vmbus_drv.c 61.6 KB
Newer Older
1
// SPDX-License-Identifier: GPL-2.0-only
2 3 4 5 6 7
/*
 * Copyright (c) 2009, Microsoft Corporation.
 *
 * Authors:
 *   Haiyang Zhang <haiyangz@microsoft.com>
 *   Hank Janssen  <hjanssen@microsoft.com>
8
 *   K. Y. Srinivasan <kys@microsoft.com>
9
 */
10 11
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt

12 13 14 15 16
#include <linux/init.h>
#include <linux/module.h>
#include <linux/device.h>
#include <linux/interrupt.h>
#include <linux/sysctl.h>
17
#include <linux/slab.h>
18
#include <linux/acpi.h>
19
#include <linux/completion.h>
20
#include <linux/hyperv.h>
21
#include <linux/kernel_stat.h>
22
#include <linux/clockchips.h>
23
#include <linux/cpu.h>
24 25
#include <linux/sched/task_stack.h>

26
#include <asm/mshyperv.h>
27
#include <linux/delay.h>
28 29
#include <linux/notifier.h>
#include <linux/ptrace.h>
30
#include <linux/screen_info.h>
31
#include <linux/kdebug.h>
32
#include <linux/efi.h>
33
#include <linux/random.h>
34
#include <linux/syscore_ops.h>
35
#include <clocksource/hyperv_timer.h>
36
#include "hyperv_vmbus.h"
37

38 39 40 41 42
struct vmbus_dynid {
	struct list_head node;
	struct hv_vmbus_device_id id;
};

43
static struct acpi_device  *hv_acpi_dev;
44

45
static struct completion probe_event;
46

47
static int hyperv_cpuhp_online;
48

49 50
static void *hv_panic_page;

51 52 53 54 55 56 57
static int hyperv_panic_event(struct notifier_block *nb, unsigned long val,
			      void *args)
{
	struct pt_regs *regs;

	regs = current_pt_regs();

58
	hyperv_report_panic(regs, val);
59 60 61
	return NOTIFY_DONE;
}

62 63 64 65 66 67
static int hyperv_die_event(struct notifier_block *nb, unsigned long val,
			    void *args)
{
	struct die_args *die = (struct die_args *)args;
	struct pt_regs *regs = die->regs;

68
	hyperv_report_panic(regs, val);
69 70 71 72 73 74
	return NOTIFY_DONE;
}

static struct notifier_block hyperv_die_block = {
	.notifier_call = hyperv_die_event,
};
75 76 77 78
static struct notifier_block hyperv_panic_block = {
	.notifier_call = hyperv_panic_event,
};

79 80
static const char *fb_mmio_name = "fb_range";
static struct resource *fb_mmio;
81 82
static struct resource *hyperv_mmio;
static DEFINE_SEMAPHORE(hyperv_mmio_lock);
83

84 85 86 87 88 89 90 91
static int vmbus_exists(void)
{
	if (hv_acpi_dev == NULL)
		return -ENODEV;

	return 0;
}

92 93 94 95 96 97 98 99
#define VMBUS_ALIAS_LEN ((sizeof((struct hv_vmbus_device_id *)0)->guid) * 2)
static void print_alias_name(struct hv_device *hv_dev, char *alias_name)
{
	int i;
	for (i = 0; i < VMBUS_ALIAS_LEN; i += 2)
		sprintf(&alias_name[i], "%02x", hv_dev->dev_type.b[i/2]);
}

100
static u8 channel_monitor_group(const struct vmbus_channel *channel)
101 102 103 104
{
	return (u8)channel->offermsg.monitorid / 32;
}

105
static u8 channel_monitor_offset(const struct vmbus_channel *channel)
106 107 108 109
{
	return (u8)channel->offermsg.monitorid % 32;
}

110 111
static u32 channel_pending(const struct vmbus_channel *channel,
			   const struct hv_monitor_page *monitor_page)
112 113
{
	u8 monitor_group = channel_monitor_group(channel);
114

115 116 117
	return monitor_page->trigger_group[monitor_group].pending;
}

118 119
static u32 channel_latency(const struct vmbus_channel *channel,
			   const struct hv_monitor_page *monitor_page)
120 121 122
{
	u8 monitor_group = channel_monitor_group(channel);
	u8 monitor_offset = channel_monitor_offset(channel);
123

