manage.c 56.7 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3
/*
 * linux/kernel/irq/manage.c
 *
I
Ingo Molnar 已提交
4 5
 * Copyright (C) 1992, 1998-2006 Linus Torvalds, Ingo Molnar
 * Copyright (C) 2005-2006 Thomas Gleixner
L
Linus Torvalds 已提交
6 7 8 9
 *
 * This file contains driver APIs to the irq subsystem.
 */

10 11
#define pr_fmt(fmt) "genirq: " fmt

L
Linus Torvalds 已提交
12
#include <linux/irq.h>
13
#include <linux/kthread.h>
L
Linus Torvalds 已提交
14 15 16
#include <linux/module.h>
#include <linux/random.h>
#include <linux/interrupt.h>
17
#include <linux/slab.h>
18
#include <linux/sched.h>
19
#include <linux/sched/rt.h>
20
#include <linux/sched/task.h>
21
#include <uapi/linux/sched/types.h>
22
#include <linux/task_work.h>
L
Linus Torvalds 已提交
23 24 25

#include "internals.h"

26 27 28 29 30 31 32 33 34 35 36
#ifdef CONFIG_IRQ_FORCED_THREADING
__read_mostly bool force_irqthreads;

static int __init setup_forced_irqthreads(char *arg)
{
	force_irqthreads = true;
	return 0;
}
early_param("threadirqs", setup_forced_irqthreads);
#endif

37
static void __synchronize_hardirq(struct irq_desc *desc)
L
Linus Torvalds 已提交
38
{
39
	bool inprogress;
L
Linus Torvalds 已提交
40

41 42 43 44 45 46 47
	do {
		unsigned long flags;

		/*
		 * Wait until we're out of the critical section.  This might
		 * give the wrong answer due to the lack of memory barriers.
		 */
48
		while (irqd_irq_inprogress(&desc->irq_data))
49 50 51
			cpu_relax();

		/* Ok, that indicated we're done: double-check carefully. */
52
		raw_spin_lock_irqsave(&desc->lock, flags);
53
		inprogress = irqd_irq_inprogress(&desc->irq_data);
54
		raw_spin_unlock_irqrestore(&desc->lock, flags);
55 56

		/* Oops, that failed? */
57
	} while (inprogress);
58 59 60 61 62 63 64 65 66 67 68 69 70 71 72
}

/**
 *	synchronize_hardirq - wait for pending hard IRQ handlers (on other CPUs)
 *	@irq: interrupt number to wait for
 *
 *	This function waits for any pending hard IRQ handlers for this
 *	interrupt to complete before returning. If you use this
 *	function while holding a resource the IRQ handler may need you
 *	will deadlock. It does not take associated threaded handlers
 *	into account.
 *
 *	Do not use this for shutdown scenarios where you must be sure
 *	that all parts (hardirq and threaded handler) have completed.
 *
73 74
 *	Returns: false if a threaded handler is active.
 *
75 76
 *	This function may be called - with care - from IRQ context.
 */
77
bool synchronize_hardirq(unsigned int irq)
78 79
{
	struct irq_desc *desc = irq_to_desc(irq);
80

81
	if (desc) {
82
		__synchronize_hardirq(desc);
83 84 85 86
		return !atomic_read(&desc->threads_active);
	}

	return true;
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
}
EXPORT_SYMBOL(synchronize_hardirq);

/**
 *	synchronize_irq - wait for pending IRQ handlers (on other CPUs)
 *	@irq: interrupt number to wait for
 *
 *	This function waits for any pending IRQ handlers for this interrupt
 *	to complete before returning. If you use this function while
 *	holding a resource the IRQ handler may need you will deadlock.
 *
 *	This function may be called - with care - from IRQ context.
 */
void synchronize_irq(unsigned int irq)
{
	struct irq_desc *desc = irq_to_desc(irq);

	if (desc) {
		__synchronize_hardirq(desc);
		/*
		 * We made sure that no hardirq handler is
		 * running. Now verify that no threaded handlers are
		 * active.
		 */
		wait_event(desc->wait_for_threads,
			   !atomic_read(&desc->threads_active));
	}
L
Linus Torvalds 已提交
114 115 116
}
EXPORT_SYMBOL(synchronize_irq);

117 118 119
#ifdef CONFIG_SMP
cpumask_var_t irq_default_affinity;

120
static bool __irq_can_set_affinity(struct irq_desc *desc)
121 122 123
{
	if (!desc || !irqd_can_balance(&desc->irq_data) ||
	    !desc->irq_data.chip || !desc->irq_data.chip->irq_set_affinity)
124 125
		return false;
	return true;
126 127
}

128 129 130 131 132 133 134
/**
 *	irq_can_set_affinity - Check if the affinity of a given irq can be set
 *	@irq:		Interrupt to check
 *
 */
int irq_can_set_affinity(unsigned int irq)
{
135
	return __irq_can_set_affinity(irq_to_desc(irq));
136 137
}

138 139 140 141 142 143 144 145 146 147 148 149 150 151 152
/**
 * irq_can_set_affinity_usr - Check if affinity of a irq can be set from user space
 * @irq:	Interrupt to check
 *
 * Like irq_can_set_affinity() above, but additionally checks for the
 * AFFINITY_MANAGED flag.
 */
bool irq_can_set_affinity_usr(unsigned int irq)
{
	struct irq_desc *desc = irq_to_desc(irq);

	return __irq_can_set_affinity(desc) &&
		!irqd_affinity_is_managed(&desc->irq_data);
}

153 154 155 156 157 158 159 160 161 162
/**
 *	irq_set_thread_affinity - Notify irq threads to adjust affinity
 *	@desc:		irq descriptor which has affitnity changed
 *
 *	We just set IRQTF_AFFINITY and delegate the affinity setting
 *	to the interrupt thread itself. We can not call
 *	set_cpus_allowed_ptr() here as we hold desc->lock and this
 *	code can be called from hard interrupt context.
 */
void irq_set_thread_affinity(struct irq_desc *desc)
163
{
164
	struct irqaction *action;
165

166
	for_each_action_of_desc(desc, action)
167
		if (action->thread)
168
			set_bit(IRQTF_AFFINITY, &action->thread_flags);
169 170
}

171 172 173 174 175 176 177 178 179 180 181 182 183
static void irq_validate_effective_affinity(struct irq_data *data)
{
#ifdef CONFIG_GENERIC_IRQ_EFFECTIVE_AFF_MASK
	const struct cpumask *m = irq_data_get_effective_affinity_mask(data);
	struct irq_chip *chip = irq_data_get_irq_chip(data);

	if (!cpumask_empty(m))
		return;
	pr_warn_once("irq_chip %s did not update eff. affinity mask of irq %u\n",
		     chip->name, data->irq);
#endif
}

184 185 186 187 188 189 190
int irq_do_set_affinity(struct irq_data *data, const struct cpumask *mask,
			bool force)
{
	struct irq_desc *desc = irq_data_to_desc(data);
	struct irq_chip *chip = irq_data_get_irq_chip(data);
	int ret;

191
	ret = chip->irq_set_affinity(data, mask, force);
192 193
	switch (ret) {
	case IRQ_SET_MASK_OK:
194
	case IRQ_SET_MASK_OK_DONE:
195
		cpumask_copy(desc->irq_common_data.affinity, mask);
196
	case IRQ_SET_MASK_OK_NOCOPY:
197
		irq_validate_effective_affinity(data);
198 199 200 201 202 203 204
		irq_set_thread_affinity(desc);
		ret = 0;
	}

	return ret;
}

205 206
int irq_set_affinity_locked(struct irq_data *data, const struct cpumask *mask,
			    bool force)
207
{
208 209
	struct irq_chip *chip = irq_data_get_irq_chip(data);
	struct irq_desc *desc = irq_data_to_desc(data);
210
	int ret = 0;
211

212
	if (!chip || !chip->irq_set_affinity)
213 214
		return -EINVAL;

T
Thomas Gleixner 已提交
215
	if (irq_can_move_pcntxt(data)) {
216
		ret = irq_do_set_affinity(data, mask, force);
217
	} else {
218
		irqd_set_move_pending(data);
219
		irq_copy_pending(desc, mask);
220
	}
221

222 223 224 225
	if (desc->affinity_notify) {
		kref_get(&desc->affinity_notify->kref);
		schedule_work(&desc->affinity_notify->work);
	}
226 227 228 229 230
	irqd_set(data, IRQD_AFFINITY_SET);

	return ret;
}

231
int __irq_set_affinity(unsigned int irq, const struct cpumask *mask, bool force)
232 233 234 235 236 237 238 239 240
{
	struct irq_desc *desc = irq_to_desc(irq);
	unsigned long flags;
	int ret;

	if (!desc)
		return -EINVAL;

	raw_spin_lock_irqsave(&desc->lock, flags);
241
	ret = irq_set_affinity_locked(irq_desc_get_irq_data(desc), mask, force);
242
	raw_spin_unlock_irqrestore(&desc->lock, flags);
243
	return ret;
244 245
}

246 247 248
int irq_set_affinity_hint(unsigned int irq, const struct cpumask *m)
{
	unsigned long flags;
249
	struct irq_desc *desc = irq_get_desc_lock(irq, &flags, IRQ_GET_DESC_CHECK_GLOBAL);
250 251 252 253

	if (!desc)
		return -EINVAL;
	desc->affinity_hint = m;
254
	irq_put_desc_unlock(desc, flags);
255
	/* set the initial affinity to prevent every interrupt being on CPU0 */
256 257
	if (m)
		__irq_set_affinity(irq, m, false);
258 259 260 261
	return 0;
}
EXPORT_SYMBOL_GPL(irq_set_affinity_hint);

262 263 264 265 266 267 268 269
static void irq_affinity_notify(struct work_struct *work)
{
	struct irq_affinity_notify *notify =
		container_of(work, struct irq_affinity_notify, work);
	struct irq_desc *desc = irq_to_desc(notify->irq);
	cpumask_var_t cpumask;
	unsigned long flags;

