events.c 36.4 KB
Newer Older
J
Jeremy Fitzhardinge 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
/*
 * Xen event channels
 *
 * Xen models interrupts with abstract event channels.  Because each
 * domain gets 1024 event channels, but NR_IRQ is not that large, we
 * must dynamically map irqs<->event channels.  The event channels
 * interface with the rest of the kernel by defining a xen interrupt
 * chip.  When an event is recieved, it is mapped to an irq and sent
 * through the normal interrupt processing path.
 *
 * There are four kinds of events which can be mapped to an event
 * channel:
 *
 * 1. Inter-domain notifications.  This includes all the virtual
 *    device events, since they're driven by front-ends in another domain
 *    (typically dom0).
 * 2. VIRQs, typically used for timers.  These are per-cpu events.
 * 3. IPIs.
19
 * 4. PIRQs - Hardware interrupts.
J
Jeremy Fitzhardinge 已提交
20 21 22 23 24 25 26 27 28
 *
 * Jeremy Fitzhardinge <jeremy@xensource.com>, XenSource Inc, 2007
 */

#include <linux/linkage.h>
#include <linux/interrupt.h>
#include <linux/irq.h>
#include <linux/module.h>
#include <linux/string.h>
29
#include <linux/bootmem.h>
30
#include <linux/slab.h>
31
#include <linux/irqnr.h>
32
#include <linux/pci.h>
J
Jeremy Fitzhardinge 已提交
33

34
#include <asm/desc.h>
J
Jeremy Fitzhardinge 已提交
35 36
#include <asm/ptrace.h>
#include <asm/irq.h>
37
#include <asm/idle.h>
38
#include <asm/io_apic.h>
J
Jeremy Fitzhardinge 已提交
39
#include <asm/sync_bitops.h>
40
#include <asm/xen/pci.h>
J
Jeremy Fitzhardinge 已提交
41
#include <asm/xen/hypercall.h>
42
#include <asm/xen/hypervisor.h>
J
Jeremy Fitzhardinge 已提交
43

44 45
#include <xen/xen.h>
#include <xen/hvm.h>
46
#include <xen/xen-ops.h>
J
Jeremy Fitzhardinge 已提交
47 48 49
#include <xen/events.h>
#include <xen/interface/xen.h>
#include <xen/interface/event_channel.h>
50 51
#include <xen/interface/hvm/hvm_op.h>
#include <xen/interface/hvm/params.h>
J
Jeremy Fitzhardinge 已提交
52 53 54 55 56 57 58 59

/*
 * This lock protects updates to the following mapping and reference-count
 * arrays. The lock does not need to be acquired to read the mapping tables.
 */
static DEFINE_SPINLOCK(irq_mapping_update_lock);

/* IRQ <-> VIRQ mapping. */
60
static DEFINE_PER_CPU(int [NR_VIRQS], virq_to_irq) = {[0 ... NR_VIRQS-1] = -1};
J
Jeremy Fitzhardinge 已提交
61

J
Jeremy Fitzhardinge 已提交
62
/* IRQ <-> IPI mapping */
63
static DEFINE_PER_CPU(int [XEN_NR_IPIS], ipi_to_irq) = {[0 ... XEN_NR_IPIS-1] = -1};
J
Jeremy Fitzhardinge 已提交
64

65 66
/* Interrupt types. */
enum xen_irq_type {
J
Jeremy Fitzhardinge 已提交
67
	IRQT_UNBOUND = 0,
J
Jeremy Fitzhardinge 已提交
68 69 70 71 72
	IRQT_PIRQ,
	IRQT_VIRQ,
	IRQT_IPI,
	IRQT_EVTCHN
};
J
Jeremy Fitzhardinge 已提交
73

74 75 76 77 78 79
/*
 * Packed IRQ information:
 * type - enum xen_irq_type
 * event channel - irq->event channel mapping
 * cpu - cpu this event channel is bound to
 * index - type-specific information:
80 81
 *    PIRQ - vector, with MSB being "needs EIO", or physical IRQ of the HVM
 *           guest, or GSI (real passthrough IRQ) of the device.
82 83 84 85 86 87 88 89 90 91 92 93 94 95
 *    VIRQ - virq number
 *    IPI - IPI vector
 *    EVTCHN -
 */
struct irq_info
{
	enum xen_irq_type type;	/* type */
	unsigned short evtchn;	/* event channel */
	unsigned short cpu;	/* cpu bound */

	union {
		unsigned short virq;
		enum ipi_vector ipi;
		struct {
S
Stefano Stabellini 已提交
96
			unsigned short pirq;
97
			unsigned short gsi;
98 99
			unsigned char vector;
			unsigned char flags;
100 101 102
		} pirq;
	} u;
};
103
#define PIRQ_NEEDS_EOI	(1 << 0)
104
#define PIRQ_SHAREABLE	(1 << 1)
105

106
static struct irq_info *irq_info;
S
Stefano Stabellini 已提交
107
static int *pirq_to_irq;
J
Jeremy Fitzhardinge 已提交
108

109
static int *evtchn_to_irq;
110 111 112
struct cpu_evtchn_s {
	unsigned long bits[NR_EVENT_CHANNELS/BITS_PER_LONG];
};
113 114 115 116 117 118

static __initdata struct cpu_evtchn_s init_evtchn_mask = {
	.bits[0 ... (NR_EVENT_CHANNELS/BITS_PER_LONG)-1] = ~0ul,
};
static struct cpu_evtchn_s *cpu_evtchn_mask_p = &init_evtchn_mask;

119 120 121 122
static inline unsigned long *cpu_evtchn_mask(int cpu)
{
	return cpu_evtchn_mask_p[cpu].bits;
}
J
Jeremy Fitzhardinge 已提交
123 124 125 126 127

/* Xen will never allocate port zero for any purpose. */
#define VALID_EVTCHN(chn)	((chn) != 0)

static struct irq_chip xen_dynamic_chip;
128
static struct irq_chip xen_percpu_chip;
129
static struct irq_chip xen_pirq_chip;
J
Jeremy Fitzhardinge 已提交
130 131

/* Constructor for packed IRQ information. */
132 133 134 135 136 137 138
static struct irq_info mk_unbound_info(void)
{
	return (struct irq_info) { .type = IRQT_UNBOUND };
}

static struct irq_info mk_evtchn_info(unsigned short evtchn)
{
139 140
	return (struct irq_info) { .type = IRQT_EVTCHN, .evtchn = evtchn,
			.cpu = 0 };
141 142 143
}

static struct irq_info mk_ipi_info(unsigned short evtchn, enum ipi_vector ipi)
J
Jeremy Fitzhardinge 已提交
144
{
145
	return (struct irq_info) { .type = IRQT_IPI, .evtchn = evtchn,
146
			.cpu = 0, .u.ipi = ipi };
147 148 149 150 151
}

static struct irq_info mk_virq_info(unsigned short evtchn, unsigned short virq)
{
	return (struct irq_info) { .type = IRQT_VIRQ, .evtchn = evtchn,
152
			.cpu = 0, .u.virq = virq };
153 154
}

S
Stefano Stabellini 已提交
155
static struct irq_info mk_pirq_info(unsigned short evtchn, unsigned short pirq,
156 157 158
				    unsigned short gsi, unsigned short vector)
{
	return (struct irq_info) { .type = IRQT_PIRQ, .evtchn = evtchn,
S
Stefano Stabellini 已提交
159 160
			.cpu = 0,
			.u.pirq = { .pirq = pirq, .gsi = gsi, .vector = vector } };
J
Jeremy Fitzhardinge 已提交
161 162 163 164 165
}

/*
 * Accessors for packed IRQ information.
 */
166
static struct irq_info *info_for_irq(unsigned irq)
J
Jeremy Fitzhardinge 已提交
167
{
168
	return &irq_info[irq];
J
Jeremy Fitzhardinge 已提交
169 170
}

