)) {
			tp = list_entry(lp->next, typeof(*tp), rcu_node_entry);
P
Paul E. McKenney 已提交
487
			raw_spin_lock(&rnp_root->lock); /* irqs already disabled */
488 489 490
			list_del(&tp->rcu_node_entry);
			tp->rcu_blocked_node = rnp_root;
			list_add(&tp->rcu_node_entry, lp_root);
P
Paul E. McKenney 已提交
491
			raw_spin_unlock(&rnp_root->lock); /* irqs remain disabled */
492 493
		}
	}
494
	return retval;
495 496
}

497 498 499 500 501 502 503 504 505 506
/*
 * Do CPU-offline processing for preemptable RCU.
 */
static void rcu_preempt_offline_cpu(int cpu)
{
	__rcu_offline_cpu(cpu, &rcu_preempt_state);
}

#endif /* #ifdef CONFIG_HOTPLUG_CPU */

507 508 509 510 511 512 513 514 515 516 517 518
/*
 * Check for a quiescent state from the current CPU.  When a task blocks,
 * the task is recorded in the corresponding CPU's rcu_node structure,
 * which is checked elsewhere.
 *
 * Caller must disable hard irqs.
 */
static void rcu_preempt_check_callbacks(int cpu)
{
	struct task_struct *t = current;

	if (t->rcu_read_lock_nesting == 0) {
519
		rcu_preempt_qs(cpu);
520 521
		return;
	}
522
	if (per_cpu(rcu_preempt_data, cpu).qs_pending)
523
		t->rcu_read_unlock_special |= RCU_READ_UNLOCK_NEED_QS;
524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543
}

/*
 * Process callbacks for preemptable RCU.
 */
static void rcu_preempt_process_callbacks(void)
{
	__rcu_process_callbacks(&rcu_preempt_state,
				&__get_cpu_var(rcu_preempt_data));
}

/*
 * Queue a preemptable-RCU callback for invocation after a grace period.
 */
void call_rcu(struct rcu_head *head, void (*func)(struct rcu_head *rcu))
{
	__call_rcu(head, func, &rcu_preempt_state);
}
EXPORT_SYMBOL_GPL(call_rcu);

544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559
/**
 * synchronize_rcu - wait until a grace period has elapsed.
 *
 * Control will return to the caller some time after a full grace
 * period has elapsed, in other words after all currently executing RCU
 * read-side critical sections have completed.  RCU read-side critical
 * sections are delimited by rcu_read_lock() and rcu_read_unlock(),
 * and may be nested.
 */
void synchronize_rcu(void)
{
	struct rcu_synchronize rcu;

	if (!rcu_scheduler_active)
		return;

560
	init_rcu_head_on_stack(&rcu.head);
561 562 563 564 565
	init_completion(&rcu.completion);
	/* Will wake me after RCU finished. */
	call_rcu(&rcu.head, wakeme_after_rcu);
	/* Wait for it. */
	wait_for_completion(&rcu.completion);
566
	destroy_rcu_head_on_stack(&rcu.head);
567 568 569
}
EXPORT_SYMBOL_GPL(synchronize_rcu);

570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615
static DECLARE_WAIT_QUEUE_HEAD(sync_rcu_preempt_exp_wq);
static long sync_rcu_preempt_exp_count;
static DEFINE_MUTEX(sync_rcu_preempt_exp_mutex);

/*
 * Return non-zero if there are any tasks in RCU read-side critical
 * sections blocking the current preemptible-RCU expedited grace period.
 * If there is no preemptible-RCU expedited grace period currently in
 * progress, returns zero unconditionally.
 */
static int rcu_preempted_readers_exp(struct rcu_node *rnp)
{
	return !list_empty(&rnp->blocked_tasks[2]) ||
	       !list_empty(&rnp->blocked_tasks[3]);
}

/*
 * return non-zero if there is no RCU expedited grace period in progress
 * for the specified rcu_node structure, in other words, if all CPUs and
 * tasks covered by the specified rcu_node structure have done their bit
 * for the current expedited grace period.  Works only for preemptible
 * RCU -- other RCU implementation use other means.
 *
 * Caller must hold sync_rcu_preempt_exp_mutex.
 */
static int sync_rcu_preempt_exp_done(struct rcu_node *rnp)
{
	return !rcu_preempted_readers_exp(rnp) &&
	       ACCESS_ONCE(rnp->expmask) == 0;
}