270
	if (!desc || !alloc_cpumask_var(&cpumask, GFP_KERNEL))
271 272 273
		goto out;

	raw_spin_lock_irqsave(&desc->lock, flags);
T
Thomas Gleixner 已提交
274
	if (irq_move_pending(&desc->irq_data))
275
		irq_get_pending(cpumask, desc);
276
	else
277
		cpumask_copy(cpumask, desc->irq_common_data.affinity);
278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329
	raw_spin_unlock_irqrestore(&desc->lock, flags);

	notify->notify(notify, cpumask);

	free_cpumask_var(cpumask);
out:
	kref_put(&notify->kref, notify->release);
}

/**
 *	irq_set_affinity_notifier - control notification of IRQ affinity changes
 *	@irq:		Interrupt for which to enable/disable notification
 *	@notify:	Context for notification, or %NULL to disable
 *			notification.  Function pointers must be initialised;
 *			the other fields will be initialised by this function.
 *
 *	Must be called in process context.  Notification may only be enabled
 *	after the IRQ is allocated and must be disabled before the IRQ is
 *	freed using free_irq().
 */
int
irq_set_affinity_notifier(unsigned int irq, struct irq_affinity_notify *notify)
{
	struct irq_desc *desc = irq_to_desc(irq);
	struct irq_affinity_notify *old_notify;
	unsigned long flags;

	/* The release function is promised process context */
	might_sleep();

	if (!desc)
		return -EINVAL;

	/* Complete initialisation of *notify */
	if (notify) {
		notify->irq = irq;
		kref_init(&notify->kref);
		INIT_WORK(&notify->work, irq_affinity_notify);
	}

	raw_spin_lock_irqsave(&desc->lock, flags);
	old_notify = desc->affinity_notify;
	desc->affinity_notify = notify;
	raw_spin_unlock_irqrestore(&desc->lock, flags);

	if (old_notify)
		kref_put(&old_notify->kref, old_notify->release);

	return 0;
}
EXPORT_SYMBOL_GPL(irq_set_affinity_notifier);

330 331 332 333
#ifndef CONFIG_AUTO_IRQ_AFFINITY
/*
 * Generic version of the affinity autoselector.
 */
334
int irq_setup_affinity(struct irq_desc *desc)
335
{
336
	struct cpumask *set = irq_default_affinity;
337 338 339
	int ret, node = irq_desc_get_node(desc);
	static DEFINE_RAW_SPINLOCK(mask_lock);
	static struct cpumask mask;
340

341
	/* Excludes PER_CPU and NO_BALANCE interrupts */
342
	if (!__irq_can_set_affinity(desc))
343 344
		return 0;

345
	raw_spin_lock(&mask_lock);
346
	/*
347
	 * Preserve the managed affinity setting and a userspace affinity
348
	 * setup, but make sure that one of the targets is online.
349
	 */
350 351
	if (irqd_affinity_is_managed(&desc->irq_data) ||
	    irqd_has_set(&desc->irq_data, IRQD_AFFINITY_SET)) {
352
		if (cpumask_intersects(desc->irq_common_data.affinity,
353
				       cpu_online_mask))
354
			set = desc->irq_common_data.affinity;
T
Thomas Gleixner 已提交
355
		else
356
			irqd_clear(&desc->irq_data, IRQD_AFFINITY_SET);
357
	}
358

359
	cpumask_and(&mask, cpu_online_mask, set);
360 361 362 363
	if (node != NUMA_NO_NODE) {
		const struct cpumask *nodemask = cpumask_of_node(node);

		/* make sure at least one of the cpus in nodemask is online */
364 365
		if (cpumask_intersects(&mask, nodemask))
			cpumask_and(&mask, &mask, nodemask);
366
	}
367 368 369
	ret = irq_do_set_affinity(&desc->irq_data, &mask, false);
	raw_spin_unlock(&mask_lock);
	return ret;
370
}
371
#else
372
/* Wrapper for ALPHA specific affinity selector magic */
373
int irq_setup_affinity(struct irq_desc *desc)
374
{
375
	return irq_select_affinity(irq_desc_get_irq(desc));
376
}
377 378
#endif

379
/*
380
 * Called when a bogus affinity is set via /proc/irq
381
 */
382
int irq_select_affinity_usr(unsigned int irq)
383 384 385 386 387
{
	struct irq_desc *desc = irq_to_desc(irq);
	unsigned long flags;
	int ret;

388
	raw_spin_lock_irqsave(&desc->lock, flags);
389
	ret = irq_setup_affinity(desc);
390
	raw_spin_unlock_irqrestore(&desc->lock, flags);
391 392
	return ret;
}
L
Linus Torvalds 已提交
393 394
#endif

395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416
/**
 *	irq_set_vcpu_affinity - Set vcpu affinity for the interrupt
 *	@irq: interrupt number to set affinity
 *	@vcpu_info: vCPU specific data
 *
 *	This function uses the vCPU specific data to set the vCPU
 *	affinity for an irq. The vCPU specific data is passed from
 *	outside, such as KVM. One example code path is as below:
 *	KVM -> IOMMU -> irq_set_vcpu_affinity().
 */
int irq_set_vcpu_affinity(unsigned int irq, void *vcpu_info)
{
	unsigned long flags;
	struct irq_desc *desc = irq_get_desc_lock(irq, &flags, 0);
	struct irq_data *data;
	struct irq_chip *chip;
	int ret = -ENOSYS;

	if (!desc)
		return -EINVAL;

	data = irq_desc_get_irq_data(desc);
417 418 419 420 421 422 423 424 425 426 427 428
	do {
		chip = irq_data_get_irq_chip(data);
		if (chip && chip->irq_set_vcpu_affinity)
			break;
#ifdef CONFIG_IRQ_DOMAIN_HIERARCHY
		data = data->parent_data;
#else
		data = NULL;
#endif
	} while (data);

	if (data)
429 430 431 432 433 434 435
		ret = chip->irq_set_vcpu_affinity(data, vcpu_info);
	irq_put_desc_unlock(desc, flags);

	return ret;
}
EXPORT_SYMBOL_GPL(irq_set_vcpu_affinity);

436
void __disable_irq(struct irq_desc *desc)
437
{
438
	if (!desc->depth++)
439
		irq_disable(desc);
440 441
}

442 443 444
static int __disable_irq_nosync(unsigned int irq)
{
	unsigned long flags;
445
	struct irq_desc *desc = irq_get_desc_buslock(irq, &flags, IRQ_GET_DESC_CHECK_GLOBAL);
446 447 448

	if (!desc)
		return -EINVAL;
449
	__disable_irq(desc);
450 451 452 453
	irq_put_desc_busunlock(desc, flags);
	return 0;
}

L
Linus Torvalds 已提交
454 455 456 457 458 459 460 461 462 463 464 465 466
/**
 *	disable_irq_nosync - disable an irq without waiting
 *	@irq: Interrupt to disable
 *
 *	Disable the selected interrupt line.  Disables and Enables are
 *	nested.
 *	Unlike disable_irq(), this function does not ensure existing
 *	instances of the IRQ handler have completed before returning.
 *
 *	This function may be called from IRQ context.
 */
void disable_irq_nosync(unsigned int irq)
{
467
	__disable_irq_nosync(irq);
L
Linus Torvalds 已提交
468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484
}
EXPORT_SYMBOL(disable_irq_nosync);

/**
 *	disable_irq - disable an irq and wait for completion
 *	@irq: Interrupt to disable
 *
 *	Disable the selected interrupt line.  Enables and Disables are
 *	nested.
 *	This function waits for any pending IRQ handlers for this interrupt
 *	to complete before returning. If you use this function while
 *	holding a resource the IRQ handler may need you will deadlock.
 *
 *	This function may be called - with care - from IRQ context.
 */
void disable_irq(unsigned int irq)
{
485
	if (!__disable_irq_nosync(irq))
L
Linus Torvalds 已提交
486 487 488 489
		synchronize_irq(irq);
}
EXPORT_SYMBOL(disable_irq);

490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515
/**
 *	disable_hardirq - disables an irq and waits for hardirq completion
 *	@irq: Interrupt to disable
 *
 *	Disable the selected interrupt line.  Enables and Disables are
 *	nested.
 *	This function waits for any pending hard IRQ handlers for this
 *	interrupt to complete before returning. If you use this function while
 *	holding a resource the hard IRQ handler may need you will deadlock.
 *
 *	When used to optimistically disable an interrupt from atomic context
 *	the return value must be checked.
 *
 *	Returns: false if a threaded handler is active.
 *
 *	This function may be called - with care - from IRQ context.
 */
bool disable_hardirq(unsigned int irq)
{
	if (!__disable_irq_nosync(irq))
		return synchronize_hardirq(irq);

	return false;
}
EXPORT_SYMBOL_GPL(disable_hardirq);

516
void __enable_irq(struct irq_desc *desc)
517 518 519
{
	switch (desc->depth) {
	case 0:
520
 err_out:
521 522
		WARN(1, KERN_WARNING "Unbalanced enable for IRQ %d\n",
		     irq_desc_get_irq(desc));
523 524
		break;
	case 1: {
525
		if (desc->istate & IRQS_SUSPENDED)
526
			goto err_out;
527
		/* Prevent probing on this irq: */
528
		irq_settings_set_noprobe(desc);
529 530 531 532 533 534 535
		/*
		 * Call irq_startup() not irq_enable() here because the
		 * interrupt might be marked NOAUTOEN. So irq_startup()
		 * needs to be invoked when it gets enabled the first
		 * time. If it was already started up, then irq_startup()
		 * will invoke irq_enable() under the hood.
		 */
536
		irq_startup(desc, IRQ_RESEND, IRQ_START_COND);
537
		break;
538 539 540 541 542 543
	}
	default:
		desc->depth--;
	}
}

L
Linus Torvalds 已提交
544 545 546 547 548 549 550 551
/**
 *	enable_irq - enable handling of an irq
 *	@irq: Interrupt to enable
 *
 *	Undoes the effect of one call to disable_irq().  If this
 *	matches the last disable, processing of interrupts on this
 *	IRQ line is re-enabled.
 *
T
Thomas Gleixner 已提交
552
 *	This function may be called from IRQ context only when
553
 *	desc->irq_data.chip->bus_lock and desc->chip->bus_sync_unlock are NULL !
L
Linus Torvalds 已提交
554 555 556 557
 */
void enable_irq(unsigned int irq)
{
	unsigned long flags;
558
	struct irq_desc *desc = irq_get_desc_buslock(irq, &flags, IRQ_GET_DESC_CHECK_GLOBAL);
L
Linus Torvalds 已提交
559