171
static unsigned int evtchn_from_irq(unsigned irq)
J
Jeremy Fitzhardinge 已提交
172
{
173
	return info_for_irq(irq)->evtchn;
J
Jeremy Fitzhardinge 已提交
174 175
}

I
Ian Campbell 已提交
176 177 178 179 180 181
unsigned irq_from_evtchn(unsigned int evtchn)
{
	return evtchn_to_irq[evtchn];
}
EXPORT_SYMBOL_GPL(irq_from_evtchn);

182
static enum ipi_vector ipi_from_irq(unsigned irq)
J
Jeremy Fitzhardinge 已提交
183
{
184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201
	struct irq_info *info = info_for_irq(irq);

	BUG_ON(info == NULL);
	BUG_ON(info->type != IRQT_IPI);

	return info->u.ipi;
}

static unsigned virq_from_irq(unsigned irq)
{
	struct irq_info *info = info_for_irq(irq);

	BUG_ON(info == NULL);
	BUG_ON(info->type != IRQT_VIRQ);

	return info->u.virq;
}

S
Stefano Stabellini 已提交
202 203 204 205 206 207 208 209 210 211
static unsigned pirq_from_irq(unsigned irq)
{
	struct irq_info *info = info_for_irq(irq);

	BUG_ON(info == NULL);
	BUG_ON(info->type != IRQT_PIRQ);

	return info->u.pirq.pirq;
}

212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250
static unsigned gsi_from_irq(unsigned irq)
{
	struct irq_info *info = info_for_irq(irq);

	BUG_ON(info == NULL);
	BUG_ON(info->type != IRQT_PIRQ);

	return info->u.pirq.gsi;
}

static unsigned vector_from_irq(unsigned irq)
{
	struct irq_info *info = info_for_irq(irq);

	BUG_ON(info == NULL);
	BUG_ON(info->type != IRQT_PIRQ);

	return info->u.pirq.vector;
}

static enum xen_irq_type type_from_irq(unsigned irq)
{
	return info_for_irq(irq)->type;
}

static unsigned cpu_from_irq(unsigned irq)
{
	return info_for_irq(irq)->cpu;
}

static unsigned int cpu_from_evtchn(unsigned int evtchn)
{
	int irq = evtchn_to_irq[evtchn];
	unsigned ret = 0;

	if (irq != -1)
		ret = cpu_from_irq(irq);

	return ret;
J
Jeremy Fitzhardinge 已提交
251 252
}

253 254 255 256 257 258 259 260 261
static bool pirq_needs_eoi(unsigned irq)
{
	struct irq_info *info = info_for_irq(irq);

	BUG_ON(info->type != IRQT_PIRQ);

	return info->u.pirq.flags & PIRQ_NEEDS_EOI;
}

J
Jeremy Fitzhardinge 已提交
262 263 264 265 266
static inline unsigned long active_evtchns(unsigned int cpu,
					   struct shared_info *sh,
					   unsigned int idx)
{
	return (sh->evtchn_pending[idx] &
267
		cpu_evtchn_mask(cpu)[idx] &
J
Jeremy Fitzhardinge 已提交
268 269 270 271 272 273 274 275 276
		~sh->evtchn_mask[idx]);
}

static void bind_evtchn_to_cpu(unsigned int chn, unsigned int cpu)
{
	int irq = evtchn_to_irq[chn];

	BUG_ON(irq == -1);
#ifdef CONFIG_SMP
277
	cpumask_copy(irq_to_desc(irq)->affinity, cpumask_of(cpu));
J
Jeremy Fitzhardinge 已提交
278 279
#endif

280 281
	clear_bit(chn, cpu_evtchn_mask(cpu_from_irq(irq)));
	set_bit(chn, cpu_evtchn_mask(cpu));
J
Jeremy Fitzhardinge 已提交
282

283
	irq_info[irq].cpu = cpu;
J
Jeremy Fitzhardinge 已提交
284 285 286 287
}

static void init_evtchn_cpu_bindings(void)
{
288
	int i;
J
Jeremy Fitzhardinge 已提交
289
#ifdef CONFIG_SMP
290 291
	struct irq_desc *desc;

J
Jeremy Fitzhardinge 已提交
292
	/* By default all event channels notify CPU#0. */
293
	for_each_irq_desc(i, desc) {
294
		cpumask_copy(desc->affinity, cpumask_of(0));
295
	}
J
Jeremy Fitzhardinge 已提交
296 297
#endif

298 299 300 301
	for_each_possible_cpu(i)
		memset(cpu_evtchn_mask(i),
		       (i == 0) ? ~0 : 0, sizeof(struct cpu_evtchn_s));

J
Jeremy Fitzhardinge 已提交
302 303 304 305 306 307 308 309 310 311 312 313 314 315
}

static inline void clear_evtchn(int port)
{
	struct shared_info *s = HYPERVISOR_shared_info;
	sync_clear_bit(port, &s->evtchn_pending[0]);
}

static inline void set_evtchn(int port)
{
	struct shared_info *s = HYPERVISOR_shared_info;
	sync_set_bit(port, &s->evtchn_pending[0]);
}

316 317 318 319 320 321
static inline int test_evtchn(int port)
{
	struct shared_info *s = HYPERVISOR_shared_info;
	return sync_test_bit(port, &s->evtchn_pending[0]);
}

J
Jeremy Fitzhardinge 已提交
322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375

/**
 * notify_remote_via_irq - send event to remote end of event channel via irq
 * @irq: irq of event channel to send event to
 *
 * Unlike notify_remote_via_evtchn(), this is safe to use across
 * save/restore. Notifications on a broken connection are silently
 * dropped.
 */
void notify_remote_via_irq(int irq)
{
	int evtchn = evtchn_from_irq(irq);

	if (VALID_EVTCHN(evtchn))
		notify_remote_via_evtchn(evtchn);
}
EXPORT_SYMBOL_GPL(notify_remote_via_irq);

static void mask_evtchn(int port)
{
	struct shared_info *s = HYPERVISOR_shared_info;
	sync_set_bit(port, &s->evtchn_mask[0]);
}

static void unmask_evtchn(int port)
{
	struct shared_info *s = HYPERVISOR_shared_info;
	unsigned int cpu = get_cpu();

	BUG_ON(!irqs_disabled());

	/* Slow path (hypercall) if this is a non-local port. */
	if (unlikely(cpu != cpu_from_evtchn(port))) {
		struct evtchn_unmask unmask = { .port = port };
		(void)HYPERVISOR_event_channel_op(EVTCHNOP_unmask, &unmask);
	} else {
		struct vcpu_info *vcpu_info = __get_cpu_var(xen_vcpu);

		sync_clear_bit(port, &s->evtchn_mask[0]);

		/*
		 * The following is basically the equivalent of
		 * 'hw_resend_irq'. Just like a real IO-APIC we 'lose
		 * the interrupt edge' if the channel is masked.
		 */
		if (sync_test_bit(port, &s->evtchn_pending[0]) &&
		    !sync_test_and_set_bit(port / BITS_PER_LONG,
					   &vcpu_info->evtchn_pending_sel))
			vcpu_info->evtchn_upcall_pending = 1;
	}

	put_cpu();
}

376 377 378 379 380 381 382 383 384 385 386
static int get_nr_hw_irqs(void)
{
	int ret = 1;

#ifdef CONFIG_X86_IO_APIC
	ret = get_nr_irqs_gsi();
#endif

	return ret;
}

387
static int find_unbound_pirq(int type)
S
Stefano Stabellini 已提交
388
{
389 390 391 392 393 394 395 396 397
	int rc, i;
	struct physdev_get_free_pirq op_get_free_pirq;
	op_get_free_pirq.type = type;

	rc = HYPERVISOR_physdev_op(PHYSDEVOP_get_free_pirq, &op_get_free_pirq);
	if (!rc)
		return op_get_free_pirq.pirq;

	for (i = 0; i < nr_irqs; i++) {
S
Stefano Stabellini 已提交
398 399 400 401 402 403
		if (pirq_to_irq[i] < 0)
			return i;
	}
	return -1;
}