/*
 * Report the exit from RCU read-side critical section for the last task
 * that queued itself during or before the current expedited preemptible-RCU
 * grace period.  This event is reported either to the rcu_node structure on
 * which the task was queued or to one of that rcu_node structure's ancestors,
 * recursively up the tree.  (Calm down, calm down, we do the recursion
 * iteratively!)
 *
 * Caller must hold sync_rcu_preempt_exp_mutex.
 */
static void rcu_report_exp_rnp(struct rcu_state *rsp, struct rcu_node *rnp)
{
	unsigned long flags;
	unsigned long mask;

P
Paul E. McKenney 已提交
616
	raw_spin_lock_irqsave(&rnp->lock, flags);
617 618 619 620 621 622 623 624
	for (;;) {
		if (!sync_rcu_preempt_exp_done(rnp))
			break;
		if (rnp->parent == NULL) {
			wake_up(&sync_rcu_preempt_exp_wq);
			break;
		}
		mask = rnp->grpmask;
P
Paul E. McKenney 已提交
625
		raw_spin_unlock(&rnp->lock); /* irqs remain disabled */
626
		rnp = rnp->parent;
P
Paul E. McKenney 已提交
627
		raw_spin_lock(&rnp->lock); /* irqs already disabled */
628 629
		rnp->expmask &= ~mask;
	}
P
Paul E. McKenney 已提交
630
	raw_spin_unlock_irqrestore(&rnp->lock, flags);
631 632 633 634 635 636 637 638 639 640 641 642 643 644
}

/*
 * Snapshot the tasks blocking the newly started preemptible-RCU expedited
 * grace period for the specified rcu_node structure.  If there are no such
 * tasks, report it up the rcu_node hierarchy.
 *
 * Caller must hold sync_rcu_preempt_exp_mutex and rsp->onofflock.
 */
static void
sync_rcu_preempt_exp_init(struct rcu_state *rsp, struct rcu_node *rnp)
{
	int must_wait;

P
Paul E. McKenney 已提交
645
	raw_spin_lock(&rnp->lock); /* irqs already disabled */
646 647 648
	list_splice_init(&rnp->blocked_tasks[0], &rnp->blocked_tasks[2]);
	list_splice_init(&rnp->blocked_tasks[1], &rnp->blocked_tasks[3]);
	must_wait = rcu_preempted_readers_exp(rnp);
P
Paul E. McKenney 已提交
649
	raw_spin_unlock(&rnp->lock); /* irqs remain disabled */
650 651 652 653
	if (!must_wait)
		rcu_report_exp_rnp(rsp, rnp);
}

654
/*
655 656 657 658 659
 * Wait for an rcu-preempt grace period, but expedite it.  The basic idea
 * is to invoke synchronize_sched_expedited() to push all the tasks to
 * the ->blocked_tasks[] lists, move all entries from the first set of
 * ->blocked_tasks[] lists to the second set, and finally wait for this
 * second set to drain.
660 661 662
 */
void synchronize_rcu_expedited(void)
{
663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693
	unsigned long flags;
	struct rcu_node *rnp;
	struct rcu_state *rsp = &rcu_preempt_state;
	long snap;
	int trycount = 0;

	smp_mb(); /* Caller's modifications seen first by other CPUs. */
	snap = ACCESS_ONCE(sync_rcu_preempt_exp_count) + 1;
	smp_mb(); /* Above access cannot bleed into critical section. */

	/*
	 * Acquire lock, falling back to synchronize_rcu() if too many
	 * lock-acquisition failures.  Of course, if someone does the
	 * expedited grace period for us, just leave.
	 */
	while (!mutex_trylock(&sync_rcu_preempt_exp_mutex)) {
		if (trycount++ < 10)
			udelay(trycount * num_online_cpus());
		else {
			synchronize_rcu();
			return;
		}
		if ((ACCESS_ONCE(sync_rcu_preempt_exp_count) - snap) > 0)
			goto mb_ret; /* Others did our work for us. */
	}
	if ((ACCESS_ONCE(sync_rcu_preempt_exp_count) - snap) > 0)
		goto unlock_mb_ret; /* Others did our work for us. */