560
	if (!desc)
561
		return;
T
Thomas Gleixner 已提交
562 563
	if (WARN(!desc->irq_data.chip,
		 KERN_ERR "enable_irq before setup/request_irq: irq %u\n", irq))
564
		goto out;
565

566
	__enable_irq(desc);
567 568
out:
	irq_put_desc_busunlock(desc, flags);
L
Linus Torvalds 已提交
569 570 571
}
EXPORT_SYMBOL(enable_irq);

D
David Brownell 已提交
572
static int set_irq_wake_real(unsigned int irq, unsigned int on)
573
{
574
	struct irq_desc *desc = irq_to_desc(irq);
575 576
	int ret = -ENXIO;

577 578 579
	if (irq_desc_get_chip(desc)->flags &  IRQCHIP_SKIP_SET_WAKE)
		return 0;

580 581
	if (desc->irq_data.chip->irq_set_wake)
		ret = desc->irq_data.chip->irq_set_wake(&desc->irq_data, on);
582 583 584 585

	return ret;
}

586
/**
T
Thomas Gleixner 已提交
587
 *	irq_set_irq_wake - control irq power management wakeup
588 589 590
 *	@irq:	interrupt to control
 *	@on:	enable/disable power management wakeup
 *
591 592 593 594 595 596
 *	Enable/disable power management wakeup mode, which is
 *	disabled by default.  Enables and disables must match,
 *	just as they match for non-wakeup mode support.
 *
 *	Wakeup mode lets this IRQ wake the system from sleep
 *	states like "suspend to RAM".
597
 */
T
Thomas Gleixner 已提交
598
int irq_set_irq_wake(unsigned int irq, unsigned int on)
599 600
{
	unsigned long flags;
601
	struct irq_desc *desc = irq_get_desc_buslock(irq, &flags, IRQ_GET_DESC_CHECK_GLOBAL);
602
	int ret = 0;
603

604 605 606
	if (!desc)
		return -EINVAL;

607 608 609 610
	/* wakeup-capable irqs can be shared between drivers that
	 * don't need to have the same sleep mode behaviors.
	 */
	if (on) {
611 612 613 614 615
		if (desc->wake_depth++ == 0) {
			ret = set_irq_wake_real(irq, on);
			if (ret)
				desc->wake_depth = 0;
			else
616
				irqd_set(&desc->irq_data, IRQD_WAKEUP_STATE);
617
		}
618 619
	} else {
		if (desc->wake_depth == 0) {
620
			WARN(1, "Unbalanced IRQ %d wake disable\n", irq);
621 622 623 624 625
		} else if (--desc->wake_depth == 0) {
			ret = set_irq_wake_real(irq, on);
			if (ret)
				desc->wake_depth = 1;
			else
626
				irqd_clear(&desc->irq_data, IRQD_WAKEUP_STATE);
627
		}
628
	}
629
	irq_put_desc_busunlock(desc, flags);
630 631
	return ret;
}
T
Thomas Gleixner 已提交
632
EXPORT_SYMBOL(irq_set_irq_wake);
633

L
Linus Torvalds 已提交
634 635 636 637 638 639 640
/*
 * Internal function that tells the architecture code whether a
 * particular irq has been exclusively allocated or is available
 * for driver use.
 */
int can_request_irq(unsigned int irq, unsigned long irqflags)
{
641
	unsigned long flags;
642
	struct irq_desc *desc = irq_get_desc_lock(irq, &flags, 0);
643
	int canrequest = 0;
L
Linus Torvalds 已提交
644

645 646 647
	if (!desc)
		return 0;

648
	if (irq_settings_can_request(desc)) {
649 650 651
		if (!desc->action ||
		    irqflags & desc->action->flags & IRQF_SHARED)
			canrequest = 1;
652 653 654
	}
	irq_put_desc_unlock(desc, flags);
	return canrequest;
L
Linus Torvalds 已提交
655 656
}

657
int __irq_set_trigger(struct irq_desc *desc, unsigned long flags)
658
{
659
	struct irq_chip *chip = desc->irq_data.chip;
660
	int ret, unmask = 0;
661

662
	if (!chip || !chip->irq_set_type) {
663 664 665 666
		/*
		 * IRQF_TRIGGER_* but the PIC does not support multiple
		 * flow-types?
		 */
667 668
		pr_debug("No set_type function for IRQ %d (%s)\n",
			 irq_desc_get_irq(desc),
669
			 chip ? (chip->name ? : "unknown") : "unknown");
670 671 672
		return 0;
	}

673
	if (chip->flags & IRQCHIP_SET_TYPE_MASKED) {
674
		if (!irqd_irq_masked(&desc->irq_data))
675
			mask_irq(desc);
676
		if (!irqd_irq_disabled(&desc->irq_data))
677 678 679
			unmask = 1;
	}

680 681
	/* Mask all flags except trigger mode */
	flags &= IRQ_TYPE_SENSE_MASK;
682
	ret = chip->irq_set_type(&desc->irq_data, flags);
683

684 685
	switch (ret) {
	case IRQ_SET_MASK_OK:
686
	case IRQ_SET_MASK_OK_DONE:
687 688 689 690 691 692 693 694 695 696 697 698
		irqd_clear(&desc->irq_data, IRQD_TRIGGER_MASK);
		irqd_set(&desc->irq_data, flags);

	case IRQ_SET_MASK_OK_NOCOPY:
		flags = irqd_get_trigger_type(&desc->irq_data);
		irq_settings_set_trigger_mask(desc, flags);
		irqd_clear(&desc->irq_data, IRQD_LEVEL);
		irq_settings_clr_level(desc);
		if (flags & IRQ_TYPE_LEVEL_MASK) {
			irq_settings_set_level(desc);
			irqd_set(&desc->irq_data, IRQD_LEVEL);
		}
699

700
		ret = 0;
701
		break;
702
	default:
703
		pr_err("Setting trigger mode %lu for irq %u failed (%pF)\n",
704
		       flags, irq_desc_get_irq(desc), chip->irq_set_type);
D
David Brownell 已提交
705
	}
706 707
	if (unmask)
		unmask_irq(desc);
708 709 710
	return ret;
}

711 712 713 714 715 716 717 718 719 720 721 722 723 724
#ifdef CONFIG_HARDIRQS_SW_RESEND
int irq_set_parent(int irq, int parent_irq)
{
	unsigned long flags;
	struct irq_desc *desc = irq_get_desc_lock(irq, &flags, 0);

	if (!desc)
		return -EINVAL;

	desc->parent_irq = parent_irq;

	irq_put_desc_unlock(desc, flags);
	return 0;
}
725
EXPORT_SYMBOL_GPL(irq_set_parent);
726 727
#endif

T
Thomas Gleixner 已提交
728 729 730 731 732 733 734 735 736 737
/*
 * Default primary interrupt handler for threaded interrupts. Is
 * assigned as primary handler when request_threaded_irq is called
 * with handler == NULL. Useful for oneshot interrupts.
 */
static irqreturn_t irq_default_primary_handler(int irq, void *dev_id)
{
	return IRQ_WAKE_THREAD;
}

738 739 740 741 742 743 744 745 746 747
/*
 * Primary handler for nested threaded interrupts. Should never be
 * called.
 */
static irqreturn_t irq_nested_primary_handler(int irq, void *dev_id)
{
	WARN(1, "Primary handler called for nested irq %d\n", irq);
	return IRQ_NONE;
}

748 749 750 751 752 753
static irqreturn_t irq_forced_secondary_handler(int irq, void *dev_id)
{
	WARN(1, "Secondary action handler called for irq %d\n", irq);
	return IRQ_NONE;
}

754 755
static int irq_wait_for_interrupt(struct irqaction *action)
{
756 757
	set_current_state(TASK_INTERRUPTIBLE);

758
	while (!kthread_should_stop()) {
759 760 761

		if (test_and_clear_bit(IRQTF_RUNTHREAD,
				       &action->thread_flags)) {
762 763
			__set_current_state(TASK_RUNNING);
			return 0;
764 765
		}
		schedule();
766
		set_current_state(TASK_INTERRUPTIBLE);
767
	}
768
	__set_current_state(TASK_RUNNING);
769 770 771
	return -1;
}

T
Thomas Gleixner 已提交
772 773 774 775 776
/*
 * Oneshot interrupts keep the irq line masked until the threaded
 * handler finished. unmask if the interrupt has not been disabled and
 * is marked MASKED.
 */
777
static void irq_finalize_oneshot(struct irq_desc *desc,
778
				 struct irqaction *action)
T
Thomas Gleixner 已提交
779
{
780 781
	if (!(desc->istate & IRQS_ONESHOT) ||
	    action->handler == irq_forced_secondary_handler)
782
		return;
783
again:
784
	chip_bus_lock(desc);
785
	raw_spin_lock_irq(&desc->lock);
786 787 788 789 790 791 792 793

	/*
	 * Implausible though it may be we need to protect us against
	 * the following scenario:
	 *
	 * The thread is faster done than the hard interrupt handler
	 * on the other CPU. If we unmask the irq line then the
	 * interrupt can come in again and masks the line, leaves due
794
	 * to IRQS_INPROGRESS and the irq line is masked forever.
795 796 797 798 799
	 *
	 * This also serializes the state of shared oneshot handlers
	 * versus "desc->threads_onehsot |= action->thread_mask;" in
	 * irq_wake_thread(). See the comment there which explains the
	 * serialization.
800
	 */
801
	if (unlikely(irqd_irq_inprogress(&desc->irq_data))) {
802
		raw_spin_unlock_irq(&desc->lock);
803
		chip_bus_sync_unlock(desc);
804 805 806 807
		cpu_relax();
		goto again;
	}

808 809 810 811 812
	/*
	 * Now check again, whether the thread should run. Otherwise
	 * we would clear the threads_oneshot bit of this thread which
	 * was just set.
	 */
813
	if (test_bit(IRQTF_RUNTHREAD, &action->thread_flags))
814 815 816 817
		goto out_unlock;

	desc->threads_oneshot &= ~action->thread_mask;

818 819
	if (!desc->threads_oneshot && !irqd_irq_disabled(&desc->irq_data) &&
	    irqd_irq_masked(&desc->irq_data))
820
		unmask_threaded_irq(desc);
821