J
Jeremy Fitzhardinge 已提交
404 405
static int find_unbound_irq(void)
{
406 407
	struct irq_data *data;
	int irq, res;
408 409
	int bottom = get_nr_hw_irqs();
	int top = nr_irqs-1;
J
Jeremy Fitzhardinge 已提交
410

411
	if (bottom == nr_irqs)
412 413
		goto no_irqs;

414 415 416 417 418 419 420
	/* This loop starts from the top of IRQ space and goes down.
	 * We need this b/c if we have a PCI device in a Xen PV guest
	 * we do not have an IO-APIC (though the backend might have them)
	 * mapped in. To not have a collision of physical IRQs with the Xen
	 * event channels start at the top of the IRQ space for virtual IRQs.
	 */
	for (irq = top; irq > bottom; irq--) {
421
		data = irq_get_irq_data(irq);
422
		/* only 15->0 have init'd desc; handle irq > 16 */
423
		if (!data)
424
			break;
425
		if (data->chip == &no_irq_chip)
426
			break;
427
		if (data->chip != &xen_dynamic_chip)
428
			continue;
J
Jeremy Fitzhardinge 已提交
429
		if (irq_info[irq].type == IRQT_UNBOUND)
430
			return irq;
431
	}
J
Jeremy Fitzhardinge 已提交
432

433
	if (irq == bottom)
434
		goto no_irqs;
J
Jeremy Fitzhardinge 已提交
435

436
	res = irq_alloc_desc_at(irq, -1);
437

438 439
	if (WARN_ON(res != irq))
		return -1;
440

J
Jeremy Fitzhardinge 已提交
441
	return irq;
442 443 444

no_irqs:
	panic("No available IRQ to bind to: increase nr_irqs!\n");
J
Jeremy Fitzhardinge 已提交
445 446
}

447 448
static bool identity_mapped_irq(unsigned irq)
{
449 450
	/* identity map all the hardware irqs */
	return irq < get_nr_hw_irqs();
451 452 453 454
}

static void pirq_unmask_notify(int irq)
{
S
Stefano Stabellini 已提交
455
	struct physdev_eoi eoi = { .irq = pirq_from_irq(irq) };
456 457 458 459 460 461 462 463 464 465 466 467 468 469

	if (unlikely(pirq_needs_eoi(irq))) {
		int rc = HYPERVISOR_physdev_op(PHYSDEVOP_eoi, &eoi);
		WARN_ON(rc);
	}
}

static void pirq_query_unmask(int irq)
{
	struct physdev_irq_status_query irq_status;
	struct irq_info *info = info_for_irq(irq);

	BUG_ON(info->type != IRQT_PIRQ);

S
Stefano Stabellini 已提交
470
	irq_status.irq = pirq_from_irq(irq);
471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490
	if (HYPERVISOR_physdev_op(PHYSDEVOP_irq_status_query, &irq_status))
		irq_status.flags = 0;

	info->u.pirq.flags &= ~PIRQ_NEEDS_EOI;
	if (irq_status.flags & XENIRQSTAT_needs_eoi)
		info->u.pirq.flags |= PIRQ_NEEDS_EOI;
}

static bool probing_irq(int irq)
{
	struct irq_desc *desc = irq_to_desc(irq);

	return desc && desc->action == NULL;
}

static unsigned int startup_pirq(unsigned int irq)
{
	struct evtchn_bind_pirq bind_pirq;
	struct irq_info *info = info_for_irq(irq);
	int evtchn = evtchn_from_irq(irq);
491
	int rc;
492 493 494 495 496 497

	BUG_ON(info->type != IRQT_PIRQ);

	if (VALID_EVTCHN(evtchn))
		goto out;

S
Stefano Stabellini 已提交
498
	bind_pirq.pirq = pirq_from_irq(irq);
499
	/* NB. We are happy to share unless we are probing. */
500 501 502 503
	bind_pirq.flags = info->u.pirq.flags & PIRQ_SHAREABLE ?
					BIND_PIRQ__WILL_SHARE : 0;
	rc = HYPERVISOR_event_channel_op(EVTCHNOP_bind_pirq, &bind_pirq);
	if (rc != 0) {
504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587
		if (!probing_irq(irq))
			printk(KERN_INFO "Failed to obtain physical IRQ %d\n",
			       irq);
		return 0;
	}
	evtchn = bind_pirq.port;

	pirq_query_unmask(irq);

	evtchn_to_irq[evtchn] = irq;
	bind_evtchn_to_cpu(evtchn, 0);
	info->evtchn = evtchn;

out:
	unmask_evtchn(evtchn);
	pirq_unmask_notify(irq);

	return 0;
}

static void shutdown_pirq(unsigned int irq)
{
	struct evtchn_close close;
	struct irq_info *info = info_for_irq(irq);
	int evtchn = evtchn_from_irq(irq);

	BUG_ON(info->type != IRQT_PIRQ);

	if (!VALID_EVTCHN(evtchn))
		return;

	mask_evtchn(evtchn);

	close.port = evtchn;
	if (HYPERVISOR_event_channel_op(EVTCHNOP_close, &close) != 0)
		BUG();

	bind_evtchn_to_cpu(evtchn, 0);
	evtchn_to_irq[evtchn] = -1;
	info->evtchn = 0;
}

static void enable_pirq(unsigned int irq)
{
	startup_pirq(irq);
}

static void disable_pirq(unsigned int irq)
{
}

static void ack_pirq(unsigned int irq)
{
	int evtchn = evtchn_from_irq(irq);

	move_native_irq(irq);

	if (VALID_EVTCHN(evtchn)) {
		mask_evtchn(evtchn);
		clear_evtchn(evtchn);
	}
}

static void end_pirq(unsigned int irq)
{
	int evtchn = evtchn_from_irq(irq);
	struct irq_desc *desc = irq_to_desc(irq);

	if (WARN_ON(!desc))
		return;

	if ((desc->status & (IRQ_DISABLED|IRQ_PENDING)) ==
	    (IRQ_DISABLED|IRQ_PENDING)) {
		shutdown_pirq(irq);
	} else if (VALID_EVTCHN(evtchn)) {
		unmask_evtchn(evtchn);
		pirq_unmask_notify(irq);
	}
}

static int find_irq_by_gsi(unsigned gsi)
{
	int irq;

588
	for (irq = 0; irq < nr_irqs; irq++) {
589 590 591 592 593 594 595 596 597 598 599 600
		struct irq_info *info = info_for_irq(irq);

		if (info == NULL || info->type != IRQT_PIRQ)
			continue;

		if (gsi_from_irq(irq) == gsi)
			return irq;
	}

	return -1;
}

S
Stefano Stabellini 已提交
601 602 603 604 605 606
int xen_allocate_pirq(unsigned gsi, int shareable, char *name)
{
	return xen_map_pirq_gsi(gsi, gsi, shareable, name);
}

/* xen_map_pirq_gsi might allocate irqs from the top down, as a
607 608 609
 * consequence don't assume that the irq number returned has a low value
 * or can be used as a pirq number unless you know otherwise.
 *
S
Stefano Stabellini 已提交
610
 * One notable exception is when xen_map_pirq_gsi is called passing an
611
 * hardware gsi as argument, in that case the irq number returned
S
Stefano Stabellini 已提交
612 613 614 615
 * matches the gsi number passed as second argument.
 *
 * Note: We don't assign an event channel until the irq actually started
 * up.  Return an existing irq if we've already got one for the gsi.
616
 */
S
Stefano Stabellini 已提交
617
int xen_map_pirq_gsi(unsigned pirq, unsigned gsi, int shareable, char *name)
618
{
S
Stefano Stabellini 已提交
619
	int irq = 0;
620 621 622 623
	struct physdev_irq irq_op;

	spin_lock(&irq_mapping_update_lock);

