manage.c 56.9 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
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;

178
	ret = chip->irq_set_affinity(data, mask, force);
179 180
	switch (ret) {
	case IRQ_SET_MASK_OK:
181
	case IRQ_SET_MASK_OK_DONE:
182
		cpumask_copy(desc->irq_common_data.affinity, mask);
183 184 185 186 187 188 189 190
	case IRQ_SET_MASK_OK_NOCOPY:
		irq_set_thread_affinity(desc);
		ret = 0;
	}

	return ret;
}

191 192
int irq_set_affinity_locked(struct irq_data *data, const struct cpumask *mask,
			    bool force)
193
{
194 195
	struct irq_chip *chip = irq_data_get_irq_chip(data);
	struct irq_desc *desc = irq_data_to_desc(data);
196
	int ret = 0;
197

198
	if (!chip || !chip->irq_set_affinity)
199 200
		return -EINVAL;

T
Thomas Gleixner 已提交
201
	if (irq_can_move_pcntxt(data)) {
202
		ret = irq_do_set_affinity(data, mask, force);
203
	} else {
204
		irqd_set_move_pending(data);
205
		irq_copy_pending(desc, mask);
206
	}
207

208 209 210 211
	if (desc->affinity_notify) {
		kref_get(&desc->affinity_notify->kref);
		schedule_work(&desc->affinity_notify->work);
	}
212 213 214 215 216
	irqd_set(data, IRQD_AFFINITY_SET);

	return ret;
}

217
int __irq_set_affinity(unsigned int irq, const struct cpumask *mask, bool force)
218 219 220 221 222 223 224 225 226
{
	struct irq_desc *desc = irq_to_desc(irq);
	unsigned long flags;
	int ret;

	if (!desc)
		return -EINVAL;

	raw_spin_lock_irqsave(&desc->lock, flags);
227
	ret = irq_set_affinity_locked(irq_desc_get_irq_data(desc), mask, force);
228
	raw_spin_unlock_irqrestore(&desc->lock, flags);
229
	return ret;
230 231
}

232 233 234
int irq_set_affinity_hint(unsigned int irq, const struct cpumask *m)
{
	unsigned long flags;
235
	struct irq_desc *desc = irq_get_desc_lock(irq, &flags, IRQ_GET_DESC_CHECK_GLOBAL);
236 237 238 239

	if (!desc)
		return -EINVAL;
	desc->affinity_hint = m;
240
	irq_put_desc_unlock(desc, flags);
241
	/* set the initial affinity to prevent every interrupt being on CPU0 */
242 243
	if (m)
		__irq_set_affinity(irq, m, false);
244 245 246 247
	return 0;
}
EXPORT_SYMBOL_GPL(irq_set_affinity_hint);

248 249 250 251 252 253 254 255
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;

256
	if (!desc || !alloc_cpumask_var(&cpumask, GFP_KERNEL))
257 258 259
		goto out;

	raw_spin_lock_irqsave(&desc->lock, flags);
T
Thomas Gleixner 已提交
260
	if (irq_move_pending(&desc->irq_data))
261
		irq_get_pending(cpumask, desc);
262
	else
263
		cpumask_copy(cpumask, desc->irq_common_data.affinity);
264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315
	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);

316 317 318 319
#ifndef CONFIG_AUTO_IRQ_AFFINITY
/*
 * Generic version of the affinity autoselector.
 */
320
int irq_setup_affinity(struct irq_desc *desc)
321
{
322
	struct cpumask *set = irq_default_affinity;
323 324 325
	int ret, node = irq_desc_get_node(desc);
	static DEFINE_RAW_SPINLOCK(mask_lock);
	static struct cpumask mask;
326

327
	/* Excludes PER_CPU and NO_BALANCE interrupts */
328
	if (!__irq_can_set_affinity(desc))
329 330
		return 0;

331
	raw_spin_lock(&mask_lock);
332
	/*
333
	 * Preserve the managed affinity setting and a userspace affinity
334
	 * setup, but make sure that one of the targets is online.
335
	 */
336 337
	if (irqd_affinity_is_managed(&desc->irq_data) ||
	    irqd_has_set(&desc->irq_data, IRQD_AFFINITY_SET)) {
338
		if (cpumask_intersects(desc->irq_common_data.affinity,
339
				       cpu_online_mask))
340
			set = desc->irq_common_data.affinity;
T
Thomas Gleixner 已提交
341
		else
342
			irqd_clear(&desc->irq_data, IRQD_AFFINITY_SET);
343
	}
344

345
	cpumask_and(&mask, cpu_online_mask, set);
346 347 348 349
	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 */
350 351
		if (cpumask_intersects(&mask, nodemask))
			cpumask_and(&mask, &mask, nodemask);
352
	}
353 354 355
	ret = irq_do_set_affinity(&desc->irq_data, &mask, false);
	raw_spin_unlock(&mask_lock);
	return ret;
356
}
357
#else
358
/* Wrapper for ALPHA specific affinity selector magic */
359
int irq_setup_affinity(struct irq_desc *desc)
360
{
361
	return irq_select_affinity(irq_desc_get_irq(desc));
362
}
363 364
#endif

365
/*
366
 * Called when a bogus affinity is set via /proc/irq
367
 */
368
int irq_select_affinity_usr(unsigned int irq)
369 370 371 372 373
{
	struct irq_desc *desc = irq_to_desc(irq);
	unsigned long flags;
	int ret;

374
	raw_spin_lock_irqsave(&desc->lock, flags);
375
	ret = irq_setup_affinity(desc);
376
	raw_spin_unlock_irqrestore(&desc->lock, flags);
377 378
	return ret;
}
L
Linus Torvalds 已提交
379 380
#endif