	/* force all RCU readers onto blocked_tasks[]. */
	synchronize_sched_expedited();

P
Paul E. McKenney 已提交
694
	raw_spin_lock_irqsave(&rsp->onofflock, flags);
695 696 697

	/* Initialize ->expmask for all non-leaf rcu_node structures. */
	rcu_for_each_nonleaf_node_breadth_first(rsp, rnp) {
P
Paul E. McKenney 已提交
698
		raw_spin_lock(&rnp->lock); /* irqs already disabled. */
699
		rnp->expmask = rnp->qsmaskinit;
P
Paul E. McKenney 已提交
700
		raw_spin_unlock(&rnp->lock); /* irqs remain disabled. */
701 702 703 704 705 706 707 708
	}

	/* Snapshot current state of ->blocked_tasks[] lists. */
	rcu_for_each_leaf_node(rsp, rnp)
		sync_rcu_preempt_exp_init(rsp, rnp);
	if (NUM_RCU_NODES > 1)
		sync_rcu_preempt_exp_init(rsp, rcu_get_root(rsp));

P
Paul E. McKenney 已提交
709
	raw_spin_unlock_irqrestore(&rsp->onofflock, flags);
710 711 712 713 714 715 716 717 718 719 720 721 722

	/* Wait for snapshotted ->blocked_tasks[] lists to drain. */
	rnp = rcu_get_root(rsp);
	wait_event(sync_rcu_preempt_exp_wq,
		   sync_rcu_preempt_exp_done(rnp));

	/* Clean up and exit. */
	smp_mb(); /* ensure expedited GP seen before counter increment. */
	ACCESS_ONCE(sync_rcu_preempt_exp_count)++;
unlock_mb_ret:
	mutex_unlock(&sync_rcu_preempt_exp_mutex);
mb_ret:
	smp_mb(); /* ensure subsequent action seen after grace period. */
723 724 725
}
EXPORT_SYMBOL_GPL(synchronize_rcu_expedited);

726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743
/*
 * Check to see if there is any immediate preemptable-RCU-related work
 * to be done.
 */
static int rcu_preempt_pending(int cpu)
{
	return __rcu_pending(&rcu_preempt_state,
			     &per_cpu(rcu_preempt_data, cpu));
}

/*
 * Does preemptable RCU need the CPU to stay out of dynticks mode?
 */
static int rcu_preempt_needs_cpu(int cpu)
{
	return !!per_cpu(rcu_preempt_data, cpu).nxtlist;
}

744 745 746 747 748 749 750 751 752
/**
 * rcu_barrier - Wait until all in-flight call_rcu() callbacks complete.
 */
void rcu_barrier(void)
{
	_rcu_barrier(&rcu_preempt_state, call_rcu);
}
EXPORT_SYMBOL_GPL(rcu_barrier);

753 754 755 756 757 758 759 760
/*
 * Initialize preemptable RCU's per-CPU data.
 */
static void __cpuinit rcu_preempt_init_percpu_data(int cpu)
{
	rcu_init_percpu_data(cpu, &rcu_preempt_state, 1);
}

761 762 763 764 765 766 767 768
/*
 * Move preemptable RCU's callbacks to ->orphan_cbs_list.
 */
static void rcu_preempt_send_cbs_to_orphanage(void)
{
	rcu_send_cbs_to_orphanage(&rcu_preempt_state);
}

769 770 771 772 773
/*
 * Initialize preemptable RCU's state structures.
 */
static void __init __rcu_init_preempt(void)
{
774
	rcu_init_one(&rcu_preempt_state, &rcu_preempt_data);
775 776
}

777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797
/*
 * Check for a task exiting while in a preemptable-RCU read-side
 * critical section, clean up if so.  No need to issue warnings,
 * as debug_check_no_locks_held() already does this if lockdep
 * is enabled.
 */
void exit_rcu(void)
{
	struct task_struct *t = current;

	if (t->rcu_read_lock_nesting == 0)
		return;
	t->rcu_read_lock_nesting = 1;
	rcu_read_unlock();
}