624
	if ((pirq > nr_irqs) || (gsi > nr_irqs)) {
625
		printk(KERN_WARNING "xen_map_pirq_gsi: %s %s is incorrect!\n",
626 627
			pirq > nr_irqs ? "pirq" :"",
			gsi > nr_irqs ? "gsi" : "");
628 629 630
		goto out;
	}

631 632
	irq = find_irq_by_gsi(gsi);
	if (irq != -1) {
S
Stefano Stabellini 已提交
633
		printk(KERN_INFO "xen_map_pirq_gsi: returning irq %d for gsi %u\n",
634 635 636 637
		       irq, gsi);
		goto out;	/* XXX need refcount? */
	}

638 639
	/* If we are a PV guest, we don't have GSIs (no ACPI passed). Therefore
	 * we are using the !xen_initial_domain() to drop in the function.*/
640 641
	if (identity_mapped_irq(gsi) || (!xen_initial_domain() &&
				xen_pv_domain())) {
642
		irq = gsi;
643
		irq_alloc_desc_at(irq, -1);
644 645 646 647
	} else
		irq = find_unbound_irq();

	set_irq_chip_and_handler_name(irq, &xen_pirq_chip,
648
				      handle_level_irq, name);
649 650

	irq_op.irq = irq;
651 652 653 654 655 656 657
	irq_op.vector = 0;

	/* Only the privileged domain can do this. For non-priv, the pcifront
	 * driver provides a PCI bus that does the call to do exactly
	 * this in the priv domain. */
	if (xen_initial_domain() &&
	    HYPERVISOR_physdev_op(PHYSDEVOP_alloc_irq_vector, &irq_op)) {
658
		irq_free_desc(irq);
659 660 661 662
		irq = -ENOSPC;
		goto out;
	}

S
Stefano Stabellini 已提交
663
	irq_info[irq] = mk_pirq_info(0, pirq, gsi, irq_op.vector);
664
	irq_info[irq].u.pirq.flags |= shareable ? PIRQ_SHAREABLE : 0;
S
Stefano Stabellini 已提交
665
	pirq_to_irq[pirq] = irq;
666 667 668 669 670 671 672

out:
	spin_unlock(&irq_mapping_update_lock);

	return irq;
}

673 674 675 676
#ifdef CONFIG_PCI_MSI
#include <linux/msi.h>
#include "../pci/msi.h"

677
void xen_allocate_pirq_msi(char *name, int *irq, int *pirq, int alloc)
S
Stefano Stabellini 已提交
678 679 680
{
	spin_lock(&irq_mapping_update_lock);

681 682 683 684 685
	if (alloc & XEN_ALLOC_IRQ) {
		*irq = find_unbound_irq();
		if (*irq == -1)
			goto out;
	}
S
Stefano Stabellini 已提交
686

687 688 689 690 691
	if (alloc & XEN_ALLOC_PIRQ) {
		*pirq = find_unbound_pirq(MAP_PIRQ_TYPE_MSI);
		if (*pirq == -1)
			goto out;
	}
S
Stefano Stabellini 已提交
692 693 694 695 696 697 698 699 700 701 702

	set_irq_chip_and_handler_name(*irq, &xen_pirq_chip,
				      handle_level_irq, name);

	irq_info[*irq] = mk_pirq_info(0, *pirq, 0, 0);
	pirq_to_irq[*pirq] = *irq;

out:
	spin_unlock(&irq_mapping_update_lock);
}

703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757
int xen_create_msi_irq(struct pci_dev *dev, struct msi_desc *msidesc, int type)
{
	int irq = -1;
	struct physdev_map_pirq map_irq;
	int rc;
	int pos;
	u32 table_offset, bir;

	memset(&map_irq, 0, sizeof(map_irq));
	map_irq.domid = DOMID_SELF;
	map_irq.type = MAP_PIRQ_TYPE_MSI;
	map_irq.index = -1;
	map_irq.pirq = -1;
	map_irq.bus = dev->bus->number;
	map_irq.devfn = dev->devfn;

	if (type == PCI_CAP_ID_MSIX) {
		pos = pci_find_capability(dev, PCI_CAP_ID_MSIX);

		pci_read_config_dword(dev, msix_table_offset_reg(pos),
					&table_offset);
		bir = (u8)(table_offset & PCI_MSIX_FLAGS_BIRMASK);

		map_irq.table_base = pci_resource_start(dev, bir);
		map_irq.entry_nr = msidesc->msi_attrib.entry_nr;
	}

	spin_lock(&irq_mapping_update_lock);

	irq = find_unbound_irq();

	if (irq == -1)
		goto out;

	rc = HYPERVISOR_physdev_op(PHYSDEVOP_map_pirq, &map_irq);
	if (rc) {
		printk(KERN_WARNING "xen map irq failed %d\n", rc);

		irq_free_desc(irq);

		irq = -1;
		goto out;
	}
	irq_info[irq] = mk_pirq_info(0, map_irq.pirq, 0, map_irq.index);

	set_irq_chip_and_handler_name(irq, &xen_pirq_chip,
			handle_level_irq,
			(type == PCI_CAP_ID_MSIX) ? "msi-x":"msi");

out:
	spin_unlock(&irq_mapping_update_lock);
	return irq;
}
#endif

758 759 760
int xen_destroy_irq(int irq)
{
	struct irq_desc *desc;
761 762
	struct physdev_unmap_pirq unmap_irq;
	struct irq_info *info = info_for_irq(irq);
763 764 765 766 767 768 769 770
	int rc = -ENOENT;

	spin_lock(&irq_mapping_update_lock);

	desc = irq_to_desc(irq);
	if (!desc)
		goto out;

771
	if (xen_initial_domain()) {
772
		unmap_irq.pirq = info->u.pirq.pirq;
773 774 775 776 777 778
		unmap_irq.domid = DOMID_SELF;
		rc = HYPERVISOR_physdev_op(PHYSDEVOP_unmap_pirq, &unmap_irq);
		if (rc) {
			printk(KERN_WARNING "unmap irq failed %d\n", rc);
			goto out;
		}
779
		pirq_to_irq[info->u.pirq.pirq] = -1;
780
	}
781 782
	irq_info[irq] = mk_unbound_info();

783
	irq_free_desc(irq);
784 785 786 787 788 789

out:
	spin_unlock(&irq_mapping_update_lock);
	return rc;
}

790 791 792 793 794 795 796 797
int xen_vector_from_irq(unsigned irq)
{
	return vector_from_irq(irq);
}

int xen_gsi_from_irq(unsigned irq)
{
	return gsi_from_irq(irq);
J
Jeremy Fitzhardinge 已提交
798 799
}

800 801 802 803 804
int xen_irq_from_pirq(unsigned pirq)
{
	return pirq_to_irq[pirq];
}

805
int bind_evtchn_to_irq(unsigned int evtchn)
J
Jeremy Fitzhardinge 已提交
806 807 808 809 810 811 812 813 814 815 816
{
	int irq;

	spin_lock(&irq_mapping_update_lock);

	irq = evtchn_to_irq[evtchn];

	if (irq == -1) {
		irq = find_unbound_irq();

		set_irq_chip_and_handler_name(irq, &xen_dynamic_chip,
817
					      handle_fasteoi_irq, "event");
J
Jeremy Fitzhardinge 已提交
818 819

		evtchn_to_irq[evtchn] = irq;
820
		irq_info[irq] = mk_evtchn_info(evtchn);
J
Jeremy Fitzhardinge 已提交
821 822 823 824 825 826
	}

	spin_unlock(&irq_mapping_update_lock);

	return irq;
}
827
EXPORT_SYMBOL_GPL(bind_evtchn_to_irq);
J
Jeremy Fitzhardinge 已提交
828