381 382 383
/**
 *	irq_set_vcpu_affinity - Set vcpu affinity for the interrupt
 *	@irq: interrupt number to set affinity
384 385
 *	@vcpu_info: vCPU specific data or pointer to a percpu array of vCPU
 *	            specific data for percpu_devid interrupts
386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403
 *
 *	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);
404 405 406 407 408 409 410 411 412 413 414 415
	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)
416 417 418 419 420 421 422
		ret = chip->irq_set_vcpu_affinity(data, vcpu_info);
	irq_put_desc_unlock(desc, flags);

	return ret;
}
EXPORT_SYMBOL_GPL(irq_set_vcpu_affinity);

423
void __disable_irq(struct irq_desc *desc)
424
{
425
	if (!desc->depth++)
426
		irq_disable(desc);
427 428
}

429 430 431
static int __disable_irq_nosync(unsigned int irq)
{
	unsigned long flags;
432
	struct irq_desc *desc = irq_get_desc_buslock(irq, &flags, IRQ_GET_DESC_CHECK_GLOBAL);
433 434 435

	if (!desc)
		return -EINVAL;
436
	__disable_irq(desc);
437 438 439 440
	irq_put_desc_busunlock(desc, flags);
	return 0;
}

L
Linus Torvalds 已提交
441 442 443 444 445 446 447 448 449 450 451 452 453
/**
 *	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)
{
454
	__disable_irq_nosync(irq);
L
Linus Torvalds 已提交
455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471
}
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)
{
472
	if (!__disable_irq_nosync(irq))
L
Linus Torvalds 已提交
473 474 475 476
		synchronize_irq(irq);
}
EXPORT_SYMBOL(disable_irq);

477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502
/**
 *	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);

503
void __enable_irq(struct irq_desc *desc)
504 505 506
{
	switch (desc->depth) {
	case 0:
507
 err_out:
508 509
		WARN(1, KERN_WARNING "Unbalanced enable for IRQ %d\n",
		     irq_desc_get_irq(desc));
510 511
		break;
	case 1: {
512
		if (desc->istate & IRQS_SUSPENDED)
513
			goto err_out;
514
		/* Prevent probing on this irq: */
515
		irq_settings_set_noprobe(desc);
516 517 518 519 520 521 522
		/*
		 * 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.
		 */
523
		irq_startup(desc, IRQ_RESEND, IRQ_START_FORCE);
524
		break;
525 526 527 528 529 530
	}
	default:
		desc->depth--;
	}
}

L
Linus Torvalds 已提交
531 532 533 534 535 536 537 538
/**
 *	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
539
 *	This function may be called from IRQ context only when
540
 *	desc->irq_data.chip->bus_lock and desc->chip->bus_sync_unlock are NULL !
L
Linus Torvalds 已提交
541 542 543 544
 */
void enable_irq(unsigned int irq)
{
	unsigned long flags;
545
	struct irq_desc *desc = irq_get_desc_buslock(irq, &flags, IRQ_GET_DESC_CHECK_GLOBAL);
L
Linus Torvalds 已提交
546

547
	if (!desc)
548
		return;
T
Thomas Gleixner 已提交
549 550
	if (WARN(!desc->irq_data.chip,
		 KERN_ERR "enable_irq before setup/request_irq: irq %u\n", irq))
551
		goto out;
552

553
	__enable_irq(desc);
554 555
out:
	irq_put_desc_busunlock(desc, flags);
L
Linus Torvalds 已提交
556 557 558
}
EXPORT_SYMBOL(enable_irq);

D
David Brownell 已提交
559
static int set_irq_wake_real(unsigned int irq, unsigned int on)
560
{
561
	struct irq_desc *desc = irq_to_desc(irq);
562 563
	int ret = -ENXIO;

564 565 566
	if (irq_desc_get_chip(desc)->flags &  IRQCHIP_SKIP_SET_WAKE)
		return 0;

567 568
	if (desc->irq_data.chip->irq_set_wake)
		ret = desc->irq_data.chip->irq_set_wake(&desc->irq_data, on);
569 570 571 572

	return ret;
}

573
/**
T
Thomas Gleixner 已提交
574
 *	irq_set_irq_wake - control irq power management wakeup
575 576 577
 *	@irq:	interrupt to control
 *	@on:	enable/disable power management wakeup
 *
578 579 580 581 582 583
 *	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".
584
 */
T
Thomas Gleixner 已提交
585
int irq_set_irq_wake(unsigned int irq, unsigned int on)
586 587
{
	unsigned long flags;
588
	struct irq_desc *desc = irq_get_desc_buslock(irq, &flags, IRQ_GET_DESC_CHECK_GLOBAL);
589
	int ret = 0;
590

591 592 593
	if (!desc)
		return -EINVAL;

594 595 596 597
	/* wakeup-capable irqs can be shared between drivers that
	 * don't need to have the same sleep mode behaviors.
	 */
	if (on) {
598 599 600 601 602
		if (desc->wake_depth++ == 0) {
			ret = set_irq_wake_real(irq, on);
			if (ret)
				desc->wake_depth = 0;
			else
603
				irqd_set(&desc->irq_data, IRQD_WAKEUP_STATE);
604
		}
605 606
	} else {
		if (desc->wake_depth == 0) {
607
			WARN(1, "Unbalanced IRQ %d wake disable\n", irq);
608 609 610 611 612
		} else if (--desc->wake_depth == 0) {
			ret = set_irq_wake_real(irq, on);
			if (ret)
				desc->wake_depth = 1;
			else
613
				irqd_clear(&desc->irq_data, IRQD_WAKEUP_STATE);
614
		}
615
	}
616
	irq_put_desc_busunlock(desc, flags);
617 618
	return ret;
}
T
Thomas Gleixner 已提交
619
EXPORT_SYMBOL(irq_set_irq_wake);
620

L
Linus Torvalds 已提交
621 622 623 624 625 626 627
/*
 * 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)
{
628
	unsigned long flags;
629
	struct irq_desc *desc = irq_get_desc_lock(irq, &flags, 0);
630
	int canrequest = 0;
L
Linus Torvalds 已提交
631