124 125 126
	return monitor_page->latency[monitor_group][monitor_offset];
}

127 128 129 130 131 132 133 134
static u32 channel_conn_id(struct vmbus_channel *channel,
			   struct hv_monitor_page *monitor_page)
{
	u8 monitor_group = channel_monitor_group(channel);
	u8 monitor_offset = channel_monitor_offset(channel);
	return monitor_page->parameter[monitor_group][monitor_offset].connectionid.u.id;
}

135 136 137 138 139 140 141 142 143 144 145
static ssize_t id_show(struct device *dev, struct device_attribute *dev_attr,
		       char *buf)
{
	struct hv_device *hv_dev = device_to_hv_device(dev);

	if (!hv_dev->channel)
		return -ENODEV;
	return sprintf(buf, "%d\n", hv_dev->channel->offermsg.child_relid);
}
static DEVICE_ATTR_RO(id);

146 147 148 149 150 151 152 153 154 155 156
static ssize_t state_show(struct device *dev, struct device_attribute *dev_attr,
			  char *buf)
{
	struct hv_device *hv_dev = device_to_hv_device(dev);

	if (!hv_dev->channel)
		return -ENODEV;
	return sprintf(buf, "%d\n", hv_dev->channel->state);
}
static DEVICE_ATTR_RO(state);

157 158 159 160 161 162 163 164 165 166 167
static ssize_t monitor_id_show(struct device *dev,
			       struct device_attribute *dev_attr, char *buf)
{
	struct hv_device *hv_dev = device_to_hv_device(dev);

	if (!hv_dev->channel)
		return -ENODEV;
	return sprintf(buf, "%d\n", hv_dev->channel->offermsg.monitorid);
}
static DEVICE_ATTR_RO(monitor_id);

168 169 170 171 172 173 174 175 176 177 178 179
static ssize_t class_id_show(struct device *dev,
			       struct device_attribute *dev_attr, char *buf)
{
	struct hv_device *hv_dev = device_to_hv_device(dev);

	if (!hv_dev->channel)
		return -ENODEV;
	return sprintf(buf, "{%pUl}\n",
		       hv_dev->channel->offermsg.offer.if_type.b);
}
static DEVICE_ATTR_RO(class_id);

180 181 182 183 184 185 186 187 188 189 190 191
static ssize_t device_id_show(struct device *dev,
			      struct device_attribute *dev_attr, char *buf)
{
	struct hv_device *hv_dev = device_to_hv_device(dev);

	if (!hv_dev->channel)
		return -ENODEV;
	return sprintf(buf, "{%pUl}\n",
		       hv_dev->channel->offermsg.offer.if_instance.b);
}
static DEVICE_ATTR_RO(device_id);

192 193 194 195 196 197 198 199 200 201 202
static ssize_t modalias_show(struct device *dev,
			     struct device_attribute *dev_attr, char *buf)
{
	struct hv_device *hv_dev = device_to_hv_device(dev);
	char alias_name[VMBUS_ALIAS_LEN + 1];

	print_alias_name(hv_dev, alias_name);
	return sprintf(buf, "vmbus:%s\n", alias_name);
}
static DEVICE_ATTR_RO(modalias);

203 204 205 206 207 208 209 210 211 212 213 214 215 216
#ifdef CONFIG_NUMA
static ssize_t numa_node_show(struct device *dev,
			      struct device_attribute *attr, char *buf)
{
	struct hv_device *hv_dev = device_to_hv_device(dev);

	if (!hv_dev->channel)
		return -ENODEV;

	return sprintf(buf, "%d\n", hv_dev->channel->numa_node);
}
static DEVICE_ATTR_RO(numa_node);
#endif

217 218 219 220 221 222 223 224 225 226
static ssize_t server_monitor_pending_show(struct device *dev,
					   struct device_attribute *dev_attr,
					   char *buf)
{
	struct hv_device *hv_dev = device_to_hv_device(dev);

	if (!hv_dev->channel)
		return -ENODEV;
	return sprintf(buf, "%d\n",
		       channel_pending(hv_dev->channel,
227
				       vmbus_connection.monitor_pages[0]));
228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243
}
static DEVICE_ATTR_RO(server_monitor_pending);

static ssize_t client_monitor_pending_show(struct device *dev,
					   struct device_attribute *dev_attr,
					   char *buf)
{
	struct hv_device *hv_dev = device_to_hv_device(dev);

	if (!hv_dev->channel)
		return -ENODEV;
	return sprintf(buf, "%d\n",
		       channel_pending(hv_dev->channel,
				       vmbus_connection.monitor_pages[1]));
}
static DEVICE_ATTR_RO(client_monitor_pending);
244