822
out_unlock:
823
	raw_spin_unlock_irq(&desc->lock);
824
	chip_bus_sync_unlock(desc);
T
Thomas Gleixner 已提交
825 826
}

827
#ifdef CONFIG_SMP
828
/*
829
 * Check whether we need to change the affinity of the interrupt thread.
830 831 832 833 834
 */
static void
irq_thread_check_affinity(struct irq_desc *desc, struct irqaction *action)
{
	cpumask_var_t mask;
835
	bool valid = true;
836 837 838 839 840 841 842 843 844 845 846 847 848

	if (!test_and_clear_bit(IRQTF_AFFINITY, &action->thread_flags))
		return;

	/*
	 * In case we are out of memory we set IRQTF_AFFINITY again and
	 * try again next time
	 */
	if (!alloc_cpumask_var(&mask, GFP_KERNEL)) {
		set_bit(IRQTF_AFFINITY, &action->thread_flags);
		return;
	}

849
	raw_spin_lock_irq(&desc->lock);
850 851 852 853
	/*
	 * This code is triggered unconditionally. Check the affinity
	 * mask pointer. For CPU_MASK_OFFSTACK=n this is optimized out.
	 */
854
	if (cpumask_available(desc->irq_common_data.affinity))
855
		cpumask_copy(mask, desc->irq_common_data.affinity);
856 857
	else
		valid = false;
858
	raw_spin_unlock_irq(&desc->lock);
859

860 861
	if (valid)
		set_cpus_allowed_ptr(current, mask);
862 863
	free_cpumask_var(mask);
}
864 865 866 867
#else
static inline void
irq_thread_check_affinity(struct irq_desc *desc, struct irqaction *action) { }
#endif
868

869 870 871 872 873 874
/*
 * Interrupts which are not explicitely requested as threaded
 * interrupts rely on the implicit bh/preempt disable of the hard irq
 * context. So we need to disable bh here to avoid deadlocks and other
 * side effects.
 */
875
static irqreturn_t
876 877
irq_forced_thread_fn(struct irq_desc *desc, struct irqaction *action)
{
878 879
	irqreturn_t ret;

880
	local_bh_disable();
881
	ret = action->thread_fn(action->irq, action->dev_id);
882
	irq_finalize_oneshot(desc, action);
883
	local_bh_enable();
884
	return ret;
885 886 887
}

/*
888
 * Interrupts explicitly requested as threaded interrupts want to be
889 890 891
 * preemtible - many of them need to sleep and wait for slow busses to
 * complete.
 */
892 893
static irqreturn_t irq_thread_fn(struct irq_desc *desc,
		struct irqaction *action)
894
{
895 896 897
	irqreturn_t ret;

	ret = action->thread_fn(action->irq, action->dev_id);
898
	irq_finalize_oneshot(desc, action);
899
	return ret;
900 901
}

902 903
static void wake_threads_waitq(struct irq_desc *desc)
{
904
	if (atomic_dec_and_test(&desc->threads_active))
905 906 907
		wake_up(&desc->wait_for_threads);
}

908
static void irq_thread_dtor(struct callback_head *unused)
909 910 911 912 913 914 915 916 917 918
{
	struct task_struct *tsk = current;
	struct irq_desc *desc;
	struct irqaction *action;

	if (WARN_ON_ONCE(!(current->flags & PF_EXITING)))
		return;

	action = kthread_data(tsk);

919
	pr_err("exiting task \"%s\" (%d) is an active IRQ thread (irq %d)\n",
A
Alan Cox 已提交
920
	       tsk->comm, tsk->pid, action->irq);
921 922 923 924 925 926 927 928 929 930 931 932 933 934


	desc = irq_to_desc(action->irq);
	/*
	 * If IRQTF_RUNTHREAD is set, we need to decrement
	 * desc->threads_active and wake possible waiters.
	 */
	if (test_and_clear_bit(IRQTF_RUNTHREAD, &action->thread_flags))
		wake_threads_waitq(desc);

	/* Prevent a stale desc->threads_oneshot */
	irq_finalize_oneshot(desc, action);
}

935 936 937 938 939 940 941 942 943 944 945 946
static void irq_wake_secondary(struct irq_desc *desc, struct irqaction *action)
{
	struct irqaction *secondary = action->secondary;

	if (WARN_ON_ONCE(!secondary))
		return;

	raw_spin_lock_irq(&desc->lock);
	__irq_wake_thread(desc, secondary);
	raw_spin_unlock_irq(&desc->lock);
}

947 948 949 950 951
/*
 * Interrupt handler thread
 */
static int irq_thread(void *data)
{
952
	struct callback_head on_exit_work;
953 954
	struct irqaction *action = data;
	struct irq_desc *desc = irq_to_desc(action->irq);
955 956
	irqreturn_t (*handler_fn)(struct irq_desc *desc,
			struct irqaction *action);
957

958
	if (force_irqthreads && test_bit(IRQTF_FORCED_THREAD,
959 960 961 962 963
					&action->thread_flags))
		handler_fn = irq_forced_thread_fn;
	else
		handler_fn = irq_thread_fn;

A
Al Viro 已提交
964
	init_task_work(&on_exit_work, irq_thread_dtor);
965
	task_work_add(current, &on_exit_work, false);
966

967 968
	irq_thread_check_affinity(desc, action);

969
	while (!irq_wait_for_interrupt(action)) {
970
		irqreturn_t action_ret;
971

972 973
		irq_thread_check_affinity(desc, action);

974
		action_ret = handler_fn(desc, action);
975 976
		if (action_ret == IRQ_HANDLED)
			atomic_inc(&desc->threads_handled);
977 978
		if (action_ret == IRQ_WAKE_THREAD)
			irq_wake_secondary(desc, action);
979

980
		wake_threads_waitq(desc);
981 982
	}

983 984 985 986
	/*
	 * This is the regular exit path. __free_irq() is stopping the
	 * thread via kthread_stop() after calling
	 * synchronize_irq(). So neither IRQTF_RUNTHREAD nor the
987 988 989 990
	 * oneshot mask bit can be set. We cannot verify that as we
	 * cannot touch the oneshot mask at this point anymore as
	 * __setup_irq() might have given out currents thread_mask
	 * again.
991
	 */
992
	task_work_cancel(current, irq_thread_dtor);
993 994 995
	return 0;
}

996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011
/**
 *	irq_wake_thread - wake the irq thread for the action identified by dev_id
 *	@irq:		Interrupt line
 *	@dev_id:	Device identity for which the thread should be woken
 *
 */
void irq_wake_thread(unsigned int irq, void *dev_id)
{
	struct irq_desc *desc = irq_to_desc(irq);
	struct irqaction *action;
	unsigned long flags;

	if (!desc || WARN_ON(irq_settings_is_per_cpu_devid(desc)))
		return;

	raw_spin_lock_irqsave(&desc->lock, flags);
1012
	for_each_action_of_desc(desc, action) {
1013 1014 1015 1016 1017 1018 1019 1020 1021 1022
		if (action->dev_id == dev_id) {
			if (action->thread)
				__irq_wake_thread(desc, action);
			break;
		}
	}
	raw_spin_unlock_irqrestore(&desc->lock, flags);
}
EXPORT_SYMBOL_GPL(irq_wake_thread);

1023
static int irq_setup_forced_threading(struct irqaction *new)
1024 1025
{
	if (!force_irqthreads)
1026
		return 0;
1027
	if (new->flags & (IRQF_NO_THREAD | IRQF_PERCPU | IRQF_ONESHOT))
1028
		return 0;
1029 1030 1031

	new->flags |= IRQF_ONESHOT;

1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046
	/*
	 * Handle the case where we have a real primary handler and a
	 * thread handler. We force thread them as well by creating a
	 * secondary action.
	 */
	if (new->handler != irq_default_primary_handler && new->thread_fn) {
		/* Allocate the secondary action */
		new->secondary = kzalloc(sizeof(struct irqaction), GFP_KERNEL);
		if (!new->secondary)
			return -ENOMEM;
		new->secondary->handler = irq_forced_secondary_handler;
		new->secondary->thread_fn = new->thread_fn;
		new->secondary->dev_id = new->dev_id;
		new->secondary->irq = new->irq;
		new->secondary->name = new->name;
1047
	}
1048 1049 1050 1051 1052
	/* Deal with the primary handler */
	set_bit(IRQTF_FORCED_THREAD, &new->thread_flags);
	new->thread_fn = new->handler;
	new->handler = irq_default_primary_handler;
	return 0;
1053 1054
}

1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071
static int irq_request_resources(struct irq_desc *desc)
{
	struct irq_data *d = &desc->irq_data;
	struct irq_chip *c = d->chip;

	return c->irq_request_resources ? c->irq_request_resources(d) : 0;
}

static void irq_release_resources(struct irq_desc *desc)
{
	struct irq_data *d = &desc->irq_data;
	struct irq_chip *c = d->chip;

	if (c->irq_release_resources)
		c->irq_release_resources(d);
}

1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113
static int
setup_irq_thread(struct irqaction *new, unsigned int irq, bool secondary)
{
	struct task_struct *t;
	struct sched_param param = {
		.sched_priority = MAX_USER_RT_PRIO/2,
	};

	if (!secondary) {
		t = kthread_create(irq_thread, new, "irq/%d-%s", irq,
				   new->name);
	} else {
		t = kthread_create(irq_thread, new, "irq/%d-s-%s", irq,
				   new->name);
		param.sched_priority -= 1;
	}

	if (IS_ERR(t))
		return PTR_ERR(t);

	sched_setscheduler_nocheck(t, SCHED_FIFO, &param);

	/*
	 * We keep the reference to the task struct even if
	 * the thread dies to avoid that the interrupt code
	 * references an already freed task_struct.
	 */
	get_task_struct(t);
	new->thread = t;
	/*
	 * Tell the thread to set its affinity. This is
	 * important for shared interrupt handlers as we do
	 * not invoke setup_affinity() for the secondary
	 * handlers as everything is already set up. Even for
	 * interrupts marked with IRQF_NO_BALANCE this is
	 * correct as we want the thread to move to the cpu(s)
	 * on which the requesting code placed the interrupt.
	 */
	set_bit(IRQTF_AFFINITY, &new->thread_flags);
	return 0;
}