J
Jeremy Fitzhardinge 已提交
829 830 831 832 833 834 835 836
static int bind_ipi_to_irq(unsigned int ipi, unsigned int cpu)
{
	struct evtchn_bind_ipi bind_ipi;
	int evtchn, irq;

	spin_lock(&irq_mapping_update_lock);

	irq = per_cpu(ipi_to_irq, cpu)[ipi];
837

J
Jeremy Fitzhardinge 已提交
838 839 840 841 842
	if (irq == -1) {
		irq = find_unbound_irq();
		if (irq < 0)
			goto out;

843 844
		set_irq_chip_and_handler_name(irq, &xen_percpu_chip,
					      handle_percpu_irq, "ipi");
J
Jeremy Fitzhardinge 已提交
845 846 847 848 849 850 851 852

		bind_ipi.vcpu = cpu;
		if (HYPERVISOR_event_channel_op(EVTCHNOP_bind_ipi,
						&bind_ipi) != 0)
			BUG();
		evtchn = bind_ipi.port;

		evtchn_to_irq[evtchn] = irq;
853
		irq_info[irq] = mk_ipi_info(evtchn, ipi);
J
Jeremy Fitzhardinge 已提交
854 855 856 857 858 859 860 861 862 863 864
		per_cpu(ipi_to_irq, cpu)[ipi] = irq;

		bind_evtchn_to_cpu(evtchn, cpu);
	}

 out:
	spin_unlock(&irq_mapping_update_lock);
	return irq;
}


865
int bind_virq_to_irq(unsigned int virq, unsigned int cpu)
J
Jeremy Fitzhardinge 已提交
866 867 868 869 870 871 872 873 874
{
	struct evtchn_bind_virq bind_virq;
	int evtchn, irq;

	spin_lock(&irq_mapping_update_lock);

	irq = per_cpu(virq_to_irq, cpu)[virq];

	if (irq == -1) {
875 876 877 878 879
		irq = find_unbound_irq();

		set_irq_chip_and_handler_name(irq, &xen_percpu_chip,
					      handle_percpu_irq, "virq");

J
Jeremy Fitzhardinge 已提交
880 881 882 883 884 885 886 887
		bind_virq.virq = virq;
		bind_virq.vcpu = cpu;
		if (HYPERVISOR_event_channel_op(EVTCHNOP_bind_virq,
						&bind_virq) != 0)
			BUG();
		evtchn = bind_virq.port;

		evtchn_to_irq[evtchn] = irq;
888
		irq_info[irq] = mk_virq_info(evtchn, virq);
J
Jeremy Fitzhardinge 已提交
889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906

		per_cpu(virq_to_irq, cpu)[virq] = irq;

		bind_evtchn_to_cpu(evtchn, cpu);
	}

	spin_unlock(&irq_mapping_update_lock);

	return irq;
}

static void unbind_from_irq(unsigned int irq)
{
	struct evtchn_close close;
	int evtchn = evtchn_from_irq(irq);

	spin_lock(&irq_mapping_update_lock);

J
Jeremy Fitzhardinge 已提交
907
	if (VALID_EVTCHN(evtchn)) {
J
Jeremy Fitzhardinge 已提交
908 909 910 911 912 913 914
		close.port = evtchn;
		if (HYPERVISOR_event_channel_op(EVTCHNOP_close, &close) != 0)
			BUG();

		switch (type_from_irq(irq)) {
		case IRQT_VIRQ:
			per_cpu(virq_to_irq, cpu_from_evtchn(evtchn))
915
				[virq_from_irq(irq)] = -1;
J
Jeremy Fitzhardinge 已提交
916
			break;
A
Alex Nixon 已提交
917 918
		case IRQT_IPI:
			per_cpu(ipi_to_irq, cpu_from_evtchn(evtchn))
919
				[ipi_from_irq(irq)] = -1;
A
Alex Nixon 已提交
920
			break;
J
Jeremy Fitzhardinge 已提交
921 922 923 924 925 926 927 928
		default:
			break;
		}

		/* Closed ports are implicitly re-bound to VCPU0. */
		bind_evtchn_to_cpu(evtchn, 0);

		evtchn_to_irq[evtchn] = -1;
929 930 931
	}

	if (irq_info[irq].type != IRQT_UNBOUND) {
932
		irq_info[irq] = mk_unbound_info();
J
Jeremy Fitzhardinge 已提交
933

934
		irq_free_desc(irq);
J
Jeremy Fitzhardinge 已提交
935 936 937 938 939 940
	}

	spin_unlock(&irq_mapping_update_lock);
}

int bind_evtchn_to_irqhandler(unsigned int evtchn,
941
			      irq_handler_t handler,
J
Jeremy Fitzhardinge 已提交
942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959
			      unsigned long irqflags,
			      const char *devname, void *dev_id)
{
	unsigned int irq;
	int retval;

	irq = bind_evtchn_to_irq(evtchn);
	retval = request_irq(irq, handler, irqflags, devname, dev_id);
	if (retval != 0) {
		unbind_from_irq(irq);
		return retval;
	}

	return irq;
}
EXPORT_SYMBOL_GPL(bind_evtchn_to_irqhandler);

int bind_virq_to_irqhandler(unsigned int virq, unsigned int cpu,
960
			    irq_handler_t handler,
J
Jeremy Fitzhardinge 已提交
961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976
			    unsigned long irqflags, const char *devname, void *dev_id)
{
	unsigned int irq;
	int retval;

	irq = bind_virq_to_irq(virq, cpu);
	retval = request_irq(irq, handler, irqflags, devname, dev_id);
	if (retval != 0) {
		unbind_from_irq(irq);
		return retval;
	}

	return irq;
}
EXPORT_SYMBOL_GPL(bind_virq_to_irqhandler);

J
Jeremy Fitzhardinge 已提交
977 978 979 980 981 982 983 984 985 986 987 988 989
int bind_ipi_to_irqhandler(enum ipi_vector ipi,
			   unsigned int cpu,
			   irq_handler_t handler,
			   unsigned long irqflags,
			   const char *devname,
			   void *dev_id)
{
	int irq, retval;

	irq = bind_ipi_to_irq(ipi, cpu);
	if (irq < 0)
		return irq;

I
Ian Campbell 已提交
990
	irqflags |= IRQF_NO_SUSPEND;
J
Jeremy Fitzhardinge 已提交
991 992 993 994 995 996 997 998 999
	retval = request_irq(irq, handler, irqflags, devname, dev_id);
	if (retval != 0) {
		unbind_from_irq(irq);
		return retval;
	}

	return irq;
}

J
Jeremy Fitzhardinge 已提交
1000 1001 1002 1003 1004 1005 1006
void unbind_from_irqhandler(unsigned int irq, void *dev_id)
{
	free_irq(irq, dev_id);
	unbind_from_irq(irq);
}
EXPORT_SYMBOL_GPL(unbind_from_irqhandler);

J
Jeremy Fitzhardinge 已提交
1007 1008 1009 1010 1011 1012 1013
void xen_send_IPI_one(unsigned int cpu, enum ipi_vector vector)
{
	int irq = per_cpu(ipi_to_irq, cpu)[vector];
	BUG_ON(irq < 0);
	notify_remote_via_irq(irq);
}

1014 1015 1016 1017
irqreturn_t xen_debug_interrupt(int irq, void *dev_id)
{
	struct shared_info *sh = HYPERVISOR_shared_info;
	int cpu = smp_processor_id();
1018
	unsigned long *cpu_evtchn = cpu_evtchn_mask(cpu);
1019 1020 1021
	int i;
	unsigned long flags;
	static DEFINE_SPINLOCK(debug_lock);
1022
	struct vcpu_info *v;
1023 1024 1025

	spin_lock_irqsave(&debug_lock, flags);