245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272
static ssize_t server_monitor_latency_show(struct device *dev,
					   struct device_attribute *dev_attr,
					   char *buf)
{
	struct hv_device *hv_dev = device_to_hv_device(dev);

	if (!hv_dev->channel)
		return -ENODEV;
	return sprintf(buf, "%d\n",
		       channel_latency(hv_dev->channel,
				       vmbus_connection.monitor_pages[0]));
}
static DEVICE_ATTR_RO(server_monitor_latency);

static ssize_t client_monitor_latency_show(struct device *dev,
					   struct device_attribute *dev_attr,
					   char *buf)
{
	struct hv_device *hv_dev = device_to_hv_device(dev);

	if (!hv_dev->channel)
		return -ENODEV;
	return sprintf(buf, "%d\n",
		       channel_latency(hv_dev->channel,
				       vmbus_connection.monitor_pages[1]));
}
static DEVICE_ATTR_RO(client_monitor_latency);

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
static ssize_t server_monitor_conn_id_show(struct device *dev,
					   struct device_attribute *dev_attr,
					   char *buf)
{
	struct hv_device *hv_dev = device_to_hv_device(dev);

	if (!hv_dev->channel)
		return -ENODEV;
	return sprintf(buf, "%d\n",
		       channel_conn_id(hv_dev->channel,
				       vmbus_connection.monitor_pages[0]));
}
static DEVICE_ATTR_RO(server_monitor_conn_id);

static ssize_t client_monitor_conn_id_show(struct device *dev,
					   struct device_attribute *dev_attr,
					   char *buf)
{
	struct hv_device *hv_dev = device_to_hv_device(dev);

	if (!hv_dev->channel)
		return -ENODEV;
	return sprintf(buf, "%d\n",
		       channel_conn_id(hv_dev->channel,
				       vmbus_connection.monitor_pages[1]));
}
static DEVICE_ATTR_RO(client_monitor_conn_id);

301 302 303 304 305
static ssize_t out_intr_mask_show(struct device *dev,
				  struct device_attribute *dev_attr, char *buf)
{
	struct hv_device *hv_dev = device_to_hv_device(dev);
	struct hv_ring_buffer_debug_info outbound;
306
	int ret;
307 308 309

	if (!hv_dev->channel)
		return -ENODEV;
310 311 312 313 314 315

	ret = hv_ringbuffer_get_debuginfo(&hv_dev->channel->outbound,
					  &outbound);
	if (ret < 0)
		return ret;

316 317 318 319 320 321 322 323 324
	return sprintf(buf, "%d\n", outbound.current_interrupt_mask);
}
static DEVICE_ATTR_RO(out_intr_mask);

static ssize_t out_read_index_show(struct device *dev,
				   struct device_attribute *dev_attr, char *buf)
{
	struct hv_device *hv_dev = device_to_hv_device(dev);
	struct hv_ring_buffer_debug_info outbound;
325
	int ret;
326 327 328

	if (!hv_dev->channel)
		return -ENODEV;
329 330 331 332 333

	ret = hv_ringbuffer_get_debuginfo(&hv_dev->channel->outbound,
					  &outbound);
	if (ret < 0)
		return ret;
334 335 336 337 338 339 340 341 342 343
	return sprintf(buf, "%d\n", outbound.current_read_index);
}
static DEVICE_ATTR_RO(out_read_index);

static ssize_t out_write_index_show(struct device *dev,
				    struct device_attribute *dev_attr,
				    char *buf)
{
	struct hv_device *hv_dev = device_to_hv_device(dev);
	struct hv_ring_buffer_debug_info outbound;
344
	int ret;
345 346 347

	if (!hv_dev->channel)
		return -ENODEV;
348 349 350 351 352

	ret = hv_ringbuffer_get_debuginfo(&hv_dev->channel->outbound,
					  &outbound);
	if (ret < 0)
		return ret;
353 354 355 356 357 358 359 360 361 362
	return sprintf(buf, "%d\n", outbound.current_write_index);
}
static DEVICE_ATTR_RO(out_write_index);