632 633 634
	if (!desc)
		return 0;

635
	if (irq_settings_can_request(desc)) {
636 637 638
		if (!desc->action ||
		    irqflags & desc->action->flags & IRQF_SHARED)
			canrequest = 1;
639 640 641
	}
	irq_put_desc_unlock(desc, flags);
	return canrequest;
L
Linus Torvalds 已提交
642 643
}

644
int __irq_set_trigger(struct irq_desc *desc, unsigned long flags)
645
{
646
	struct irq_chip *chip = desc->irq_data.chip;
647
	int ret, unmask = 0;
648

649
	if (!chip || !chip->irq_set_type) {
650 651 652 653
		/*
		 * IRQF_TRIGGER_* but the PIC does not support multiple
		 * flow-types?
		 */
654 655
		pr_debug("No set_type function for IRQ %d (%s)\n",
			 irq_desc_get_irq(desc),
656
			 chip ? (chip->name ? : "unknown") : "unknown");
657 658 659
		return 0;
	}

660
	if (chip->flags & IRQCHIP_SET_TYPE_MASKED) {
661
		if (!irqd_irq_masked(&desc->irq_data))
662
			mask_irq(desc);
663
		if (!irqd_irq_disabled(&desc->irq_data))
664 665 666
			unmask = 1;
	}

667 668
	/* Mask all flags except trigger mode */
	flags &= IRQ_TYPE_SENSE_MASK;
669
	ret = chip->irq_set_type(&desc->irq_data, flags);
670

671 672
	switch (ret) {
	case IRQ_SET_MASK_OK:
673
	case IRQ_SET_MASK_OK_DONE:
674 675 676 677 678 679 680 681 682 683 684 685
		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);
		}
686

687
		ret = 0;
688
		break;
689
	default:
690
		pr_err("Setting trigger mode %lu for irq %u failed (%pF)\n",
691
		       flags, irq_desc_get_irq(desc), chip->irq_set_type);
D
David Brownell 已提交
692
	}
693 694
	if (unmask)
		unmask_irq(desc);
695 696 697
	return ret;
}

698 699 700 701 702 703 704 705 706 707 708 709 710 711
#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;
}
712
EXPORT_SYMBOL_GPL(irq_set_parent);
713 714
#endif

T
Thomas Gleixner 已提交
715 716 717 718 719 720 721 722 723 724
/*
 * 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;
}

725 726 727 728 729 730 731 732 733 734
/*
 * 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;
}

735 736 737 738 739 740
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;
}

741 742
static int irq_wait_for_interrupt(struct irqaction *action)
{
743 744
	set_current_state(TASK_INTERRUPTIBLE);

745
	while (!kthread_should_stop()) {
746 747 748

		if (test_and_clear_bit(IRQTF_RUNTHREAD,
				       &action->thread_flags)) {
749 750
			__set_current_state(TASK_RUNNING);
			return 0;
751 752
		}
		schedule();
753
		set_current_state(TASK_INTERRUPTIBLE);
754
	}
755
	__set_current_state(TASK_RUNNING);
756 757 758
	return -1;
}

T
Thomas Gleixner 已提交
759 760 761 762 763
/*
 * Oneshot interrupts keep the irq line masked until the threaded
 * handler finished. unmask if the interrupt has not been disabled and
 * is marked MASKED.
 */
764
static void irq_finalize_oneshot(struct irq_desc *desc,
765
				 struct irqaction *action)
T
Thomas Gleixner 已提交
766
{
767 768
	if (!(desc->istate & IRQS_ONESHOT) ||
	    action->handler == irq_forced_secondary_handler)
769
		return;
770
again:
771
	chip_bus_lock(desc);
772
	raw_spin_lock_irq(&desc->lock);
773 774 775 776 777 778 779 780

	/*
	 * 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
781
	 * to IRQS_INPROGRESS and the irq line is masked forever.
782 783 784 785 786
	 *
	 * 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.
787
	 */
788
	if (unlikely(irqd_irq_inprogress(&desc->irq_data))) {
789
		raw_spin_unlock_irq(&desc->lock);
790
		chip_bus_sync_unlock(desc);
791 792 793 794
		cpu_relax();
		goto again;
	}

795 796 797 798 799
	/*
	 * Now check again, whether the thread should run. Otherwise
	 * we would clear the threads_oneshot bit of this thread which
	 * was just set.
	 */
800
	if (test_bit(IRQTF_RUNTHREAD, &action->thread_flags))
801 802 803 804
		goto out_unlock;

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

805 806
	if (!desc->threads_oneshot && !irqd_irq_disabled(&desc->irq_data) &&
	    irqd_irq_masked(&desc->irq_data))
807
		unmask_threaded_irq(desc);
808

809
out_unlock:
810
	raw_spin_unlock_irq(&desc->lock);
811
	chip_bus_sync_unlock(desc);
T
Thomas Gleixner 已提交
812 813
}

814
#ifdef CONFIG_SMP
815
/*
816
 * Check whether we need to change the affinity of the interrupt thread.
817 818 819 820 821
 */
static void
irq_thread_check_affinity(struct irq_desc *desc, struct irqaction *action)
{
	cpumask_var_t mask;
822
	bool valid = true;
823 824 825 826 827 828 829 830 831 832 833 834 835

	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;
	}

836
	raw_spin_lock_irq(&desc->lock);
837 838 839 840
	/*
	 * This code is triggered unconditionally. Check the affinity
	 * mask pointer. For CPU_MASK_OFFSTACK=n this is optimized out.
	 */
841
	if (cpumask_available(desc->irq_common_data.affinity))
842
		cpumask_copy(mask, desc->irq_common_data.affinity);
843 844
	else
		valid = false;
845
	raw_spin_unlock_irq(&desc->lock);
846