#else /* #ifdef CONFIG_TREE_PREEMPT_RCU */

/*
 * Tell them what RCU they are running.
 */
798
static void __init rcu_bootup_announce(void)
799 800
{
	printk(KERN_INFO "Hierarchical RCU implementation.\n");
801
	rcu_bootup_announce_oddness();
802 803 804 805 806 807 808 809 810 811 812
}

/*
 * Return the number of RCU batches processed thus far for debug & stats.
 */
long rcu_batches_completed(void)
{
	return rcu_batches_completed_sched();
}
EXPORT_SYMBOL_GPL(rcu_batches_completed);

813 814 815 816 817 818 819 820 821 822
/*
 * Force a quiescent state for RCU, which, because there is no preemptible
 * RCU, becomes the same as rcu-sched.
 */
void rcu_force_quiescent_state(void)
{
	rcu_sched_force_quiescent_state();
}
EXPORT_SYMBOL_GPL(rcu_force_quiescent_state);

823 824 825 826
/*
 * Because preemptable RCU does not exist, we never have to check for
 * CPUs being in quiescent states.
 */
827
static void rcu_preempt_note_context_switch(int cpu)
828 829 830
{
}

831 832 833 834 835 836 837 838 839
/*
 * Because preemptable RCU does not exist, there are never any preempted
 * RCU readers.
 */
static int rcu_preempted_readers(struct rcu_node *rnp)
{
	return 0;
}

840 841 842
#ifdef CONFIG_HOTPLUG_CPU

/* Because preemptible RCU does not exist, no quieting of tasks. */
P
Paul E. McKenney 已提交
843
static void rcu_report_unblock_qs_rnp(struct rcu_node *rnp, unsigned long flags)
844
{
P
Paul E. McKenney 已提交
845
	raw_spin_unlock_irqrestore(&rnp->lock, flags);
846 847 848 849
}

#endif /* #ifdef CONFIG_HOTPLUG_CPU */

850 851
#ifdef CONFIG_RCU_CPU_STALL_DETECTOR

852 853 854 855 856 857 858 859
/*
 * Because preemptable RCU does not exist, we never have to check for
 * tasks blocked within RCU read-side critical sections.
 */
static void rcu_print_detail_task_stall(struct rcu_state *rsp)
{
}

860 861 862 863 864 865 866 867 868 869
/*
 * Because preemptable RCU does not exist, we never have to check for
 * tasks blocked within RCU read-side critical sections.
 */
static void rcu_print_task_stall(struct rcu_node *rnp)
{
}

#endif /* #ifdef CONFIG_RCU_CPU_STALL_DETECTOR */

870 871
/*
 * Because there is no preemptable RCU, there can be no readers blocked,
872 873
 * so there is no need to check for blocked tasks.  So check only for
 * bogus qsmask values.
874 875 876
 */
static void rcu_preempt_check_blocked_tasks(struct rcu_node *rnp)
{
877
	WARN_ON_ONCE(rnp->qsmask);
878 879
}

880 881
#ifdef CONFIG_HOTPLUG_CPU

882 883
/*
 * Because preemptable RCU does not exist, it never needs to migrate
884 885 886
 * tasks that were blocked within RCU read-side critical sections, and
 * such non-existent tasks cannot possibly have been blocking the current
 * grace period.
887
 */
888 889 890
static int rcu_preempt_offline_tasks(struct rcu_state *rsp,
				     struct rcu_node *rnp,
				     struct rcu_data *rdp)
891
{
892
	return 0;
893 894
}

895 896 897 898 899 900 901 902 903 904
/*
 * Because preemptable RCU does not exist, it never needs CPU-offline
 * processing.
 */
static void rcu_preempt_offline_cpu(int cpu)
{
}

#endif /* #ifdef CONFIG_HOTPLUG_CPU */

905 906 907 908
/*
 * Because preemptable RCU does not exist, it never has any callbacks
 * to check.
 */
909
static void rcu_preempt_check_callbacks(int cpu)
910 911 912 913 914 915 916
{
}