1026
	printk("\nvcpu %d\n  ", cpu);
1027 1028

	for_each_online_cpu(i) {
1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071
		int pending;
		v = per_cpu(xen_vcpu, i);
		pending = (get_irq_regs() && i == cpu)
			? xen_irqs_disabled(get_irq_regs())
			: v->evtchn_upcall_mask;
		printk("%d: masked=%d pending=%d event_sel %0*lx\n  ", i,
		       pending, v->evtchn_upcall_pending,
		       (int)(sizeof(v->evtchn_pending_sel)*2),
		       v->evtchn_pending_sel);
	}
	v = per_cpu(xen_vcpu, cpu);

	printk("\npending:\n   ");
	for (i = ARRAY_SIZE(sh->evtchn_pending)-1; i >= 0; i--)
		printk("%0*lx%s", (int)sizeof(sh->evtchn_pending[0])*2,
		       sh->evtchn_pending[i],
		       i % 8 == 0 ? "\n   " : " ");
	printk("\nglobal mask:\n   ");
	for (i = ARRAY_SIZE(sh->evtchn_mask)-1; i >= 0; i--)
		printk("%0*lx%s",
		       (int)(sizeof(sh->evtchn_mask[0])*2),
		       sh->evtchn_mask[i],
		       i % 8 == 0 ? "\n   " : " ");

	printk("\nglobally unmasked:\n   ");
	for (i = ARRAY_SIZE(sh->evtchn_mask)-1; i >= 0; i--)
		printk("%0*lx%s", (int)(sizeof(sh->evtchn_mask[0])*2),
		       sh->evtchn_pending[i] & ~sh->evtchn_mask[i],
		       i % 8 == 0 ? "\n   " : " ");

	printk("\nlocal cpu%d mask:\n   ", cpu);
	for (i = (NR_EVENT_CHANNELS/BITS_PER_LONG)-1; i >= 0; i--)
		printk("%0*lx%s", (int)(sizeof(cpu_evtchn[0])*2),
		       cpu_evtchn[i],
		       i % 8 == 0 ? "\n   " : " ");

	printk("\nlocally unmasked:\n   ");
	for (i = ARRAY_SIZE(sh->evtchn_mask)-1; i >= 0; i--) {
		unsigned long pending = sh->evtchn_pending[i]
			& ~sh->evtchn_mask[i]
			& cpu_evtchn[i];
		printk("%0*lx%s", (int)(sizeof(sh->evtchn_mask[0])*2),
		       pending, i % 8 == 0 ? "\n   " : " ");
1072 1073 1074
	}

	printk("\npending list:\n");
1075
	for (i = 0; i < NR_EVENT_CHANNELS; i++) {
1076
		if (sync_test_bit(i, sh->evtchn_pending)) {
1077 1078
			int word_idx = i / BITS_PER_LONG;
			printk("  %d: event %d -> irq %d%s%s%s\n",
1079
			       cpu_from_evtchn(i), i,
1080 1081 1082 1083 1084 1085 1086
			       evtchn_to_irq[i],
			       sync_test_bit(word_idx, &v->evtchn_pending_sel)
					     ? "" : " l2-clear",
			       !sync_test_bit(i, sh->evtchn_mask)
					     ? "" : " globally-masked",
			       sync_test_bit(i, cpu_evtchn)
					     ? "" : " locally-masked");
1087 1088 1089 1090 1091 1092 1093 1094
		}
	}

	spin_unlock_irqrestore(&debug_lock, flags);

	return IRQ_HANDLED;
}

1095 1096
static DEFINE_PER_CPU(unsigned, xed_nesting_count);

J
Jeremy Fitzhardinge 已提交
1097 1098 1099 1100 1101 1102 1103 1104 1105
/*
 * Search the CPUs pending events bitmasks.  For each one found, map
 * the event number to an irq, and feed it into do_IRQ() for
 * handling.
 *
 * Xen uses a two-level bitmap to speed searching.  The first level is
 * a bitset of words which contain pending event bits.  The second
 * level is a bitset of pending events themselves.
 */
1106
static void __xen_evtchn_do_upcall(void)
J
Jeremy Fitzhardinge 已提交
1107 1108 1109 1110
{
	int cpu = get_cpu();
	struct shared_info *s = HYPERVISOR_shared_info;
	struct vcpu_info *vcpu_info = __get_cpu_var(xen_vcpu);
1111
 	unsigned count;
J
Jeremy Fitzhardinge 已提交
1112

1113 1114
	do {
		unsigned long pending_words;
J
Jeremy Fitzhardinge 已提交
1115

1116
		vcpu_info->evtchn_upcall_pending = 0;
J
Jeremy Fitzhardinge 已提交
1117

1118
		if (__get_cpu_var(xed_nesting_count)++)
1119
			goto out;
J
Jeremy Fitzhardinge 已提交
1120

1121 1122
#ifndef CONFIG_X86 /* No need for a barrier -- XCHG is a barrier on x86. */
		/* Clear master flag /before/ clearing selector flag. */
1123
		wmb();
1124
#endif
1125 1126 1127 1128 1129 1130 1131 1132 1133 1134
		pending_words = xchg(&vcpu_info->evtchn_pending_sel, 0);
		while (pending_words != 0) {
			unsigned long pending_bits;
			int word_idx = __ffs(pending_words);
			pending_words &= ~(1UL << word_idx);

			while ((pending_bits = active_evtchns(cpu, s, word_idx)) != 0) {
				int bit_idx = __ffs(pending_bits);
				int port = (word_idx * BITS_PER_LONG) + bit_idx;
				int irq = evtchn_to_irq[port];
1135
				struct irq_desc *desc;
1136

1137 1138 1139
				mask_evtchn(port);
				clear_evtchn(port);

1140 1141 1142 1143 1144
				if (irq != -1) {
					desc = irq_to_desc(irq);
					if (desc)
						generic_handle_irq_desc(irq, desc);
				}
J
Jeremy Fitzhardinge 已提交
1145 1146 1147
			}
		}

1148 1149
		BUG_ON(!irqs_disabled());

1150 1151
		count = __get_cpu_var(xed_nesting_count);
		__get_cpu_var(xed_nesting_count) = 0;
1152
	} while (count != 1 || vcpu_info->evtchn_upcall_pending);
1153 1154

out:
1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167

	put_cpu();
}

void xen_evtchn_do_upcall(struct pt_regs *regs)
{
	struct pt_regs *old_regs = set_irq_regs(regs);

	exit_idle();
	irq_enter();

	__xen_evtchn_do_upcall();

1168 1169
	irq_exit();
	set_irq_regs(old_regs);
1170
}
1171

1172 1173 1174
void xen_hvm_evtchn_do_upcall(void)
{
	__xen_evtchn_do_upcall();
J
Jeremy Fitzhardinge 已提交
1175
}
1176
EXPORT_SYMBOL_GPL(xen_hvm_evtchn_do_upcall);
J
Jeremy Fitzhardinge 已提交
1177

J
Jeremy Fitzhardinge 已提交
1178 1179 1180
/* Rebind a new event channel to an existing irq. */
void rebind_evtchn_irq(int evtchn, int irq)
{
J
Jeremy Fitzhardinge 已提交
1181 1182
	struct irq_info *info = info_for_irq(irq);

J
Jeremy Fitzhardinge 已提交
1183 1184 1185 1186 1187 1188 1189 1190 1191
	/* Make sure the irq is masked, since the new event channel
	   will also be masked. */
	disable_irq(irq);

	spin_lock(&irq_mapping_update_lock);

	/* After resume the irq<->evtchn mappings are all cleared out */
	BUG_ON(evtchn_to_irq[evtchn] != -1);
	/* Expect irq to have been bound before,
J
Jeremy Fitzhardinge 已提交
1192 1193
	   so there should be a proper type */
	BUG_ON(info->type == IRQT_UNBOUND);
J
Jeremy Fitzhardinge 已提交
1194 1195

	evtchn_to_irq[evtchn] = irq;
1196
	irq_info[irq] = mk_evtchn_info(evtchn);
J
Jeremy Fitzhardinge 已提交
1197 1198 1199 1200

	spin_unlock(&irq_mapping_update_lock);