847 848
	if (valid)
		set_cpus_allowed_ptr(current, mask);
849 850
	free_cpumask_var(mask);
}
851 852 853 854
#else
static inline void
irq_thread_check_affinity(struct irq_desc *desc, struct irqaction *action) { }
#endif
855

856 857 858 859 860 861
/*
 * 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.
 */
862
static irqreturn_t
863 864
irq_forced_thread_fn(struct irq_desc *desc, struct irqaction *action)
{
865 866
	irqreturn_t ret;

867
	local_bh_disable();
868
	ret = action->thread_fn(action->irq, action->dev_id);
869
	irq_finalize_oneshot(desc, action);
870
	local_bh_enable();
871
	return ret;
872 873 874
}

/*
875
 * Interrupts explicitly requested as threaded interrupts want to be
876 877 878
 * preemtible - many of them need to sleep and wait for slow busses to
 * complete.
 */
879 880
static irqreturn_t irq_thread_fn(struct irq_desc *desc,
		struct irqaction *action)
881
{
882 883 884
	irqreturn_t ret;

	ret = action->thread_fn(action->irq, action->dev_id);
885
	irq_finalize_oneshot(desc, action);
886
	return ret;
887 888
}

889 890
static void wake_threads_waitq(struct irq_desc *desc)
{
891
	if (atomic_dec_and_test(&desc->threads_active))
892 893 894
		wake_up(&desc->wait_for_threads);
}

895
static void irq_thread_dtor(struct callback_head *unused)
896 897 898 899 900 901 902 903 904 905
{
	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);

906
	pr_err("exiting task \"%s\" (%d) is an active IRQ thread (irq %d)\n",
A
Alan Cox 已提交
907
	       tsk->comm, tsk->pid, action->irq);
908 909 910 911 912 913 914 915 916 917 918 919 920 921


	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);
}

922 923 924 925 926 927 928 929 930 931 932 933
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);
}

934 935 936 937 938
/*
 * Interrupt handler thread
 */
static int irq_thread(void *data)
{
939
	struct callback_head on_exit_work;
940 941
	struct irqaction *action = data;
	struct irq_desc *desc = irq_to_desc(action->irq);
942 943
	irqreturn_t (*handler_fn)(struct irq_desc *desc,
			struct irqaction *action);
944

945
	if (force_irqthreads && test_bit(IRQTF_FORCED_THREAD,
946 947 948 949 950
					&action->thread_flags))
		handler_fn = irq_forced_thread_fn;
	else
		handler_fn = irq_thread_fn;

A
Al Viro 已提交
951
	init_task_work(&on_exit_work, irq_thread_dtor);
952
	task_work_add(current, &on_exit_work, false);
953

954 955
	irq_thread_check_affinity(desc, action);

956
	while (!irq_wait_for_interrupt(action)) {
957
		irqreturn_t action_ret;
958

959 960
		irq_thread_check_affinity(desc, action);

961
		action_ret = handler_fn(desc, action);
962 963
		if (action_ret == IRQ_HANDLED)
			atomic_inc(&desc->threads_handled);
964 965
		if (action_ret == IRQ_WAKE_THREAD)
			irq_wake_secondary(desc, action);
966

967
		wake_threads_waitq(desc);
968 969
	}

970 971 972 973
	/*
	 * 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
974 975 976 977
	 * 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.
978
	 */
979
	task_work_cancel(current, irq_thread_dtor);
980 981 982
	return 0;
}

983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998
/**
 *	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);
999
	for_each_action_of_desc(desc, action) {
1000 1001 1002 1003 1004 1005 1006 1007 1008 1009
		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);

1010
static int irq_setup_forced_threading(struct irqaction *new)
1011 1012
{
	if (!force_irqthreads)
1013
		return 0;
1014
	if (new->flags & (IRQF_NO_THREAD | IRQF_PERCPU | IRQF_ONESHOT))
1015
		return 0;
1016 1017 1018

	new->flags |= IRQF_ONESHOT;

1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033
	/*
	 * 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;
1034
	}
1035 1036 1037 1038 1039
	/* 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;
1040 1041
}

1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058
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);
}

1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 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
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
1101 1102 1103
/*
 * Internal function to register an irqaction - typically used to
 * allocate special interrupts that are part of the architecture.
1104 1105 1106 1107 1108 1109 1110 1111 1112 1113
 *
 * 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
1114
 */
1115
static int
I
Ingo Molnar 已提交
1116
__setup_irq(unsigned int irq, struct irq_desc *desc, struct irqaction *new)
L
Linus Torvalds 已提交
1117
{
1118
	struct irqaction *old, **old_ptr;
1119
	unsigned long flags, thread_mask = 0;
1120
	int ret, nested, shared = 0;
L
Linus Torvalds 已提交
1121

1122
	if (!desc)
1123 1124
		return -EINVAL;

1125
	if (desc->irq_data.chip == &no_irq_chip)
L
Linus Torvalds 已提交
1126
		return -ENOSYS;
1127 1128
	if (!try_module_get(desc->owner))
		return -ENODEV;
L
Linus Torvalds 已提交
1129

1130 1131
	new->irq = irq;

1132 1133 1134 1135 1136 1137 1138
	/*
	 * 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);

1139
	/*
1140 1141 1142
	 * Check whether the interrupt nests into another interrupt
	 * thread.
	 */
1143
	nested = irq_settings_is_nested_thread(desc);
1144
	if (nested) {
1145 1146 1147 1148
		if (!new->thread_fn) {
			ret = -EINVAL;
			goto out_mput;
		}
1149 1150 1151 1152 1153 1154
		/*
		 * 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;
1155
	} else {
1156 1157 1158 1159 1160
		if (irq_settings_can_thread(desc)) {
			ret = irq_setup_forced_threading(new);
			if (ret)
				goto out_mput;
		}
1161 1162
	}