static ssize_t out_read_bytes_avail_show(struct device *dev,
					 struct device_attribute *dev_attr,
					 char *buf)
{
	struct hv_device *hv_dev = device_to_hv_device(dev);
	struct hv_ring_buffer_debug_info outbound;
363
	int ret;
364 365 366

	if (!hv_dev->channel)
		return -ENODEV;
367 368 369 370 371

	ret = hv_ringbuffer_get_debuginfo(&hv_dev->channel->outbound,
					  &outbound);
	if (ret < 0)
		return ret;
372 373 374 375 376 377 378 379 380 381
	return sprintf(buf, "%d\n", outbound.bytes_avail_toread);
}
static DEVICE_ATTR_RO(out_read_bytes_avail);

static ssize_t out_write_bytes_avail_show(struct device *dev,
					  struct device_attribute *dev_attr,
					  char *buf)
{
	struct hv_device *hv_dev = device_to_hv_device(dev);
	struct hv_ring_buffer_debug_info outbound;
382
	int ret;
383 384 385

	if (!hv_dev->channel)
		return -ENODEV;
386 387 388 389 390

	ret = hv_ringbuffer_get_debuginfo(&hv_dev->channel->outbound,
					  &outbound);
	if (ret < 0)
		return ret;
391 392 393 394 395 396 397 398 399
	return sprintf(buf, "%d\n", outbound.bytes_avail_towrite);
}
static DEVICE_ATTR_RO(out_write_bytes_avail);

static ssize_t in_intr_mask_show(struct device *dev,
				 struct device_attribute *dev_attr, char *buf)
{
	struct hv_device *hv_dev = device_to_hv_device(dev);
	struct hv_ring_buffer_debug_info inbound;
400
	int ret;
401 402 403

	if (!hv_dev->channel)
		return -ENODEV;
404 405 406 407 408

	ret = hv_ringbuffer_get_debuginfo(&hv_dev->channel->inbound, &inbound);
	if (ret < 0)
		return ret;

409 410 411 412 413 414 415 416 417
	return sprintf(buf, "%d\n", inbound.current_interrupt_mask);
}
static DEVICE_ATTR_RO(in_intr_mask);

static ssize_t in_read_index_show(struct device *dev,
				  struct device_attribute *dev_attr, char *buf)
{
	struct hv_device *hv_dev = device_to_hv_device(dev);
	struct hv_ring_buffer_debug_info inbound;
418
	int ret;
419 420 421

	if (!hv_dev->channel)
		return -ENODEV;
422 423 424 425 426

	ret = hv_ringbuffer_get_debuginfo(&hv_dev->channel->inbound, &inbound);
	if (ret < 0)
		return ret;

427 428 429 430 431 432 433 434 435
	return sprintf(buf, "%d\n", inbound.current_read_index);
}
static DEVICE_ATTR_RO(in_read_index);

static ssize_t in_write_index_show(struct device *dev,
				   struct device_attribute *dev_attr, char *buf)
{
	struct hv_device *hv_dev = device_to_hv_device(dev);
	struct hv_ring_buffer_debug_info inbound;
436
	int ret;
437 438 439

	if (!hv_dev->channel)
		return -ENODEV;
440 441 442 443 444

	ret = hv_ringbuffer_get_debuginfo(&hv_dev->channel->inbound, &inbound);
	if (ret < 0)
		return ret;

445 446 447 448 449 450 451 452 453 454
	return sprintf(buf, "%d\n", inbound.current_write_index);
}
static DEVICE_ATTR_RO(in_write_index);

static ssize_t in_read_bytes_avail_show(struct device *dev,
					struct device_attribute *dev_attr,
					char *buf)
{
	struct hv_device *hv_dev = device_to_hv_device(dev);
	struct hv_ring_buffer_debug_info inbound;
455
	int ret;
456 457 458

	if (!hv_dev->channel)
		return -ENODEV;
459 460 461 462 463

	ret = hv_ringbuffer_get_debuginfo(&hv_dev->channel->inbound, &inbound);
	if (ret < 0)
		return ret;