/*
 * Because preemptable RCU does not exist, it never has any callbacks
 * to process.
 */
917
static void rcu_preempt_process_callbacks(void)
918 919 920 921 922 923 924 925 926 927 928 929
{
}

/*
 * In classic RCU, call_rcu() is just call_rcu_sched().
 */
void call_rcu(struct rcu_head *head, void (*func)(struct rcu_head *rcu))
{
	call_rcu_sched(head, func);
}
EXPORT_SYMBOL_GPL(call_rcu);

930 931 932 933 934 935 936 937 938 939
/*
 * Wait for an rcu-preempt grace period, but make it happen quickly.
 * But because preemptable RCU does not exist, map to rcu-sched.
 */
void synchronize_rcu_expedited(void)
{
	synchronize_sched_expedited();
}
EXPORT_SYMBOL_GPL(synchronize_rcu_expedited);

940 941 942 943 944 945 946 947 948 949 950 951 952 953
#ifdef CONFIG_HOTPLUG_CPU

/*
 * Because preemptable RCU does not exist, there is never any need to
 * report on tasks preempted in RCU read-side critical sections during
 * expedited RCU grace periods.
 */
static void rcu_report_exp_rnp(struct rcu_state *rsp, struct rcu_node *rnp)
{
	return;
}

#endif /* #ifdef CONFIG_HOTPLUG_CPU */

954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969
/*
 * Because preemptable RCU does not exist, it never has any work to do.
 */
static int rcu_preempt_pending(int cpu)
{
	return 0;
}

/*
 * Because preemptable RCU does not exist, it never needs any CPU.
 */
static int rcu_preempt_needs_cpu(int cpu)
{
	return 0;
}

970 971 972 973 974 975 976 977 978 979
/*
 * Because preemptable RCU does not exist, rcu_barrier() is just
 * another name for rcu_barrier_sched().
 */
void rcu_barrier(void)
{
	rcu_barrier_sched();
}
EXPORT_SYMBOL_GPL(rcu_barrier);

980 981 982 983 984 985 986 987
/*
 * Because preemptable RCU does not exist, there is no per-CPU
 * data to initialize.
 */
static void __cpuinit rcu_preempt_init_percpu_data(int cpu)
{
}

988 989 990 991 992 993 994
/*
 * Because there is no preemptable RCU, there are no callbacks to move.
 */
static void rcu_preempt_send_cbs_to_orphanage(void)
{
}

995 996 997 998 999 1000 1001
/*
 * Because preemptable RCU does not exist, it need not be initialized.
 */
static void __init __rcu_init_preempt(void)
{
}

1002
#endif /* #else #ifdef CONFIG_TREE_PREEMPT_RCU */
1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020

#if !defined(CONFIG_RCU_FAST_NO_HZ)

/*
 * Check to see if any future RCU-related work will need to be done
 * by the current CPU, even if none need be done immediately, returning
 * 1 if so.  This function is part of the RCU implementation; it is -not-
 * an exported member of the RCU API.
 *
 * Because we have preemptible RCU, just check whether this CPU needs
 * any flavor of RCU.  Do not chew up lots of CPU cycles with preemption
 * disabled in a most-likely vain attempt to cause RCU not to need this CPU.
 */
int rcu_needs_cpu(int cpu)
{
	return rcu_needs_cpu_quick_check(cpu);
}

1021 1022 1023 1024 1025 1026 1027 1028 1029
/*
 * Check to see if we need to continue a callback-flush operations to
 * allow the last CPU to enter dyntick-idle mode.  But fast dyntick-idle
 * entry is not configured, so we never do need to.
 */
static void rcu_needs_cpu_flush(void)
{
}

1030 1031 1032
#else /* #if !defined(CONFIG_RCU_FAST_NO_HZ) */

#define RCU_NEEDS_CPU_FLUSHES 5
1033
static DEFINE_PER_CPU(int, rcu_dyntick_drain);
1034
static DEFINE_PER_CPU(unsigned long, rcu_dyntick_holdoff);
1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046