1163
	/*
1164 1165 1166
	 * Create a handler thread when a thread function is supplied
	 * and the interrupt does not nest into another interrupt
	 * thread.
1167
	 */
1168
	if (new->thread_fn && !nested) {
1169 1170
		ret = setup_irq_thread(new, irq, false);
		if (ret)
1171
			goto out_mput;
1172 1173 1174 1175
		if (new->secondary) {
			ret = setup_irq_thread(new->secondary, irq, true);
			if (ret)
				goto out_thread;
1176
		}
1177 1178
	}

1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190
	/*
	 * 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;

1191 1192 1193 1194 1195
	/*
	 * 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.
	 */
1196
	mutex_lock(&desc->request_mutex);
1197 1198 1199 1200 1201 1202 1203 1204 1205

	/*
	 * 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. */
1206 1207 1208 1209 1210
	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);
1211
			goto out_bus_unlock;
1212 1213
		}
	}
1214

L
Linus Torvalds 已提交
1215 1216
	/*
	 * The following block of code has to be executed atomically
1217 1218 1219
	 * 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
1220
	 */
1221
	raw_spin_lock_irqsave(&desc->lock, flags);
1222 1223
	old_ptr = &desc->action;
	old = *old_ptr;
1224
	if (old) {
1225 1226 1227
		/*
		 * Can't share interrupts unless both agree to and are
		 * the same type (level, edge, polarity). So both flag
1228
		 * fields must have IRQF_SHARED set and the bits which
1229 1230
		 * set the trigger type must match. Also all must
		 * agree on ONESHOT.
1231
		 */
1232 1233
		unsigned int oldtype = irqd_get_trigger_type(&desc->irq_data);

1234
		if (!((old->flags & new->flags) & IRQF_SHARED) ||
1235
		    (oldtype != (new->flags & IRQF_TRIGGER_MASK)) ||
1236
		    ((old->flags ^ new->flags) & IRQF_ONESHOT))
1237 1238 1239
			goto mismatch;

		/* All handlers must agree on per-cpuness */
1240 1241
		if ((old->flags & IRQF_PERCPU) !=
		    (new->flags & IRQF_PERCPU))
1242
			goto mismatch;
L
Linus Torvalds 已提交
1243 1244 1245

		/* add new interrupt at end of irq queue */
		do {
1246 1247 1248 1249 1250
			/*
			 * Or all existing action->thread_mask bits,
			 * so we can find the next zero bit for this
			 * new action.
			 */
1251
			thread_mask |= old->thread_mask;
1252 1253
			old_ptr = &old->next;
			old = *old_ptr;
L
Linus Torvalds 已提交
1254 1255 1256 1257
		} while (old);
		shared = 1;
	}

1258
	/*
1259 1260 1261
	 * 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().
1262
	 */
1263 1264 1265 1266 1267 1268 1269
	if (new->flags & IRQF_ONESHOT) {
		/*
		 * Unlikely to have 32 resp 64 irqs sharing one line,
		 * but who knows.
		 */
		if (thread_mask == ~0UL) {
			ret = -EBUSY;
1270
			goto out_unlock;
1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291
		}
		/*
		 * 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.
		 */
1292
		new->thread_mask = 1UL << ffz(thread_mask);
1293

1294 1295
	} else if (new->handler == irq_default_primary_handler &&
		   !(desc->irq_data.chip->flags & IRQCHIP_ONESHOT_SAFE)) {
1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310
		/*
		 * 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.
		 */
1311
		pr_err("Threaded irq requested with handler=NULL and !ONESHOT for irq %d\n",
1312 1313
		       irq);
		ret = -EINVAL;
1314
		goto out_unlock;
1315 1316
	}

L
Linus Torvalds 已提交
1317
	if (!shared) {
1318 1319
		init_waitqueue_head(&desc->wait_for_threads);

1320
		/* Setup the type (level, edge polarity) if configured: */
1321
		if (new->flags & IRQF_TRIGGER_MASK) {
1322 1323
			ret = __irq_set_trigger(desc,
						new->flags & IRQF_TRIGGER_MASK);
1324

1325
			if (ret)
1326
				goto out_unlock;
1327
		}
T
Thomas Gleixner 已提交
1328

1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343
		/*
		 * Activate the interrupt. That activation must happen
		 * independently of IRQ_NOAUTOEN. request_irq() can fail
		 * and the callers are supposed to handle
		 * that. enable_irq() of an interrupt requested with
		 * IRQ_NOAUTOEN is not supposed to fail. The activation
		 * keeps it in shutdown mode, it merily associates
		 * resources if necessary and if that's not possible it
		 * fails. Interrupts which are in managed shutdown mode
		 * will simply ignore that activation request.
		 */
		ret = irq_activate(desc);
		if (ret)
			goto out_unlock;

1344
		desc->istate &= ~(IRQS_AUTODETECT | IRQS_SPURIOUS_DISABLED | \
1345 1346
				  IRQS_ONESHOT | IRQS_WAITING);
		irqd_clear(&desc->irq_data, IRQD_IRQ_INPROGRESS);
1347

1348 1349 1350 1351
		if (new->flags & IRQF_PERCPU) {
			irqd_set(&desc->irq_data, IRQD_PER_CPU);
			irq_settings_set_per_cpu(desc);
		}
1352

T
Thomas Gleixner 已提交
1353
		if (new->flags & IRQF_ONESHOT)
1354
			desc->istate |= IRQS_ONESHOT;
T
Thomas Gleixner 已提交
1355

1356 1357 1358 1359 1360 1361
		/* 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);
		}

1362
		if (irq_settings_can_autoenable(desc)) {
1363
			irq_startup(desc, IRQ_RESEND, IRQ_START_COND);
1364 1365 1366 1367 1368 1369 1370 1371
		} 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);
1372 1373
			/* Undo nested disables: */
			desc->depth = 1;
1374
		}
1375