464 465 466 467 468 469 470 471 472 473
	return sprintf(buf, "%d\n", inbound.bytes_avail_toread);
}
static DEVICE_ATTR_RO(in_read_bytes_avail);

static ssize_t in_write_bytes_avail_show(struct device *dev,
					 struct device_attribute *dev_attr,
					 char *buf)
{
	struct hv_device *hv_dev = device_to_hv_device(dev);
	struct hv_ring_buffer_debug_info inbound;
474
	int ret;
475 476 477

	if (!hv_dev->channel)
		return -ENODEV;
478 479 480 481 482

	ret = hv_ringbuffer_get_debuginfo(&hv_dev->channel->inbound, &inbound);
	if (ret < 0)
		return ret;

483 484 485 486
	return sprintf(buf, "%d\n", inbound.bytes_avail_towrite);
}
static DEVICE_ATTR_RO(in_write_bytes_avail);

487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523
static ssize_t channel_vp_mapping_show(struct device *dev,
				       struct device_attribute *dev_attr,
				       char *buf)
{
	struct hv_device *hv_dev = device_to_hv_device(dev);
	struct vmbus_channel *channel = hv_dev->channel, *cur_sc;
	unsigned long flags;
	int buf_size = PAGE_SIZE, n_written, tot_written;
	struct list_head *cur;

	if (!channel)
		return -ENODEV;

	tot_written = snprintf(buf, buf_size, "%u:%u\n",
		channel->offermsg.child_relid, channel->target_cpu);

	spin_lock_irqsave(&channel->lock, flags);

	list_for_each(cur, &channel->sc_list) {
		if (tot_written >= buf_size - 1)
			break;

		cur_sc = list_entry(cur, struct vmbus_channel, sc_list);
		n_written = scnprintf(buf + tot_written,
				     buf_size - tot_written,
				     "%u:%u\n",
				     cur_sc->offermsg.child_relid,
				     cur_sc->target_cpu);
		tot_written += n_written;
	}

	spin_unlock_irqrestore(&channel->lock, flags);

	return tot_written;
}
static DEVICE_ATTR_RO(channel_vp_mapping);

524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541
static ssize_t vendor_show(struct device *dev,
			   struct device_attribute *dev_attr,
			   char *buf)
{
	struct hv_device *hv_dev = device_to_hv_device(dev);
	return sprintf(buf, "0x%x\n", hv_dev->vendor_id);
}
static DEVICE_ATTR_RO(vendor);

static ssize_t device_show(struct device *dev,
			   struct device_attribute *dev_attr,
			   char *buf)
{
	struct hv_device *hv_dev = device_to_hv_device(dev);
	return sprintf(buf, "0x%x\n", hv_dev->device_id);
}
static DEVICE_ATTR_RO(device);

542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589
static ssize_t driver_override_store(struct device *dev,
				     struct device_attribute *attr,
				     const char *buf, size_t count)
{
	struct hv_device *hv_dev = device_to_hv_device(dev);
	char *driver_override, *old, *cp;

	/* We need to keep extra room for a newline */
	if (count >= (PAGE_SIZE - 1))
		return -EINVAL;

	driver_override = kstrndup(buf, count, GFP_KERNEL);
	if (!driver_override)
		return -ENOMEM;

	cp = strchr(driver_override, '\n');
	if (cp)
		*cp = '\0';

	device_lock(dev);
	old = hv_dev->driver_override;
	if (strlen(driver_override)) {
		hv_dev->driver_override = driver_override;
	} else {
		kfree(driver_override);
		hv_dev->driver_override = NULL;
	}
	device_unlock(dev);

	kfree(old);

	return count;
}

static ssize_t driver_override_show(struct device *dev,
				    struct device_attribute *attr, char *buf)
{
	struct hv_device *hv_dev = device_to_hv_device(dev);
	ssize_t len;

	device_lock(dev);
	len = snprintf(buf, PAGE_SIZE, "%s\n", hv_dev->driver_override);
	device_unlock(dev);

	return len;
}
static DEVICE_ATTR_RW(driver_override);