/*
 * Check to see if any future RCU-related work will need to be done
 * by the current CPU, even if none need be done immediately, returning
 * 1 if so.  This function is part of the RCU implementation; it is -not-
 * an exported member of the RCU API.
 *
 * Because we are not supporting preemptible RCU, attempt to accelerate
 * any current grace periods so that RCU no longer needs this CPU, but
 * only if all other CPUs are already in dynticks-idle mode.  This will
 * allow the CPU cores to be powered down immediately, as opposed to after
 * waiting many milliseconds for grace periods to elapse.
1047 1048 1049 1050 1051
 *
 * Because it is not legal to invoke rcu_process_callbacks() with irqs
 * disabled, we do one pass of force_quiescent_state(), then do a
 * raise_softirq() to cause rcu_process_callbacks() to be invoked later.
 * The per-cpu rcu_dyntick_drain variable controls the sequencing.
1052 1053 1054
 */
int rcu_needs_cpu(int cpu)
{
1055
	int c = 0;
1056 1057
	int snap;
	int snap_nmi;
1058 1059
	int thatcpu;

1060 1061 1062 1063
	/* Check for being in the holdoff period. */
	if (per_cpu(rcu_dyntick_holdoff, cpu) == jiffies)
		return rcu_needs_cpu_quick_check(cpu);

1064
	/* Don't bother unless we are the last non-dyntick-idle CPU. */
1065 1066 1067
	for_each_online_cpu(thatcpu) {
		if (thatcpu == cpu)
			continue;
1068 1069
		snap = per_cpu(rcu_dynticks, thatcpu).dynticks;
		snap_nmi = per_cpu(rcu_dynticks, thatcpu).dynticks_nmi;
1070 1071
		smp_mb(); /* Order sampling of snap with end of grace period. */
		if (((snap & 0x1) != 0) || ((snap_nmi & 0x1) != 0)) {
1072
			per_cpu(rcu_dyntick_drain, cpu) = 0;
1073
			per_cpu(rcu_dyntick_holdoff, cpu) = jiffies - 1;
1074 1075
			return rcu_needs_cpu_quick_check(cpu);
		}
1076
	}
1077 1078 1079 1080 1081 1082 1083

	/* Check and update the rcu_dyntick_drain sequencing. */
	if (per_cpu(rcu_dyntick_drain, cpu) <= 0) {
		/* First time through, initialize the counter. */
		per_cpu(rcu_dyntick_drain, cpu) = RCU_NEEDS_CPU_FLUSHES;
	} else if (--per_cpu(rcu_dyntick_drain, cpu) <= 0) {
		/* We have hit the limit, so time to give up. */
1084
		per_cpu(rcu_dyntick_holdoff, cpu) = jiffies;
1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097
		return rcu_needs_cpu_quick_check(cpu);
	}

	/* Do one step pushing remaining RCU callbacks through. */
	if (per_cpu(rcu_sched_data, cpu).nxtlist) {
		rcu_sched_qs(cpu);
		force_quiescent_state(&rcu_sched_state, 0);
		c = c || per_cpu(rcu_sched_data, cpu).nxtlist;
	}
	if (per_cpu(rcu_bh_data, cpu).nxtlist) {
		rcu_bh_qs(cpu);
		force_quiescent_state(&rcu_bh_state, 0);
		c = c || per_cpu(rcu_bh_data, cpu).nxtlist;
1098 1099 1100
	}

	/* If RCU callbacks are still pending, RCU still needs this CPU. */
1101
	if (c)
1102
		raise_softirq(RCU_SOFTIRQ);
1103 1104 1105
	return c;
}

1106 1107 1108 1109 1110 1111 1112
/*
 * Check to see if we need to continue a callback-flush operations to
 * allow the last CPU to enter dyntick-idle mode.
 */
static void rcu_needs_cpu_flush(void)
{
	int cpu = smp_processor_id();
1113
	unsigned long flags;
1114 1115 1116

	if (per_cpu(rcu_dyntick_drain, cpu) <= 0)
		return;
1117
	local_irq_save(flags);
1118
	(void)rcu_needs_cpu(cpu);
1119
	local_irq_restore(flags);
1120 1121
}

1122
#endif /* #else #if !defined(CONFIG_RCU_FAST_NO_HZ) */