1376 1377
	} else if (new->flags & IRQF_TRIGGER_MASK) {
		unsigned int nmsk = new->flags & IRQF_TRIGGER_MASK;
1378
		unsigned int omsk = irqd_get_trigger_type(&desc->irq_data);
1379 1380 1381

		if (nmsk != omsk)
			/* hope the handler works with current  trigger mode */
1382
			pr_warn("irq %d uses trigger mode %u; requested %u\n",
1383
				irq, omsk, nmsk);
L
Linus Torvalds 已提交
1384
	}
1385

1386
	*old_ptr = new;
1387

1388 1389
	irq_pm_install_action(desc, new);

1390 1391 1392
	/* Reset broken irq detection when installing new handler */
	desc->irq_count = 0;
	desc->irqs_unhandled = 0;
1393 1394 1395 1396 1397

	/*
	 * Check whether we disabled the irq via the spurious handler
	 * before. Reenable it and give it another chance.
	 */
1398 1399
	if (shared && (desc->istate & IRQS_SPURIOUS_DISABLED)) {
		desc->istate &= ~IRQS_SPURIOUS_DISABLED;
1400
		__enable_irq(desc);
1401 1402
	}

1403
	raw_spin_unlock_irqrestore(&desc->lock, flags);
1404
	chip_bus_sync_unlock(desc);
1405
	mutex_unlock(&desc->request_mutex);
L
Linus Torvalds 已提交
1406

1407 1408
	irq_setup_timings(desc, new);

1409 1410 1411 1412 1413 1414
	/*
	 * 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);
1415 1416
	if (new->secondary)
		wake_up_process(new->secondary->thread);
1417

1418
	register_irq_proc(irq, desc);
L
Linus Torvalds 已提交
1419 1420 1421
	new->dir = NULL;
	register_handler_proc(irq, new);
	return 0;
1422 1423

mismatch:
1424
	if (!(new->flags & IRQF_PROBE_SHARED)) {
1425
		pr_err("Flags mismatch irq %d. %08x (%s) vs. %08x (%s)\n",
1426 1427
		       irq, new->flags, new->name, old->flags, old->name);
#ifdef CONFIG_DEBUG_SHIRQ
1428
		dump_stack();
1429
#endif
1430
	}
1431 1432
	ret = -EBUSY;

1433
out_unlock:
1434
	raw_spin_unlock_irqrestore(&desc->lock, flags);
1435

1436 1437
	if (!desc->action)
		irq_release_resources(desc);
1438 1439
out_bus_unlock:
	chip_bus_sync_unlock(desc);
1440 1441
	mutex_unlock(&desc->request_mutex);

1442 1443 1444 1445 1446
out_thread:
	if (new->thread) {
		struct task_struct *t = new->thread;

		new->thread = NULL;
1447
		kthread_stop(t);
1448 1449
		put_task_struct(t);
	}
1450 1451 1452 1453 1454 1455 1456
	if (new->secondary && new->secondary->thread) {
		struct task_struct *t = new->secondary->thread;

		new->secondary->thread = NULL;
		kthread_stop(t);
		put_task_struct(t);
	}
1457 1458
out_mput:
	module_put(desc->owner);
1459
	return ret;
L
Linus Torvalds 已提交
1460 1461
}

1462 1463 1464 1465 1466 1467 1468 1469 1470
/**
 *	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)
{
1471
	int retval;
1472 1473
	struct irq_desc *desc = irq_to_desc(irq);

1474
	if (!desc || WARN_ON(irq_settings_is_per_cpu_devid(desc)))
1475
		return -EINVAL;
1476 1477 1478 1479 1480

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

1481 1482
	retval = __setup_irq(irq, desc, act);

1483 1484 1485
	if (retval)
		irq_chip_pm_put(&desc->irq_data);

1486
	return retval;
1487
}
1488
EXPORT_SYMBOL_GPL(setup_irq);
1489

1490
/*
1491 1492
 * 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
1493
 */
1494
static struct irqaction *__free_irq(unsigned int irq, void *dev_id)
L
Linus Torvalds 已提交
1495
{
1496
	struct irq_desc *desc = irq_to_desc(irq);
1497
	struct irqaction *action, **action_ptr;
L
Linus Torvalds 已提交
1498 1499
	unsigned long flags;

1500
	WARN(in_interrupt(), "Trying to free IRQ %d from IRQ context!\n", irq);
1501 1502

	if (!desc)
1503
		return NULL;
L
Linus Torvalds 已提交
1504

1505
	mutex_lock(&desc->request_mutex);
1506
	chip_bus_lock(desc);
1507
	raw_spin_lock_irqsave(&desc->lock, flags);
1508 1509 1510 1511 1512

	/*
	 * There can be multiple actions per IRQ descriptor, find the right
	 * one based on the dev_id:
	 */
1513
	action_ptr = &desc->action;
L
Linus Torvalds 已提交
1514
	for (;;) {
1515
		action = *action_ptr;
L
Linus Torvalds 已提交
1516

1517 1518
		if (!action) {
			WARN(1, "Trying to free already-free IRQ %d\n", irq);
1519
			raw_spin_unlock_irqrestore(&desc->lock, flags);
1520
			chip_bus_sync_unlock(desc);
1521
			mutex_unlock(&desc->request_mutex);
1522
			return NULL;
1523
		}
L
Linus Torvalds 已提交
1524

1525 1526
		if (action->dev_id == dev_id)
			break;
1527
		action_ptr = &action->next;
1528
	}
1529

1530
	/* Found it - now remove it from the list of entries: */
1531
	*action_ptr = action->next;
1532

1533 1534
	irq_pm_remove_action(desc, action);