590
/* Set up per device attributes in /sys/bus/vmbus/devices/<bus device> */
591
static struct attribute *vmbus_dev_attrs[] = {
592
	&dev_attr_id.attr,
593
	&dev_attr_state.attr,
594
	&dev_attr_monitor_id.attr,
595
	&dev_attr_class_id.attr,
596
	&dev_attr_device_id.attr,
597
	&dev_attr_modalias.attr,
598 599 600
#ifdef CONFIG_NUMA
	&dev_attr_numa_node.attr,
#endif
601 602
	&dev_attr_server_monitor_pending.attr,
	&dev_attr_client_monitor_pending.attr,
603 604
	&dev_attr_server_monitor_latency.attr,
	&dev_attr_client_monitor_latency.attr,
605 606
	&dev_attr_server_monitor_conn_id.attr,
	&dev_attr_client_monitor_conn_id.attr,
607 608 609 610 611 612 613 614 615 616
	&dev_attr_out_intr_mask.attr,
	&dev_attr_out_read_index.attr,
	&dev_attr_out_write_index.attr,
	&dev_attr_out_read_bytes_avail.attr,
	&dev_attr_out_write_bytes_avail.attr,
	&dev_attr_in_intr_mask.attr,
	&dev_attr_in_read_index.attr,
	&dev_attr_in_write_index.attr,
	&dev_attr_in_read_bytes_avail.attr,
	&dev_attr_in_write_bytes_avail.attr,
617
	&dev_attr_channel_vp_mapping.attr,
618 619
	&dev_attr_vendor.attr,
	&dev_attr_device.attr,
620
	&dev_attr_driver_override.attr,
621 622
	NULL,
};
623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652

/*
 * Device-level attribute_group callback function. Returns the permission for
 * each attribute, and returns 0 if an attribute is not visible.
 */
static umode_t vmbus_dev_attr_is_visible(struct kobject *kobj,
					 struct attribute *attr, int idx)
{
	struct device *dev = kobj_to_dev(kobj);
	const struct hv_device *hv_dev = device_to_hv_device(dev);

	/* Hide the monitor attributes if the monitor mechanism is not used. */
	if (!hv_dev->channel->offermsg.monitor_allocated &&
	    (attr == &dev_attr_monitor_id.attr ||
	     attr == &dev_attr_server_monitor_pending.attr ||
	     attr == &dev_attr_client_monitor_pending.attr ||
	     attr == &dev_attr_server_monitor_latency.attr ||
	     attr == &dev_attr_client_monitor_latency.attr ||
	     attr == &dev_attr_server_monitor_conn_id.attr ||
	     attr == &dev_attr_client_monitor_conn_id.attr))
		return 0;

	return attr->mode;
}

static const struct attribute_group vmbus_dev_group = {
	.attrs = vmbus_dev_attrs,
	.is_visible = vmbus_dev_attr_is_visible
};
__ATTRIBUTE_GROUPS(vmbus_dev);
653

654 655 656 657 658 659
/*
 * vmbus_uevent - add uevent for our device
 *
 * This routine is invoked when a device is added or removed on the vmbus to
 * generate a uevent to udev in the userspace. The udev will then look at its
 * rule and the uevent generated here to load the appropriate driver
660 661 662 663
 *
 * The alias string will be of the form vmbus:guid where guid is the string
 * representation of the device guid (each byte of the guid will be
 * represented with two hex characters.
664 665 666 667
 */
static int vmbus_uevent(struct device *device, struct kobj_uevent_env *env)
{
	struct hv_device *dev = device_to_hv_device(device);
668 669
	int ret;
	char alias_name[VMBUS_ALIAS_LEN + 1];
670

671
	print_alias_name(dev, alias_name);
672 673
	ret = add_uevent_var(env, "MODALIAS=vmbus:%s", alias_name);
	return ret;
674 675
}

676
static const struct hv_vmbus_device_id *
677
hv_vmbus_dev_match(const struct hv_vmbus_device_id *id, const guid_t *guid)
678 679 680 681
{
	if (id == NULL)
		return NULL; /* empty device table */

682 683
	for (; !guid_is_null(&id->guid); id++)
		if (guid_equal(&id->guid, guid))
684 685 686 687 688 689
			return id;

	return NULL;
}

static const struct hv_vmbus_device_id *
690
hv_vmbus_dynid_match(struct hv_driver *drv, const guid_t *guid)
691
{
692 693 694 695 696
	const struct hv_vmbus_device_id *id = NULL;
	struct vmbus_dynid *dynid;

	spin_lock(&drv->dynids.lock);
	list_for_each_entry(dynid, &drv->dynids.list, node) {
697
		if (guid_equal(&dynid->id.guid, guid)) {
698 699 700 701 702 703
			id = &dynid->id;
			break;
		}
	}
	spin_unlock(&drv->dynids.lock);