	/* new event channels are always bound to cpu 0 */
1201
	irq_set_affinity(irq, cpumask_of(0));
J
Jeremy Fitzhardinge 已提交
1202 1203 1204 1205 1206

	/* Unmask the event channel. */
	enable_irq(irq);
}

J
Jeremy Fitzhardinge 已提交
1207
/* Rebind an evtchn so that it gets delivered to a specific cpu */
1208
static int rebind_irq_to_cpu(unsigned irq, unsigned tcpu)
J
Jeremy Fitzhardinge 已提交
1209 1210 1211 1212
{
	struct evtchn_bind_vcpu bind_vcpu;
	int evtchn = evtchn_from_irq(irq);

1213 1214 1215 1216
	/* events delivered via platform PCI interrupts are always
	 * routed to vcpu 0 */
	if (!VALID_EVTCHN(evtchn) ||
		(xen_hvm_domain() && !xen_have_vector_callback))
1217
		return -1;
J
Jeremy Fitzhardinge 已提交
1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230

	/* Send future instances of this interrupt to other vcpu. */
	bind_vcpu.port = evtchn;
	bind_vcpu.vcpu = tcpu;

	/*
	 * If this fails, it usually just indicates that we're dealing with a
	 * virq or IPI channel, which don't actually need to be rebound. Ignore
	 * it, but don't do the xenlinux-level rebind in that case.
	 */
	if (HYPERVISOR_event_channel_op(EVTCHNOP_bind_vcpu, &bind_vcpu) >= 0)
		bind_evtchn_to_cpu(evtchn, tcpu);

1231 1232
	return 0;
}
J
Jeremy Fitzhardinge 已提交
1233

1234
static int set_affinity_irq(unsigned irq, const struct cpumask *dest)
J
Jeremy Fitzhardinge 已提交
1235
{
1236
	unsigned tcpu = cpumask_first(dest);
1237 1238

	return rebind_irq_to_cpu(irq, tcpu);
J
Jeremy Fitzhardinge 已提交
1239 1240
}

1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256
int resend_irq_on_evtchn(unsigned int irq)
{
	int masked, evtchn = evtchn_from_irq(irq);
	struct shared_info *s = HYPERVISOR_shared_info;

	if (!VALID_EVTCHN(evtchn))
		return 1;

	masked = sync_test_and_set_bit(evtchn, s->evtchn_mask);
	sync_set_bit(evtchn, s->evtchn_pending);
	if (!masked)
		unmask_evtchn(evtchn);

	return 1;
}

J
Jeremy Fitzhardinge 已提交
1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276
static void enable_dynirq(unsigned int irq)
{
	int evtchn = evtchn_from_irq(irq);

	if (VALID_EVTCHN(evtchn))
		unmask_evtchn(evtchn);
}

static void disable_dynirq(unsigned int irq)
{
	int evtchn = evtchn_from_irq(irq);

	if (VALID_EVTCHN(evtchn))
		mask_evtchn(evtchn);
}

static void ack_dynirq(unsigned int irq)
{
	int evtchn = evtchn_from_irq(irq);

1277
	move_masked_irq(irq);
J
Jeremy Fitzhardinge 已提交
1278 1279

	if (VALID_EVTCHN(evtchn))
1280
		unmask_evtchn(evtchn);
J
Jeremy Fitzhardinge 已提交
1281 1282 1283 1284 1285
}

static int retrigger_dynirq(unsigned int irq)
{
	int evtchn = evtchn_from_irq(irq);
1286
	struct shared_info *sh = HYPERVISOR_shared_info;
J
Jeremy Fitzhardinge 已提交
1287 1288 1289
	int ret = 0;

	if (VALID_EVTCHN(evtchn)) {
1290 1291 1292 1293 1294 1295
		int masked;

		masked = sync_test_and_set_bit(evtchn, sh->evtchn_mask);
		sync_set_bit(evtchn, sh->evtchn_pending);
		if (!masked)
			unmask_evtchn(evtchn);
J
Jeremy Fitzhardinge 已提交
1296 1297 1298 1299 1300 1301
		ret = 1;
	}

	return ret;
}

1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337
static void restore_cpu_pirqs(void)
{
	int pirq, rc, irq, gsi;
	struct physdev_map_pirq map_irq;

	for (pirq = 0; pirq < nr_irqs; pirq++) {
		irq = pirq_to_irq[pirq];
		if (irq == -1)
			continue;

		/* save/restore of PT devices doesn't work, so at this point the
		 * only devices present are GSI based emulated devices */
		gsi = gsi_from_irq(irq);
		if (!gsi)
			continue;

		map_irq.domid = DOMID_SELF;
		map_irq.type = MAP_PIRQ_TYPE_GSI;
		map_irq.index = gsi;
		map_irq.pirq = pirq;

		rc = HYPERVISOR_physdev_op(PHYSDEVOP_map_pirq, &map_irq);
		if (rc) {
			printk(KERN_WARNING "xen map irq failed gsi=%d irq=%d pirq=%d rc=%d\n",
					gsi, irq, pirq, rc);
			irq_info[irq] = mk_unbound_info();
			pirq_to_irq[pirq] = -1;
			continue;
		}

		printk(KERN_DEBUG "xen: --> irq=%d, pirq=%d\n", irq, map_irq.pirq);

		startup_pirq(irq);
	}
}

1338 1339 1340 1341 1342 1343 1344 1345 1346
static void restore_cpu_virqs(unsigned int cpu)
{
	struct evtchn_bind_virq bind_virq;
	int virq, irq, evtchn;

	for (virq = 0; virq < NR_VIRQS; virq++) {
		if ((irq = per_cpu(virq_to_irq, cpu)[virq]) == -1)
			continue;

1347
		BUG_ON(virq_from_irq(irq) != virq);
1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358

		/* Get a new binding from Xen. */
		bind_virq.virq = virq;
		bind_virq.vcpu = cpu;
		if (HYPERVISOR_event_channel_op(EVTCHNOP_bind_virq,
						&bind_virq) != 0)
			BUG();
		evtchn = bind_virq.port;

		/* Record the new mapping. */
		evtchn_to_irq[evtchn] = irq;
1359
		irq_info[irq] = mk_virq_info(evtchn, virq);
1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372
		bind_evtchn_to_cpu(evtchn, cpu);
	}
}

static void restore_cpu_ipis(unsigned int cpu)
{
	struct evtchn_bind_ipi bind_ipi;
	int ipi, irq, evtchn;

	for (ipi = 0; ipi < XEN_NR_IPIS; ipi++) {
		if ((irq = per_cpu(ipi_to_irq, cpu)[ipi]) == -1)
			continue;

1373
		BUG_ON(ipi_from_irq(irq) != ipi);
1374 1375 1376 1377 1378 1379 1380 1381 1382 1383

		/* Get a new binding from Xen. */
		bind_ipi.vcpu = cpu;
		if (HYPERVISOR_event_channel_op(EVTCHNOP_bind_ipi,
						&bind_ipi) != 0)
			BUG();
		evtchn = bind_ipi.port;

		/* Record the new mapping. */
		evtchn_to_irq[evtchn] = irq;
1384
		irq_info[irq] = mk_ipi_info(evtchn, ipi);
1385 1386 1387 1388
		bind_evtchn_to_cpu(evtchn, cpu);
	}
}