1535
	/* If this was the last handler, shut down the IRQ line: */
1536
	if (!desc->action) {
1537
		irq_settings_clr_disable_unlazy(desc);
1538
		irq_shutdown(desc);
1539
	}
1540

1541 1542 1543 1544 1545 1546
#ifdef CONFIG_SMP
	/* make sure affinity_hint is cleaned up */
	if (WARN_ON_ONCE(desc->affinity_hint))
		desc->affinity_hint = NULL;
#endif

1547
	raw_spin_unlock_irqrestore(&desc->lock, flags);
1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561
	/*
	 * 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.
	 */
1562
	chip_bus_sync_unlock(desc);
1563 1564 1565 1566 1567

	unregister_handler_proc(irq, action);

	/* Make sure it's not being used on another CPU: */
	synchronize_irq(irq);
L
Linus Torvalds 已提交
1568

1569
#ifdef CONFIG_DEBUG_SHIRQ
1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581
	/*
	 * 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
1582
	}
1583
#endif
1584 1585

	if (action->thread) {
1586
		kthread_stop(action->thread);
1587
		put_task_struct(action->thread);
1588 1589 1590 1591
		if (action->secondary && action->secondary->thread) {
			kthread_stop(action->secondary->thread);
			put_task_struct(action->secondary->thread);
		}
1592 1593
	}

1594
	/* Last action releases resources */
1595
	if (!desc->action) {
1596 1597 1598 1599 1600
		/*
		 * Reaquire bus lock as irq_release_resources() might
		 * require it to deallocate resources over the slow bus.
		 */
		chip_bus_lock(desc);
1601
		irq_release_resources(desc);
1602
		chip_bus_sync_unlock(desc);
1603 1604
		irq_remove_timings(desc);
	}
1605

1606 1607
	mutex_unlock(&desc->request_mutex);

1608
	irq_chip_pm_put(&desc->irq_data);
1609
	module_put(desc->owner);
1610
	kfree(action->secondary);
1611 1612 1613
	return action;
}

1614 1615 1616 1617 1618 1619 1620 1621 1622
/**
 *	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)
{
1623 1624 1625
	struct irq_desc *desc = irq_to_desc(irq);

	if (desc && !WARN_ON(irq_settings_is_per_cpu_devid(desc)))
1626
		__free_irq(irq, act->dev_id);
1627
}
1628
EXPORT_SYMBOL_GPL(remove_irq);
1629

1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642
/**
 *	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.
1643 1644
 *
 *	Returns the devname argument passed to request_irq.
1645
 */
1646
const void *free_irq(unsigned int irq, void *dev_id)
1647
{
T
Thomas Gleixner 已提交
1648
	struct irq_desc *desc = irq_to_desc(irq);
1649 1650
	struct irqaction *action;
	const char *devname;
T
Thomas Gleixner 已提交
1651

1652
	if (!desc || WARN_ON(irq_settings_is_per_cpu_devid(desc)))
1653
		return NULL;
T
Thomas Gleixner 已提交
1654

1655 1656 1657 1658 1659
#ifdef CONFIG_SMP
	if (WARN_ON(desc->affinity_notify))
		desc->affinity_notify = NULL;
#endif

1660
	action = __free_irq(irq, dev_id);
1661 1662 1663 1664

	if (!action)
		return NULL;

1665 1666 1667
	devname = action->name;
	kfree(action);
	return devname;
L
Linus Torvalds 已提交
1668 1669 1670 1671
}
EXPORT_SYMBOL(free_irq);

/**
1672
 *	request_threaded_irq - allocate an interrupt line
L
Linus Torvalds 已提交
1673
 *	@irq: Interrupt line to allocate
1674 1675
 *	@handler: Function to be called when the IRQ occurs.
 *		  Primary handler for threaded interrupts
T
Thomas Gleixner 已提交
1676 1677
 *		  If NULL and thread_fn != NULL the default
 *		  primary handler is installed
1678 1679
 *	@thread_fn: Function called from the irq handler thread
 *		    If NULL, no irq thread is created
L
Linus Torvalds 已提交
1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690
 *	@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.
 *
1691
 *	If you want to set up a threaded irq handler for your device
J
Javi Merino 已提交
1692
 *	then you need to supply @handler and @thread_fn. @handler is
1693 1694 1695
 *	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
1696
 *	IRQ_WAKE_THREAD which will wake up the handler thread and run
1697 1698 1699
 *	@thread_fn. This split handler design is necessary to support
 *	shared interrupts.
 *
L
Linus Torvalds 已提交
1700 1701 1702 1703 1704 1705 1706 1707 1708
 *	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:
 *
1709
 *	IRQF_SHARED		Interrupt is shared
D
David Brownell 已提交
1710
 *	IRQF_TRIGGER_*		Specify active edge(s) or level
L
Linus Torvalds 已提交
1711 1712
 *
 */
1713 1714 1715
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
1716
{
1717
	struct irqaction *action;
1718
	struct irq_desc *desc;
1719
	int retval;
L
Linus Torvalds 已提交
1720

1721 1722 1723
	if (irq == IRQ_NOTCONNECTED)
		return -ENOTCONN;

L
Linus Torvalds 已提交
1724 1725 1726 1727 1728
	/*
	 * 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).
1729 1730 1731
	 *
	 * 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
1732
	 */
1733 1734 1735
	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
1736
		return -EINVAL;
1737

1738
	desc = irq_to_desc(irq);
1739
	if (!desc)
L
Linus Torvalds 已提交
1740
		return -EINVAL;
1741

1742 1743
	if (!irq_settings_can_request(desc) ||
	    WARN_ON(irq_settings_is_per_cpu_devid(desc)))
1744
		return -EINVAL;
T
Thomas Gleixner 已提交
1745 1746 1747 1748 1749 1750

	if (!handler) {
		if (!thread_fn)
			return -EINVAL;
		handler = irq_default_primary_handler;
	}
L
Linus Torvalds 已提交
1751