704 705
	return id;
}
706

707
static const struct hv_vmbus_device_id vmbus_device_null;
708

709 710 711 712 713 714 715
/*
 * Return a matching hv_vmbus_device_id pointer.
 * If there is no match, return NULL.
 */
static const struct hv_vmbus_device_id *hv_vmbus_get_id(struct hv_driver *drv,
							struct hv_device *dev)
{
716
	const guid_t *guid = &dev->dev_type;
717
	const struct hv_vmbus_device_id *id;
718

719 720 721 722 723 724 725 726 727 728 729 730 731 732
	/* When driver_override is set, only bind to the matching driver */
	if (dev->driver_override && strcmp(dev->driver_override, drv->name))
		return NULL;

	/* Look at the dynamic ids first, before the static ones */
	id = hv_vmbus_dynid_match(drv, guid);
	if (!id)
		id = hv_vmbus_dev_match(drv->id_table, guid);

	/* driver_override will always match, send a dummy id */
	if (!id && dev->driver_override)
		id = &vmbus_device_null;

	return id;
733 734
}

735
/* vmbus_add_dynid - add a new device ID to this driver and re-probe devices */
736
static int vmbus_add_dynid(struct hv_driver *drv, guid_t *guid)
737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773
{
	struct vmbus_dynid *dynid;

	dynid = kzalloc(sizeof(*dynid), GFP_KERNEL);
	if (!dynid)
		return -ENOMEM;

	dynid->id.guid = *guid;

	spin_lock(&drv->dynids.lock);
	list_add_tail(&dynid->node, &drv->dynids.list);
	spin_unlock(&drv->dynids.lock);

	return driver_attach(&drv->driver);
}

static void vmbus_free_dynids(struct hv_driver *drv)
{
	struct vmbus_dynid *dynid, *n;

	spin_lock(&drv->dynids.lock);
	list_for_each_entry_safe(dynid, n, &drv->dynids.list, node) {
		list_del(&dynid->node);
		kfree(dynid);
	}
	spin_unlock(&drv->dynids.lock);
}

/*
 * store_new_id - sysfs frontend to vmbus_add_dynid()
 *
 * Allow GUIDs to be added to an existing driver via sysfs.
 */
static ssize_t new_id_store(struct device_driver *driver, const char *buf,
			    size_t count)
{
	struct hv_driver *drv = drv_to_hv_drv(driver);
774
	guid_t guid;
775 776
	ssize_t retval;

777
	retval = guid_parse(buf, &guid);
778 779
	if (retval)
		return retval;
780

781
	if (hv_vmbus_dynid_match(drv, &guid))
782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800
		return -EEXIST;

	retval = vmbus_add_dynid(drv, &guid);
	if (retval)
		return retval;
	return count;
}
static DRIVER_ATTR_WO(new_id);

/*
 * store_remove_id - remove a PCI device ID from this driver
 *
 * Removes a dynamic pci device ID to this driver.
 */
static ssize_t remove_id_store(struct device_driver *driver, const char *buf,
			       size_t count)
{
	struct hv_driver *drv = drv_to_hv_drv(driver);
	struct vmbus_dynid *dynid, *n;
801
	guid_t guid;
802
	ssize_t retval;
803

804
	retval = guid_parse(buf, &guid);
805 806
	if (retval)
		return retval;
807

808
	retval = -ENODEV;
809 810 811 812
	spin_lock(&drv->dynids.lock);
	list_for_each_entry_safe(dynid, n, &drv->dynids.list, node) {
		struct hv_vmbus_device_id *id = &dynid->id;

813
		if (guid_equal(&id->guid, &guid)) {
814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831
			list_del(&dynid->node);
			kfree(dynid);
			retval = count;
			break;
		}
	}
	spin_unlock(&drv->dynids.lock);

	return retval;
}
static DRIVER_ATTR_WO(remove_id);

static struct attribute *vmbus_drv_attrs[] = {
	&driver_attr_new_id.attr,
	&driver_attr_remove_id.attr,
	NULL,
};
ATTRIBUTE_GROUPS(vmbus_drv);
832

833 834 835 836 837 838 839

/*
 * vmbus_match - Attempt to match the specified device to the specified driver
 */
static int vmbus_match(struct device *device, struct device_driver *driver)
{
	struct hv_driver *drv = drv_to_hv_drv(driver);
840
	struct hv_device *hv_dev = device_to_hv_device(device);
841