1389 1390 1391 1392 1393 1394 1395 1396
/* Clear an irq's pending state, in preparation for polling on it */
void xen_clear_irq_pending(int irq)
{
	int evtchn = evtchn_from_irq(irq);

	if (VALID_EVTCHN(evtchn))
		clear_evtchn(evtchn);
}
1397
EXPORT_SYMBOL(xen_clear_irq_pending);
1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416
void xen_set_irq_pending(int irq)
{
	int evtchn = evtchn_from_irq(irq);

	if (VALID_EVTCHN(evtchn))
		set_evtchn(evtchn);
}

bool xen_test_irq_pending(int irq)
{
	int evtchn = evtchn_from_irq(irq);
	bool ret = false;

	if (VALID_EVTCHN(evtchn))
		ret = test_evtchn(evtchn);

	return ret;
}

1417 1418 1419
/* Poll waiting for an irq to become pending with timeout.  In the usual case,
 * the irq will be disabled so it won't deliver an interrupt. */
void xen_poll_irq_timeout(int irq, u64 timeout)
1420 1421 1422 1423 1424 1425 1426
{
	evtchn_port_t evtchn = evtchn_from_irq(irq);

	if (VALID_EVTCHN(evtchn)) {
		struct sched_poll poll;

		poll.nr_ports = 1;
1427
		poll.timeout = timeout;
1428
		set_xen_guest_handle(poll.ports, &evtchn);
1429 1430 1431 1432 1433

		if (HYPERVISOR_sched_op(SCHEDOP_poll, &poll) != 0)
			BUG();
	}
}
1434 1435 1436 1437 1438 1439 1440
EXPORT_SYMBOL(xen_poll_irq_timeout);
/* Poll waiting for an irq to become pending.  In the usual case, the
 * irq will be disabled so it won't deliver an interrupt. */
void xen_poll_irq(int irq)
{
	xen_poll_irq_timeout(irq, 0 /* no timeout */);
}
1441

1442 1443 1444
void xen_irq_resume(void)
{
	unsigned int cpu, irq, evtchn;
1445
	struct irq_desc *desc;
1446 1447 1448 1449 1450 1451 1452 1453

	init_evtchn_cpu_bindings();

	/* New event-channel space is not 'live' yet. */
	for (evtchn = 0; evtchn < NR_EVENT_CHANNELS; evtchn++)
		mask_evtchn(evtchn);

	/* No IRQ <-> event-channel mappings. */
1454
	for (irq = 0; irq < nr_irqs; irq++)
1455 1456 1457 1458 1459 1460 1461 1462 1463
		irq_info[irq].evtchn = 0; /* zap event-channel binding */

	for (evtchn = 0; evtchn < NR_EVENT_CHANNELS; evtchn++)
		evtchn_to_irq[evtchn] = -1;

	for_each_possible_cpu(cpu) {
		restore_cpu_virqs(cpu);
		restore_cpu_ipis(cpu);
	}
1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480

	/*
	 * Unmask any IRQF_NO_SUSPEND IRQs which are enabled. These
	 * are not handled by the IRQ core.
	 */
	for_each_irq_desc(irq, desc) {
		if (!desc->action || !(desc->action->flags & IRQF_NO_SUSPEND))
			continue;
		if (desc->status & IRQ_DISABLED)
			continue;

		evtchn = evtchn_from_irq(irq);
		if (evtchn == -1)
			continue;

		unmask_evtchn(evtchn);
	}
1481 1482

	restore_cpu_pirqs();
1483 1484
}

J
Jeremy Fitzhardinge 已提交
1485 1486
static struct irq_chip xen_dynamic_chip __read_mostly = {
	.name		= "xen-dyn",
J
Jeremy Fitzhardinge 已提交
1487 1488

	.disable	= disable_dynirq,
J
Jeremy Fitzhardinge 已提交
1489 1490
	.mask		= disable_dynirq,
	.unmask		= enable_dynirq,
J
Jeremy Fitzhardinge 已提交
1491

1492
	.eoi		= ack_dynirq,
J
Jeremy Fitzhardinge 已提交
1493 1494 1495 1496
	.set_affinity	= set_affinity_irq,
	.retrigger	= retrigger_dynirq,
};

1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516
static struct irq_chip xen_pirq_chip __read_mostly = {
	.name		= "xen-pirq",

	.startup	= startup_pirq,
	.shutdown	= shutdown_pirq,

	.enable		= enable_pirq,
	.unmask		= enable_pirq,

	.disable	= disable_pirq,
	.mask		= disable_pirq,

	.ack		= ack_pirq,
	.end		= end_pirq,

	.set_affinity	= set_affinity_irq,

	.retrigger	= retrigger_dynirq,
};

1517 1518 1519 1520 1521 1522 1523 1524 1525 1526
static struct irq_chip xen_percpu_chip __read_mostly = {
	.name		= "xen-percpu",

	.disable	= disable_dynirq,
	.mask		= disable_dynirq,
	.unmask		= enable_dynirq,

	.ack		= ack_dynirq,
};

1527 1528 1529 1530 1531 1532 1533 1534 1535 1536
int xen_set_callback_via(uint64_t via)
{
	struct xen_hvm_param a;
	a.domid = DOMID_SELF;
	a.index = HVM_PARAM_CALLBACK_IRQ;
	a.value = via;
	return HYPERVISOR_hvm_op(HVMOP_set_param, &a);
}
EXPORT_SYMBOL_GPL(xen_set_callback_via);

1537
#ifdef CONFIG_XEN_PVHVM
1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560
/* Vector callbacks are better than PCI interrupts to receive event
 * channel notifications because we can receive vector callbacks on any
 * vcpu and we don't need PCI support or APIC interactions. */
void xen_callback_vector(void)
{
	int rc;
	uint64_t callback_via;
	if (xen_have_vector_callback) {
		callback_via = HVM_CALLBACK_VECTOR(XEN_HVM_EVTCHN_CALLBACK);
		rc = xen_set_callback_via(callback_via);
		if (rc) {
			printk(KERN_ERR "Request for Xen HVM callback vector"
					" failed.\n");
			xen_have_vector_callback = 0;
			return;
		}
		printk(KERN_INFO "Xen HVM callback vector for event delivery is "
				"enabled\n");
		/* in the restore case the vector has already been allocated */
		if (!test_bit(XEN_HVM_EVTCHN_CALLBACK, used_vectors))
			alloc_intr_gate(XEN_HVM_EVTCHN_CALLBACK, xen_hvm_callback_vector);
	}
}
1561 1562 1563
#else
void xen_callback_vector(void) {}
#endif
1564

J
Jeremy Fitzhardinge 已提交
1565 1566
void __init xen_init_IRQ(void)
{
1567
	int i;
1568

1569 1570
	cpu_evtchn_mask_p = kcalloc(nr_cpu_ids, sizeof(struct cpu_evtchn_s),
				    GFP_KERNEL);
1571 1572
	irq_info = kcalloc(nr_irqs, sizeof(*irq_info), GFP_KERNEL);

1573 1574 1575 1576 1577
	/* We are using nr_irqs as the maximum number of pirq available but
	 * that number is actually chosen by Xen and we don't know exactly
	 * what it is. Be careful choosing high pirq numbers. */
	pirq_to_irq = kcalloc(nr_irqs, sizeof(*pirq_to_irq), GFP_KERNEL);
	for (i = 0; i < nr_irqs; i++)
S
Stefano Stabellini 已提交
1578 1579
		pirq_to_irq[i] = -1;

1580 1581 1582 1583
	evtchn_to_irq = kcalloc(NR_EVENT_CHANNELS, sizeof(*evtchn_to_irq),
				    GFP_KERNEL);
	for (i = 0; i < NR_EVENT_CHANNELS; i++)
		evtchn_to_irq[i] = -1;
J
Jeremy Fitzhardinge 已提交
1584 1585 1586 1587 1588 1589 1590

	init_evtchn_cpu_bindings();

	/* No event channels are 'live' right now. */
	for (i = 0; i < NR_EVENT_CHANNELS; i++)
		mask_evtchn(i);

1591 1592 1593
	if (xen_hvm_domain()) {
		xen_callback_vector();
		native_init_IRQ();
1594 1595 1596
		/* pci_xen_hvm_init must be called after native_init_IRQ so that
		 * __acpi_register_gsi can point at the right function */
		pci_xen_hvm_init();
1597 1598
	} else {
		irq_ctx_init(smp_processor_id());
1599 1600
		if (xen_initial_domain())
			xen_setup_pirqs();
1601
	}
J
Jeremy Fitzhardinge 已提交
1602
}