L
Linus Torvalds 已提交
1114 1115 1116
/*
 * Internal function to register an irqaction - typically used to
 * allocate special interrupts that are part of the architecture.
1117 1118 1119 1120 1121 1122 1123 1124 1125 1126
 *
 * Locking rules:
 *
 * desc->request_mutex	Provides serialization against a concurrent free_irq()
 *   chip_bus_lock	Provides serialization for slow bus operations
 *     desc->lock	Provides serialization against hard interrupts
 *
 * chip_bus_lock and desc->lock are sufficient for all other management and
 * interrupt related functions. desc->request_mutex solely serializes
 * request/free_irq().
L
Linus Torvalds 已提交
1127
 */
1128
static int
I
Ingo Molnar 已提交
1129
__setup_irq(unsigned int irq, struct irq_desc *desc, struct irqaction *new)
L
Linus Torvalds 已提交
1130
{
1131
	struct irqaction *old, **old_ptr;
1132
	unsigned long flags, thread_mask = 0;
1133
	int ret, nested, shared = 0;
L
Linus Torvalds 已提交
1134

1135
	if (!desc)
1136 1137
		return -EINVAL;

1138
	if (desc->irq_data.chip == &no_irq_chip)
L
Linus Torvalds 已提交
1139
		return -ENOSYS;
1140 1141
	if (!try_module_get(desc->owner))
		return -ENODEV;
L
Linus Torvalds 已提交
1142

1143 1144
	new->irq = irq;

1145 1146 1147 1148 1149 1150 1151
	/*
	 * If the trigger type is not specified by the caller,
	 * then use the default for this interrupt.
	 */
	if (!(new->flags & IRQF_TRIGGER_MASK))
		new->flags |= irqd_get_trigger_type(&desc->irq_data);

1152
	/*
1153 1154 1155
	 * Check whether the interrupt nests into another interrupt
	 * thread.
	 */
1156
	nested = irq_settings_is_nested_thread(desc);
1157
	if (nested) {
1158 1159 1160 1161
		if (!new->thread_fn) {
			ret = -EINVAL;
			goto out_mput;
		}
1162 1163 1164 1165 1166 1167
		/*
		 * Replace the primary handler which was provided from
		 * the driver for non nested interrupt handling by the
		 * dummy function which warns when called.
		 */
		new->handler = irq_nested_primary_handler;
1168
	} else {
1169 1170 1171 1172 1173
		if (irq_settings_can_thread(desc)) {
			ret = irq_setup_forced_threading(new);
			if (ret)
				goto out_mput;
		}
1174 1175
	}

1176
	/*
1177 1178 1179
	 * Create a handler thread when a thread function is supplied
	 * and the interrupt does not nest into another interrupt
	 * thread.
1180
	 */
1181
	if (new->thread_fn && !nested) {
1182 1183
		ret = setup_irq_thread(new, irq, false);
		if (ret)
1184
			goto out_mput;
1185 1186 1187 1188
		if (new->secondary) {
			ret = setup_irq_thread(new->secondary, irq, true);
			if (ret)
				goto out_thread;
1189
		}
1190 1191
	}

1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203
	/*
	 * Drivers are often written to work w/o knowledge about the
	 * underlying irq chip implementation, so a request for a
	 * threaded irq without a primary hard irq context handler
	 * requires the ONESHOT flag to be set. Some irq chips like
	 * MSI based interrupts are per se one shot safe. Check the
	 * chip flags, so we can avoid the unmask dance at the end of
	 * the threaded handler for those.
	 */
	if (desc->irq_data.chip->flags & IRQCHIP_ONESHOT_SAFE)
		new->flags &= ~IRQF_ONESHOT;

1204 1205 1206 1207 1208
	/*
	 * Protects against a concurrent __free_irq() call which might wait
	 * for synchronize_irq() to complete without holding the optional
	 * chip bus lock and desc->lock.
	 */
1209
	mutex_lock(&desc->request_mutex);
1210 1211 1212 1213 1214 1215 1216 1217 1218

	/*
	 * Acquire bus lock as the irq_request_resources() callback below
	 * might rely on the serialization or the magic power management
	 * functions which are abusing the irq_bus_lock() callback,
	 */
	chip_bus_lock(desc);

	/* First installed action requests resources. */
1219 1220 1221 1222 1223
	if (!desc->action) {
		ret = irq_request_resources(desc);
		if (ret) {
			pr_err("Failed to request resources for %s (irq %d) on irqchip %s\n",
			       new->name, irq, desc->irq_data.chip->name);
1224
			goto out_bus_unlock;
1225 1226
		}
	}
1227

L
Linus Torvalds 已提交
1228 1229
	/*
	 * The following block of code has to be executed atomically
1230 1231 1232
	 * protected against a concurrent interrupt and any of the other
	 * management calls which are not serialized via
	 * desc->request_mutex or the optional bus lock.
L
Linus Torvalds 已提交
1233
	 */
1234
	raw_spin_lock_irqsave(&desc->lock, flags);
1235 1236
	old_ptr = &desc->action;
	old = *old_ptr;
1237
	if (old) {
1238 1239 1240
		/*
		 * Can't share interrupts unless both agree to and are
		 * the same type (level, edge, polarity). So both flag
1241
		 * fields must have IRQF_SHARED set and the bits which
1242 1243
		 * set the trigger type must match. Also all must
		 * agree on ONESHOT.
1244
		 */
1245 1246
		unsigned int oldtype = irqd_get_trigger_type(&desc->irq_data);

1247
		if (!((old->flags & new->flags) & IRQF_SHARED) ||
1248
		    (oldtype != (new->flags & IRQF_TRIGGER_MASK)) ||
1249
		    ((old->flags ^ new->flags) & IRQF_ONESHOT))
1250 1251 1252
			goto mismatch;

		/* All handlers must agree on per-cpuness */
1253 1254
		if ((old->flags & IRQF_PERCPU) !=
		    (new->flags & IRQF_PERCPU))
1255
			goto mismatch;
L
Linus Torvalds 已提交
1256 1257 1258

		/* add new interrupt at end of irq queue */
		do {
1259 1260 1261 1262 1263
			/*
			 * Or all existing action->thread_mask bits,
			 * so we can find the next zero bit for this
			 * new action.
			 */
1264
			thread_mask |= old->thread_mask;
1265 1266
			old_ptr = &old->next;
			old = *old_ptr;
L
Linus Torvalds 已提交
1267 1268 1269 1270
		} while (old);
		shared = 1;
	}

1271
	/*
1272 1273 1274
	 * Setup the thread mask for this irqaction for ONESHOT. For
	 * !ONESHOT irqs the thread mask is 0 so we can avoid a
	 * conditional in irq_wake_thread().
1275
	 */
1276 1277 1278 1279 1280 1281 1282
	if (new->flags & IRQF_ONESHOT) {
		/*
		 * Unlikely to have 32 resp 64 irqs sharing one line,
		 * but who knows.
		 */
		if (thread_mask == ~0UL) {
			ret = -EBUSY;
1283
			goto out_unlock;
1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305
		}
		/*
		 * The thread_mask for the action is or'ed to
		 * desc->thread_active to indicate that the
		 * IRQF_ONESHOT thread handler has been woken, but not
		 * yet finished. The bit is cleared when a thread
		 * completes. When all threads of a shared interrupt
		 * line have completed desc->threads_active becomes
		 * zero and the interrupt line is unmasked. See
		 * handle.c:irq_wake_thread() for further information.
		 *
		 * If no thread is woken by primary (hard irq context)
		 * interrupt handlers, then desc->threads_active is
		 * also checked for zero to unmask the irq line in the
		 * affected hard irq flow handlers
		 * (handle_[fasteoi|level]_irq).
		 *
		 * The new action gets the first zero bit of
		 * thread_mask assigned. See the loop above which or's
		 * all existing action->thread_mask bits.
		 */
		new->thread_mask = 1 << ffz(thread_mask);
1306

1307 1308
	} else if (new->handler == irq_default_primary_handler &&
		   !(desc->irq_data.chip->flags & IRQCHIP_ONESHOT_SAFE)) {
1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323
		/*
		 * The interrupt was requested with handler = NULL, so
		 * we use the default primary handler for it. But it
		 * does not have the oneshot flag set. In combination
		 * with level interrupts this is deadly, because the
		 * default primary handler just wakes the thread, then
		 * the irq lines is reenabled, but the device still
		 * has the level irq asserted. Rinse and repeat....
		 *
		 * While this works for edge type interrupts, we play
		 * it safe and reject unconditionally because we can't
		 * say for sure which type this interrupt really
		 * has. The type flags are unreliable as the
		 * underlying chip implementation can override them.
		 */
1324
		pr_err("Threaded irq requested with handler=NULL and !ONESHOT for irq %d\n",
1325 1326
		       irq);
		ret = -EINVAL;
1327
		goto out_unlock;
1328 1329
	}

L
Linus Torvalds 已提交
1330
	if (!shared) {
1331 1332
		init_waitqueue_head(&desc->wait_for_threads);

1333
		/* Setup the type (level, edge polarity) if configured: */
1334
		if (new->flags & IRQF_TRIGGER_MASK) {
1335 1336
			ret = __irq_set_trigger(desc,
						new->flags & IRQF_TRIGGER_MASK);
1337

1338
			if (ret)
1339
				goto out_unlock;
1340
		}
T
Thomas Gleixner 已提交
1341

1342
		desc->istate &= ~(IRQS_AUTODETECT | IRQS_SPURIOUS_DISABLED | \
1343 1344
				  IRQS_ONESHOT | IRQS_WAITING);
		irqd_clear(&desc->irq_data, IRQD_IRQ_INPROGRESS);
1345

1346 1347 1348 1349
		if (new->flags & IRQF_PERCPU) {
			irqd_set(&desc->irq_data, IRQD_PER_CPU);
			irq_settings_set_per_cpu(desc);
		}
1350

T
Thomas Gleixner 已提交
1351
		if (new->flags & IRQF_ONESHOT)
1352
			desc->istate |= IRQS_ONESHOT;
T
Thomas Gleixner 已提交
1353

1354 1355 1356 1357 1358 1359
		/* Exclude IRQ from balancing if requested */
		if (new->flags & IRQF_NOBALANCING) {
			irq_settings_set_no_balancing(desc);
			irqd_set(&desc->irq_data, IRQD_NO_BALANCING);
		}