1752
	action = kzalloc(sizeof(struct irqaction), GFP_KERNEL);
L
Linus Torvalds 已提交
1753 1754 1755 1756
	if (!action)
		return -ENOMEM;

	action->handler = handler;
1757
	action->thread_fn = thread_fn;
L
Linus Torvalds 已提交
1758 1759 1760 1761
	action->flags = irqflags;
	action->name = devname;
	action->dev_id = dev_id;

1762
	retval = irq_chip_pm_get(&desc->irq_data);
1763 1764
	if (retval < 0) {
		kfree(action);
1765
		return retval;
1766
	}
1767

1768
	retval = __setup_irq(irq, desc, action);
T
Thomas Gleixner 已提交
1769

1770
	if (retval) {
1771
		irq_chip_pm_put(&desc->irq_data);
1772
		kfree(action->secondary);
1773
		kfree(action);
1774
	}
1775

1776
#ifdef CONFIG_DEBUG_SHIRQ_FIXME
1777
	if (!retval && (irqflags & IRQF_SHARED)) {
D
David Woodhouse 已提交
1778 1779 1780
		/*
		 * It's a shared IRQ -- the driver ought to be prepared for it
		 * to happen immediately, so let's make sure....
1781 1782
		 * 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
1783
		 */
1784
		unsigned long flags;
D
David Woodhouse 已提交
1785

1786
		disable_irq(irq);
1787
		local_irq_save(flags);
1788

1789
		handler(irq, dev_id);
1790

1791
		local_irq_restore(flags);
1792
		enable_irq(irq);
D
David Woodhouse 已提交
1793 1794
	}
#endif
L
Linus Torvalds 已提交
1795 1796
	return retval;
}
1797
EXPORT_SYMBOL(request_threaded_irq);
1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818

/**
 *	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)
{
1819
	struct irq_desc *desc;
1820 1821
	int ret;

1822 1823 1824 1825
	if (irq == IRQ_NOTCONNECTED)
		return -ENOTCONN;

	desc = irq_to_desc(irq);
1826 1827 1828
	if (!desc)
		return -EINVAL;

1829
	if (irq_settings_is_nested_thread(desc)) {
1830 1831 1832 1833 1834 1835 1836 1837 1838
		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);
1839

1840
void enable_percpu_irq(unsigned int irq, unsigned int type)
1841 1842 1843 1844 1845 1846 1847 1848
{
	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;

1849 1850 1851 1852
	/*
	 * If the trigger type is not specified by the caller, then
	 * use the default for this interrupt.
	 */
1853
	type &= IRQ_TYPE_SENSE_MASK;
1854 1855 1856
	if (type == IRQ_TYPE_NONE)
		type = irqd_get_trigger_type(&desc->irq_data);

1857 1858 1859
	if (type != IRQ_TYPE_NONE) {
		int ret;

1860
		ret = __irq_set_trigger(desc, type);
1861 1862

		if (ret) {
1863
			WARN(1, "failed to set type for IRQ%d\n", irq);
1864 1865 1866 1867
			goto out;
		}
	}

1868
	irq_percpu_enable(desc, cpu);
1869
out:
1870 1871
	irq_put_desc_unlock(desc, flags);
}
1872
EXPORT_SYMBOL_GPL(enable_percpu_irq);
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 1898
/**
 * 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);

1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910
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);
}
1911
EXPORT_SYMBOL_GPL(disable_percpu_irq);
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 1947

/*
 * 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);

1948
	irq_chip_pm_put(&desc->irq_data);
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 1994
	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);
}
1995
EXPORT_SYMBOL_GPL(free_percpu_irq);
1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

/**
 *	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;
2011 2012 2013 2014 2015

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

2016 2017
	retval = __setup_irq(irq, desc, act);

2018 2019 2020
	if (retval)
		irq_chip_pm_put(&desc->irq_data);

2021 2022 2023 2024
	return retval;
}

/**
2025
 *	__request_percpu_irq - allocate a percpu interrupt line
2026 2027
 *	@irq: Interrupt line to allocate
 *	@handler: Function to be called when the IRQ occurs.
2028
 *	@flags: Interrupt type flags (IRQF_TIMER only)
2029 2030 2031
 *	@devname: An ascii name for the claiming device
 *	@dev_id: A percpu cookie passed back to the handler function
 *
2032 2033 2034 2035
 *	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().
2036 2037 2038 2039 2040
 *
 *	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.
 */
2041 2042 2043
int __request_percpu_irq(unsigned int irq, irq_handler_t handler,
			 unsigned long flags, const char *devname,
			 void __percpu *dev_id)
2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056
{
	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;

2057 2058 2059
	if (flags && flags != IRQF_TIMER)
		return -EINVAL;

2060 2061 2062 2063 2064
	action = kzalloc(sizeof(struct irqaction), GFP_KERNEL);
	if (!action)
		return -ENOMEM;

	action->handler = handler;
2065
	action->flags = flags | IRQF_PERCPU | IRQF_NO_SUSPEND;
2066 2067 2068
	action->name = devname;
	action->percpu_dev_id = dev_id;

2069
	retval = irq_chip_pm_get(&desc->irq_data);
2070 2071
	if (retval < 0) {
		kfree(action);
2072
		return retval;
2073
	}
2074

2075 2076
	retval = __setup_irq(irq, desc, action);

2077 2078
	if (retval) {
		irq_chip_pm_put(&desc->irq_data);
2079
		kfree(action);
2080
	}
2081 2082 2083

	return retval;
}
2084
EXPORT_SYMBOL_GPL(__request_percpu_irq);
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 2130

/**
 *	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;
}
2131
EXPORT_SYMBOL_GPL(irq_get_irqchip_state);
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 2176

/**
 *	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;
}
2177
EXPORT_SYMBOL_GPL(irq_set_irqchip_state);