1360
		if (irq_settings_can_autoenable(desc)) {
1361
			irq_startup(desc, IRQ_RESEND, IRQ_START_COND);
1362 1363 1364 1365 1366 1367 1368 1369
		} else {
			/*
			 * Shared interrupts do not go well with disabling
			 * auto enable. The sharing interrupt might request
			 * it while it's still disabled and then wait for
			 * interrupts forever.
			 */
			WARN_ON_ONCE(new->flags & IRQF_SHARED);
1370 1371
			/* Undo nested disables: */
			desc->depth = 1;
1372
		}
1373

1374 1375
	} else if (new->flags & IRQF_TRIGGER_MASK) {
		unsigned int nmsk = new->flags & IRQF_TRIGGER_MASK;
1376
		unsigned int omsk = irqd_get_trigger_type(&desc->irq_data);
1377 1378 1379

		if (nmsk != omsk)
			/* hope the handler works with current  trigger mode */
1380
			pr_warn("irq %d uses trigger mode %u; requested %u\n",
1381
				irq, omsk, nmsk);
L
Linus Torvalds 已提交
1382
	}
1383

1384
	*old_ptr = new;
1385

1386 1387
	irq_pm_install_action(desc, new);

1388 1389 1390
	/* Reset broken irq detection when installing new handler */
	desc->irq_count = 0;
	desc->irqs_unhandled = 0;
1391 1392 1393 1394 1395

	/*
	 * Check whether we disabled the irq via the spurious handler
	 * before. Reenable it and give it another chance.
	 */
1396 1397
	if (shared && (desc->istate & IRQS_SPURIOUS_DISABLED)) {
		desc->istate &= ~IRQS_SPURIOUS_DISABLED;
1398
		__enable_irq(desc);
1399 1400
	}

1401
	raw_spin_unlock_irqrestore(&desc->lock, flags);
1402
	chip_bus_sync_unlock(desc);
1403
	mutex_unlock(&desc->request_mutex);
L
Linus Torvalds 已提交
1404

1405 1406
	irq_setup_timings(desc, new);

1407 1408 1409 1410 1411 1412
	/*
	 * Strictly no need to wake it up, but hung_task complains
	 * when no hard interrupt wakes the thread up.
	 */
	if (new->thread)
		wake_up_process(new->thread);
1413 1414
	if (new->secondary)
		wake_up_process(new->secondary->thread);
1415

1416
	register_irq_proc(irq, desc);
1417
	irq_add_debugfs_entry(irq, desc);
L
Linus Torvalds 已提交
1418 1419 1420
	new->dir = NULL;
	register_handler_proc(irq, new);
	return 0;
1421 1422

mismatch:
1423
	if (!(new->flags & IRQF_PROBE_SHARED)) {
1424
		pr_err("Flags mismatch irq %d. %08x (%s) vs. %08x (%s)\n",
1425 1426
		       irq, new->flags, new->name, old->flags, old->name);
#ifdef CONFIG_DEBUG_SHIRQ
1427
		dump_stack();
1428
#endif
1429
	}
1430 1431
	ret = -EBUSY;

1432
out_unlock:
1433
	raw_spin_unlock_irqrestore(&desc->lock, flags);
1434

1435 1436
	if (!desc->action)
		irq_release_resources(desc);
1437 1438
out_bus_unlock:
	chip_bus_sync_unlock(desc);
1439 1440
	mutex_unlock(&desc->request_mutex);

1441 1442 1443 1444 1445
out_thread:
	if (new->thread) {
		struct task_struct *t = new->thread;

		new->thread = NULL;
1446
		kthread_stop(t);
1447 1448
		put_task_struct(t);
	}
1449 1450 1451 1452 1453 1454 1455
	if (new->secondary && new->secondary->thread) {
		struct task_struct *t = new->secondary->thread;

		new->secondary->thread = NULL;
		kthread_stop(t);
		put_task_struct(t);
	}
1456 1457
out_mput:
	module_put(desc->owner);
1458
	return ret;
L
Linus Torvalds 已提交
1459 1460
}

1461 1462 1463 1464 1465 1466 1467 1468 1469
/**
 *	setup_irq - setup an interrupt
 *	@irq: Interrupt line to setup
 *	@act: irqaction for the interrupt
 *
 * Used to statically setup interrupts in the early boot process.
 */
int setup_irq(unsigned int irq, struct irqaction *act)
{
1470
	int retval;
1471 1472
	struct irq_desc *desc = irq_to_desc(irq);

1473
	if (!desc || WARN_ON(irq_settings_is_per_cpu_devid(desc)))
1474
		return -EINVAL;
1475 1476 1477 1478 1479

	retval = irq_chip_pm_get(&desc->irq_data);
	if (retval < 0)
		return retval;

1480 1481
	retval = __setup_irq(irq, desc, act);

1482 1483 1484
	if (retval)
		irq_chip_pm_put(&desc->irq_data);

1485
	return retval;
1486
}
1487
EXPORT_SYMBOL_GPL(setup_irq);
1488

1489
/*
1490 1491
 * Internal function to unregister an irqaction - used to free
 * regular and special interrupts that are part of the architecture.
L
Linus Torvalds 已提交
1492
 */
1493
static struct irqaction *__free_irq(unsigned int irq, void *dev_id)
L
Linus Torvalds 已提交
1494
{
1495
	struct irq_desc *desc = irq_to_desc(irq);
1496
	struct irqaction *action, **action_ptr;
L
Linus Torvalds 已提交
1497 1498
	unsigned long flags;

1499
	WARN(in_interrupt(), "Trying to free IRQ %d from IRQ context!\n", irq);
1500 1501

	if (!desc)
1502
		return NULL;
L
Linus Torvalds 已提交
1503

1504
	mutex_lock(&desc->request_mutex);
1505
	chip_bus_lock(desc);
1506
	raw_spin_lock_irqsave(&desc->lock, flags);
1507 1508 1509 1510 1511

	/*
	 * There can be multiple actions per IRQ descriptor, find the right
	 * one based on the dev_id:
	 */
1512
	action_ptr = &desc->action;
L
Linus Torvalds 已提交
1513
	for (;;) {
1514
		action = *action_ptr;
L
Linus Torvalds 已提交
1515

1516 1517
		if (!action) {
			WARN(1, "Trying to free already-free IRQ %d\n", irq);
1518
			raw_spin_unlock_irqrestore(&desc->lock, flags);
1519
			chip_bus_sync_unlock(desc);
1520
			mutex_unlock(&desc->request_mutex);
1521
			return NULL;
1522
		}
L
Linus Torvalds 已提交
1523

1524 1525
		if (action->dev_id == dev_id)
			break;
1526
		action_ptr = &action->next;
1527
	}
1528

1529
	/* Found it - now remove it from the list of entries: */
1530
	*action_ptr = action->next;
1531

1532 1533
	irq_pm_remove_action(desc, action);

1534
	/* If this was the last handler, shut down the IRQ line: */
1535
	if (!desc->action) {
1536
		irq_settings_clr_disable_unlazy(desc);
1537
		irq_shutdown(desc);
1538
	}
1539

1540 1541 1542 1543 1544 1545
#ifdef CONFIG_SMP
	/* make sure affinity_hint is cleaned up */
	if (WARN_ON_ONCE(desc->affinity_hint))
		desc->affinity_hint = NULL;
#endif

1546
	raw_spin_unlock_irqrestore(&desc->lock, flags);
1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560
	/*
	 * Drop bus_lock here so the changes which were done in the chip
	 * callbacks above are synced out to the irq chips which hang
	 * behind a slow bus (I2C, SPI) before calling synchronize_irq().
	 *
	 * Aside of that the bus_lock can also be taken from the threaded
	 * handler in irq_finalize_oneshot() which results in a deadlock
	 * because synchronize_irq() would wait forever for the thread to
	 * complete, which is blocked on the bus lock.
	 *
	 * The still held desc->request_mutex() protects against a
	 * concurrent request_irq() of this irq so the release of resources
	 * and timing data is properly serialized.
	 */
1561
	chip_bus_sync_unlock(desc);
1562 1563 1564 1565 1566

	unregister_handler_proc(irq, action);

	/* Make sure it's not being used on another CPU: */
	synchronize_irq(irq);
L
Linus Torvalds 已提交
1567

1568
#ifdef CONFIG_DEBUG_SHIRQ
1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580
	/*
	 * It's a shared IRQ -- the driver ought to be prepared for an IRQ
	 * event to happen even now it's being freed, so let's make sure that
	 * is so by doing an extra call to the handler ....
	 *
	 * ( We do this after actually deregistering it, to make sure that a
	 *   'real' IRQ doesn't run in * parallel with our fake. )
	 */
	if (action->flags & IRQF_SHARED) {
		local_irq_save(flags);
		action->handler(irq, dev_id);
		local_irq_restore(flags);
L
Linus Torvalds 已提交
1581
	}
1582
#endif
1583 1584

	if (action->thread) {
1585
		kthread_stop(action->thread);
1586
		put_task_struct(action->thread);
1587 1588 1589 1590
		if (action->secondary && action->secondary->thread) {
			kthread_stop(action->secondary->thread);
			put_task_struct(action->secondary->thread);
		}
1591 1592
	}

1593
	/* Last action releases resources */
1594
	if (!desc->action) {
1595 1596 1597 1598 1599
		/*
		 * Reaquire bus lock as irq_release_resources() might
		 * require it to deallocate resources over the slow bus.
		 */
		chip_bus_lock(desc);
1600
		irq_release_resources(desc);
1601
		chip_bus_sync_unlock(desc);
1602 1603
		irq_remove_timings(desc);
	}
1604

1605 1606
	mutex_unlock(&desc->request_mutex);

1607
	irq_chip_pm_put(&desc->irq_data);
1608
	module_put(desc->owner);
1609
	kfree(action->secondary);
1610 1611 1612
	return action;
}

1613 1614 1615 1616 1617 1618 1619 1620 1621
/**
 *	remove_irq - free an interrupt
 *	@irq: Interrupt line to free
 *	@act: irqaction for the interrupt
 *
 * Used to remove interrupts statically setup by the early boot process.
 */
void remove_irq(unsigned int irq, struct irqaction *act)
{
1622 1623 1624
	struct irq_desc *desc = irq_to_desc(irq);

	if (desc && !WARN_ON(irq_settings_is_per_cpu_devid(desc)))
1625
		__free_irq(irq, act->dev_id);
1626
}
1627
EXPORT_SYMBOL_GPL(remove_irq);
1628

1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641
/**
 *	free_irq - free an interrupt allocated with request_irq
 *	@irq: Interrupt line to free
 *	@dev_id: Device identity to free
 *
 *	Remove an interrupt handler. The handler is removed and if the
 *	interrupt line is no longer in use by any driver it is disabled.
 *	On a shared IRQ the caller must ensure the interrupt is disabled
 *	on the card it drives before calling this function. The function
 *	does not return until any executing interrupts for this IRQ
 *	have completed.
 *
 *	This function must not be called from interrupt context.
1642 1643
 *
 *	Returns the devname argument passed to request_irq.
1644
 */
1645
const void *free_irq(unsigned int irq, void *dev_id)
1646
{
T
Thomas Gleixner 已提交
1647
	struct irq_desc *desc = irq_to_desc(irq);
1648 1649
	struct irqaction *action;
	const char *devname;
T
Thomas Gleixner 已提交
1650

1651
	if (!desc || WARN_ON(irq_settings_is_per_cpu_devid(desc)))
1652
		return NULL;
T
Thomas Gleixner 已提交
1653

1654 1655 1656 1657 1658
#ifdef CONFIG_SMP
	if (WARN_ON(desc->affinity_notify))
		desc->affinity_notify = NULL;
#endif

1659
	action = __free_irq(irq, dev_id);
1660 1661 1662 1663

	if (!action)
		return NULL;

1664 1665 1666
	devname = action->name;
	kfree(action);
	return devname;
L
Linus Torvalds 已提交
1667 1668 1669 1670
}
EXPORT_SYMBOL(free_irq);

/**
1671
 *	request_threaded_irq - allocate an interrupt line
L
Linus Torvalds 已提交
1672
 *	@irq: Interrupt line to allocate
1673 1674
 *	@handler: Function to be called when the IRQ occurs.
 *		  Primary handler for threaded interrupts
T
Thomas Gleixner 已提交
1675 1676
 *		  If NULL and thread_fn != NULL the default
 *		  primary handler is installed
1677 1678
 *	@thread_fn: Function called from the irq handler thread
 *		    If NULL, no irq thread is created
L
Linus Torvalds 已提交
1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689
 *	@irqflags: Interrupt type flags
 *	@devname: An ascii name for the claiming device
 *	@dev_id: A cookie passed back to the handler function
 *
 *	This call allocates interrupt resources and enables the
 *	interrupt line and IRQ handling. From the point this
 *	call is made your handler function may be invoked. Since
 *	your handler function must clear any interrupt the board
 *	raises, you must take care both to initialise your hardware
 *	and to set up the interrupt handler in the right order.
 *
1690
 *	If you want to set up a threaded irq handler for your device
J
Javi Merino 已提交
1691
 *	then you need to supply @handler and @thread_fn. @handler is
1692 1693 1694
 *	still called in hard interrupt context and has to check
 *	whether the interrupt originates from the device. If yes it
 *	needs to disable the interrupt on the device and return
1695
 *	IRQ_WAKE_THREAD which will wake up the handler thread and run
1696 1697 1698
 *	@thread_fn. This split handler design is necessary to support
 *	shared interrupts.
 *
L
Linus Torvalds 已提交
1699 1700 1701 1702 1703 1704 1705 1706 1707
 *	Dev_id must be globally unique. Normally the address of the
 *	device data structure is used as the cookie. Since the handler
 *	receives this value it makes sense to use it.
 *
 *	If your interrupt is shared you must pass a non NULL dev_id
 *	as this is required when freeing the interrupt.
 *
 *	Flags:
 *
1708
 *	IRQF_SHARED		Interrupt is shared
D
David Brownell 已提交
1709
 *	IRQF_TRIGGER_*		Specify active edge(s) or level
L
Linus Torvalds 已提交
1710 1711
 *
 */
1712 1713 1714
int request_threaded_irq(unsigned int irq, irq_handler_t handler,
			 irq_handler_t thread_fn, unsigned long irqflags,
			 const char *devname, void *dev_id)
L
Linus Torvalds 已提交
1715
{
1716
	struct irqaction *action;
1717
	struct irq_desc *desc;
1718
	int retval;
L
Linus Torvalds 已提交
1719

1720 1721 1722
	if (irq == IRQ_NOTCONNECTED)
		return -ENOTCONN;

L
Linus Torvalds 已提交
1723 1724 1725 1726 1727
	/*
	 * Sanity-check: shared interrupts must pass in a real dev-ID,
	 * otherwise we'll have trouble later trying to figure out
	 * which interrupt is which (messes up the interrupt freeing
	 * logic etc).
1728 1729 1730
	 *
	 * Also IRQF_COND_SUSPEND only makes sense for shared interrupts and
	 * it cannot be set along with IRQF_NO_SUSPEND.
L
Linus Torvalds 已提交
1731
	 */
1732 1733 1734
	if (((irqflags & IRQF_SHARED) && !dev_id) ||
	    (!(irqflags & IRQF_SHARED) && (irqflags & IRQF_COND_SUSPEND)) ||
	    ((irqflags & IRQF_NO_SUSPEND) && (irqflags & IRQF_COND_SUSPEND)))
L
Linus Torvalds 已提交
1735
		return -EINVAL;
1736

1737
	desc = irq_to_desc(irq);
1738
	if (!desc)
L
Linus Torvalds 已提交
1739
		return -EINVAL;
1740

1741 1742
	if (!irq_settings_can_request(desc) ||
	    WARN_ON(irq_settings_is_per_cpu_devid(desc)))
1743
		return -EINVAL;
T
Thomas Gleixner 已提交
1744 1745 1746 1747 1748 1749

	if (!handler) {
		if (!thread_fn)
			return -EINVAL;
		handler = irq_default_primary_handler;
	}
L
Linus Torvalds 已提交
1750

1751
	action = kzalloc(sizeof(struct irqaction), GFP_KERNEL);
L
Linus Torvalds 已提交
1752 1753 1754 1755
	if (!action)
		return -ENOMEM;

	action->handler = handler;
1756
	action->thread_fn = thread_fn;
L
Linus Torvalds 已提交
1757 1758 1759 1760
	action->flags = irqflags;
	action->name = devname;
	action->dev_id = dev_id;

1761
	retval = irq_chip_pm_get(&desc->irq_data);
1762 1763
	if (retval < 0) {
		kfree(action);
1764
		return retval;
1765
	}
1766

1767
	retval = __setup_irq(irq, desc, action);
T
Thomas Gleixner 已提交
1768

1769
	if (retval) {
1770
		irq_chip_pm_put(&desc->irq_data);
1771
		kfree(action->secondary);
1772
		kfree(action);
1773
	}
1774

1775
#ifdef CONFIG_DEBUG_SHIRQ_FIXME
1776
	if (!retval && (irqflags & IRQF_SHARED)) {
D
David Woodhouse 已提交
1777 1778 1779
		/*
		 * It's a shared IRQ -- the driver ought to be prepared for it
		 * to happen immediately, so let's make sure....
1780 1781
		 * We disable the irq to make sure that a 'real' IRQ doesn't
		 * run in parallel with our fake.
D
David Woodhouse 已提交
1782
		 */
1783
		unsigned long flags;
D
David Woodhouse 已提交
1784

1785
		disable_irq(irq);
1786
		local_irq_save(flags);
1787

1788
		handler(irq, dev_id);
1789

1790
		local_irq_restore(flags);
1791
		enable_irq(irq);
D
David Woodhouse 已提交
1792 1793
	}
#endif
L
Linus Torvalds 已提交
1794 1795
	return retval;
}
1796
EXPORT_SYMBOL(request_threaded_irq);
1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817

/**
 *	request_any_context_irq - allocate an interrupt line
 *	@irq: Interrupt line to allocate
 *	@handler: Function to be called when the IRQ occurs.
 *		  Threaded handler for threaded interrupts.
 *	@flags: Interrupt type flags
 *	@name: An ascii name for the claiming device
 *	@dev_id: A cookie passed back to the handler function
 *
 *	This call allocates interrupt resources and enables the
 *	interrupt line and IRQ handling. It selects either a
 *	hardirq or threaded handling method depending on the
 *	context.
 *
 *	On failure, it returns a negative value. On success,
 *	it returns either IRQC_IS_HARDIRQ or IRQC_IS_NESTED.
 */
int request_any_context_irq(unsigned int irq, irq_handler_t handler,
			    unsigned long flags, const char *name, void *dev_id)
{
1818
	struct irq_desc *desc;
1819 1820
	int ret;

1821 1822 1823 1824
	if (irq == IRQ_NOTCONNECTED)
		return -ENOTCONN;

	desc = irq_to_desc(irq);
1825 1826 1827
	if (!desc)
		return -EINVAL;

1828
	if (irq_settings_is_nested_thread(desc)) {
1829 1830 1831 1832 1833 1834 1835 1836 1837
		ret = request_threaded_irq(irq, NULL, handler,
					   flags, name, dev_id);
		return !ret ? IRQC_IS_NESTED : ret;
	}

	ret = request_irq(irq, handler, flags, name, dev_id);
	return !ret ? IRQC_IS_HARDIRQ : ret;
}
EXPORT_SYMBOL_GPL(request_any_context_irq);
1838

1839
void enable_percpu_irq(unsigned int irq, unsigned int type)
1840 1841 1842 1843 1844 1845 1846 1847
{
	unsigned int cpu = smp_processor_id();
	unsigned long flags;
	struct irq_desc *desc = irq_get_desc_lock(irq, &flags, IRQ_GET_DESC_CHECK_PERCPU);

	if (!desc)
		return;

1848 1849 1850 1851
	/*
	 * If the trigger type is not specified by the caller, then
	 * use the default for this interrupt.
	 */
1852
	type &= IRQ_TYPE_SENSE_MASK;
1853 1854 1855
	if (type == IRQ_TYPE_NONE)
		type = irqd_get_trigger_type(&desc->irq_data);

1856 1857 1858
	if (type != IRQ_TYPE_NONE) {
		int ret;

1859
		ret = __irq_set_trigger(desc, type);
1860 1861

		if (ret) {
1862
			WARN(1, "failed to set type for IRQ%d\n", irq);
1863 1864 1865 1866
			goto out;
		}
	}

1867
	irq_percpu_enable(desc, cpu);
1868
out:
1869 1870
	irq_put_desc_unlock(desc, flags);
}
1871
EXPORT_SYMBOL_GPL(enable_percpu_irq);
1872

1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897
/**
 * irq_percpu_is_enabled - Check whether the per cpu irq is enabled
 * @irq:	Linux irq number to check for
 *
 * Must be called from a non migratable context. Returns the enable
 * state of a per cpu interrupt on the current cpu.
 */
bool irq_percpu_is_enabled(unsigned int irq)
{
	unsigned int cpu = smp_processor_id();
	struct irq_desc *desc;
	unsigned long flags;
	bool is_enabled;

	desc = irq_get_desc_lock(irq, &flags, IRQ_GET_DESC_CHECK_PERCPU);
	if (!desc)
		return false;

	is_enabled = cpumask_test_cpu(cpu, desc->percpu_enabled);
	irq_put_desc_unlock(desc, flags);

	return is_enabled;
}
EXPORT_SYMBOL_GPL(irq_percpu_is_enabled);

1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909
void disable_percpu_irq(unsigned int irq)
{
	unsigned int cpu = smp_processor_id();
	unsigned long flags;
	struct irq_desc *desc = irq_get_desc_lock(irq, &flags, IRQ_GET_DESC_CHECK_PERCPU);

	if (!desc)
		return;

	irq_percpu_disable(desc, cpu);
	irq_put_desc_unlock(desc, flags);
}
1910
EXPORT_SYMBOL_GPL(disable_percpu_irq);
1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946

/*
 * Internal function to unregister a percpu irqaction.
 */
static struct irqaction *__free_percpu_irq(unsigned int irq, void __percpu *dev_id)
{
	struct irq_desc *desc = irq_to_desc(irq);
	struct irqaction *action;
	unsigned long flags;

	WARN(in_interrupt(), "Trying to free IRQ %d from IRQ context!\n", irq);

	if (!desc)
		return NULL;

	raw_spin_lock_irqsave(&desc->lock, flags);

	action = desc->action;
	if (!action || action->percpu_dev_id != dev_id) {
		WARN(1, "Trying to free already-free IRQ %d\n", irq);
		goto bad;
	}

	if (!cpumask_empty(desc->percpu_enabled)) {
		WARN(1, "percpu IRQ %d still enabled on CPU%d!\n",
		     irq, cpumask_first(desc->percpu_enabled));
		goto bad;
	}

	/* Found it - now remove it from the list of entries: */
	desc->action = NULL;

	raw_spin_unlock_irqrestore(&desc->lock, flags);

	unregister_handler_proc(irq, action);

1947
	irq_chip_pm_put(&desc->irq_data);
1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993
	module_put(desc->owner);
	return action;

bad:
	raw_spin_unlock_irqrestore(&desc->lock, flags);
	return NULL;
}

/**
 *	remove_percpu_irq - free a per-cpu interrupt
 *	@irq: Interrupt line to free
 *	@act: irqaction for the interrupt
 *
 * Used to remove interrupts statically setup by the early boot process.
 */
void remove_percpu_irq(unsigned int irq, struct irqaction *act)
{
	struct irq_desc *desc = irq_to_desc(irq);

	if (desc && irq_settings_is_per_cpu_devid(desc))
	    __free_percpu_irq(irq, act->percpu_dev_id);
}

/**
 *	free_percpu_irq - free an interrupt allocated with request_percpu_irq
 *	@irq: Interrupt line to free
 *	@dev_id: Device identity to free
 *
 *	Remove a percpu interrupt handler. The handler is removed, but
 *	the interrupt line is not disabled. This must be done on each
 *	CPU before calling this function. The function does not return
 *	until any executing interrupts for this IRQ have completed.
 *
 *	This function must not be called from interrupt context.
 */
void free_percpu_irq(unsigned int irq, void __percpu *dev_id)
{
	struct irq_desc *desc = irq_to_desc(irq);

	if (!desc || !irq_settings_is_per_cpu_devid(desc))
		return;

	chip_bus_lock(desc);
	kfree(__free_percpu_irq(irq, dev_id));
	chip_bus_sync_unlock(desc);
}
1994
EXPORT_SYMBOL_GPL(free_percpu_irq);
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009

/**
 *	setup_percpu_irq - setup a per-cpu interrupt
 *	@irq: Interrupt line to setup
 *	@act: irqaction for the interrupt
 *
 * Used to statically setup per-cpu interrupts in the early boot process.
 */
int setup_percpu_irq(unsigned int irq, struct irqaction *act)
{
	struct irq_desc *desc = irq_to_desc(irq);
	int retval;

	if (!desc || !irq_settings_is_per_cpu_devid(desc))
		return -EINVAL;
2010 2011 2012 2013 2014

	retval = irq_chip_pm_get(&desc->irq_data);
	if (retval < 0)
		return retval;

2015 2016
	retval = __setup_irq(irq, desc, act);

2017 2018 2019
	if (retval)
		irq_chip_pm_put(&desc->irq_data);

2020 2021 2022 2023
	return retval;
}

/**
2024
 *	__request_percpu_irq - allocate a percpu interrupt line
2025 2026
 *	@irq: Interrupt line to allocate
 *	@handler: Function to be called when the IRQ occurs.
2027
 *	@flags: Interrupt type flags (IRQF_TIMER only)
2028 2029 2030
 *	@devname: An ascii name for the claiming device
 *	@dev_id: A percpu cookie passed back to the handler function
 *
2031 2032 2033 2034
 *	This call allocates interrupt resources and enables the
 *	interrupt on the local CPU. If the interrupt is supposed to be
 *	enabled on other CPUs, it has to be done on each CPU using
 *	enable_percpu_irq().
2035 2036 2037 2038 2039
 *
 *	Dev_id must be globally unique. It is a per-cpu variable, and
 *	the handler gets called with the interrupted CPU's instance of
 *	that variable.
 */
2040 2041 2042
int __request_percpu_irq(unsigned int irq, irq_handler_t handler,
			 unsigned long flags, const char *devname,
			 void __percpu *dev_id)
2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055
{
	struct irqaction *action;
	struct irq_desc *desc;
	int retval;

	if (!dev_id)
		return -EINVAL;

	desc = irq_to_desc(irq);
	if (!desc || !irq_settings_can_request(desc) ||
	    !irq_settings_is_per_cpu_devid(desc))
		return -EINVAL;

2056 2057 2058
	if (flags && flags != IRQF_TIMER)
		return -EINVAL;

2059 2060 2061 2062 2063
	action = kzalloc(sizeof(struct irqaction), GFP_KERNEL);
	if (!action)
		return -ENOMEM;

	action->handler = handler;
2064
	action->flags = flags | IRQF_PERCPU | IRQF_NO_SUSPEND;
2065 2066 2067
	action->name = devname;
	action->percpu_dev_id = dev_id;

2068
	retval = irq_chip_pm_get(&desc->irq_data);
2069 2070
	if (retval < 0) {
		kfree(action);
2071
		return retval;
2072
	}
2073

2074 2075
	retval = __setup_irq(irq, desc, action);

2076 2077
	if (retval) {
		irq_chip_pm_put(&desc->irq_data);
2078
		kfree(action);
2079
	}
2080 2081 2082

	return retval;
}
2083
EXPORT_SYMBOL_GPL(__request_percpu_irq);
2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129

/**
 *	irq_get_irqchip_state - returns the irqchip state of a interrupt.
 *	@irq: Interrupt line that is forwarded to a VM
 *	@which: One of IRQCHIP_STATE_* the caller wants to know about
 *	@state: a pointer to a boolean where the state is to be storeed
 *
 *	This call snapshots the internal irqchip state of an
 *	interrupt, returning into @state the bit corresponding to
 *	stage @which
 *
 *	This function should be called with preemption disabled if the
 *	interrupt controller has per-cpu registers.
 */
int irq_get_irqchip_state(unsigned int irq, enum irqchip_irq_state which,
			  bool *state)
{
	struct irq_desc *desc;
	struct irq_data *data;
	struct irq_chip *chip;
	unsigned long flags;
	int err = -EINVAL;

	desc = irq_get_desc_buslock(irq, &flags, 0);
	if (!desc)
		return err;

	data = irq_desc_get_irq_data(desc);

	do {
		chip = irq_data_get_irq_chip(data);
		if (chip->irq_get_irqchip_state)
			break;
#ifdef CONFIG_IRQ_DOMAIN_HIERARCHY
		data = data->parent_data;
#else
		data = NULL;
#endif
	} while (data);

	if (data)
		err = chip->irq_get_irqchip_state(data, which, state);

	irq_put_desc_busunlock(desc, flags);
	return err;
}
2130
EXPORT_SYMBOL_GPL(irq_get_irqchip_state);
2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175

/**
 *	irq_set_irqchip_state - set the state of a forwarded interrupt.
 *	@irq: Interrupt line that is forwarded to a VM
 *	@which: State to be restored (one of IRQCHIP_STATE_*)
 *	@val: Value corresponding to @which
 *
 *	This call sets the internal irqchip state of an interrupt,
 *	depending on the value of @which.
 *
 *	This function should be called with preemption disabled if the
 *	interrupt controller has per-cpu registers.
 */
int irq_set_irqchip_state(unsigned int irq, enum irqchip_irq_state which,
			  bool val)
{
	struct irq_desc *desc;
	struct irq_data *data;
	struct irq_chip *chip;
	unsigned long flags;
	int err = -EINVAL;

	desc = irq_get_desc_buslock(irq, &flags, 0);
	if (!desc)
		return err;

	data = irq_desc_get_irq_data(desc);

	do {
		chip = irq_data_get_irq_chip(data);
		if (chip->irq_set_irqchip_state)
			break;
#ifdef CONFIG_IRQ_DOMAIN_HIERARCHY
		data = data->parent_data;
#else
		data = NULL;
#endif
	} while (data);

	if (data)
		err = chip->irq_set_irqchip_state(data, which, val);

	irq_put_desc_busunlock(desc, flags);
	return err;
}
2176
EXPORT_SYMBOL_GPL(irq_set_irqchip_state);