842 843 844 845
	/* The hv_sock driver handles all hv_sock offers. */
	if (is_hvsock_channel(hv_dev->channel))
		return drv->hvsock;

846
	if (hv_vmbus_get_id(drv, hv_dev))
847
		return 1;
848

849
	return 0;
850 851
}

852 853 854 855 856 857 858 859
/*
 * vmbus_probe - Add the new vmbus's child device
 */
static int vmbus_probe(struct device *child_device)
{
	int ret = 0;
	struct hv_driver *drv =
			drv_to_hv_drv(child_device->driver);
860
	struct hv_device *dev = device_to_hv_device(child_device);
861
	const struct hv_vmbus_device_id *dev_id;
862

863
	dev_id = hv_vmbus_get_id(drv, dev);
864
	if (drv->probe) {
865
		ret = drv->probe(dev, dev_id);
866
		if (ret != 0)
867 868
			pr_err("probe failed for device %s (%d)\n",
			       dev_name(child_device), ret);
869 870

	} else {
871 872
		pr_err("probe not set for driver %s\n",
		       dev_name(child_device));
873
		ret = -ENODEV;
874 875 876 877
	}
	return ret;
}

878 879 880 881 882
/*
 * vmbus_remove - Remove a vmbus device
 */
static int vmbus_remove(struct device *child_device)
{
883
	struct hv_driver *drv;
884
	struct hv_device *dev = device_to_hv_device(child_device);
885

886 887 888 889 890
	if (child_device->driver) {
		drv = drv_to_hv_drv(child_device->driver);
		if (drv->remove)
			drv->remove(dev);
	}
891 892 893 894

	return 0;
}

895 896 897 898 899 900 901

/*
 * vmbus_shutdown - Shutdown a vmbus device
 */
static void vmbus_shutdown(struct device *child_device)
{
	struct hv_driver *drv;
902
	struct hv_device *dev = device_to_hv_device(child_device);
903 904 905 906 907 908 909 910


	/* The device may not be attached yet */
	if (!child_device->driver)
		return;

	drv = drv_to_hv_drv(child_device->driver);

911 912
	if (drv->shutdown)
		drv->shutdown(dev);
913 914
}

915
#ifdef CONFIG_PM_SLEEP
916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952
/*
 * vmbus_suspend - Suspend a vmbus device
 */
static int vmbus_suspend(struct device *child_device)
{
	struct hv_driver *drv;
	struct hv_device *dev = device_to_hv_device(child_device);

	/* The device may not be attached yet */
	if (!child_device->driver)
		return 0;

	drv = drv_to_hv_drv(child_device->driver);
	if (!drv->suspend)
		return -EOPNOTSUPP;

	return drv->suspend(dev);
}

/*
 * vmbus_resume - Resume a vmbus device
 */
static int vmbus_resume(struct device *child_device)
{
	struct hv_driver *drv;
	struct hv_device *dev = device_to_hv_device(child_device);

	/* The device may not be attached yet */
	if (!child_device->driver)
		return 0;

	drv = drv_to_hv_drv(child_device->driver);
	if (!drv->resume)
		return -EOPNOTSUPP;

	return drv->resume(dev);
}
953
#endif /* CONFIG_PM_SLEEP */
954 955 956 957 958 959

/*
 * vmbus_device_release - Final callback release of the vmbus child device
 */
static void vmbus_device_release(struct device *device)
{
960
	struct hv_device *hv_dev = device_to_hv_device(device);
961
	struct vmbus_channel *channel = hv_dev->channel;
962

963
	mutex_lock(&vmbus_connection.channel_mutex);
964
	hv_process_channel_removal(channel);
965
	mutex_unlock(&vmbus_connection.channel_mutex);
966
	kfree(hv_dev);
967 968
}

969 970 971 972 973 974 975 976
/*
 * Note: we must use SET_NOIRQ_SYSTEM_SLEEP_PM_OPS rather than
 * SET_SYSTEM_SLEEP_PM_OPS: see the comment before vmbus_bus_pm.
 */
static const struct dev_pm_ops vmbus_pm = {
	SET_NOIRQ_SYSTEM_SLEEP_PM_OPS(vmbus_suspend, vmbus_resume)
};

977
/* The one and only one */
978 979 980 981 982 983 984
static struct bus_type  hv_bus = {
	.name =		"vmbus",
	.match =		vmbus_match,
	.shutdown =		vmbus_shutdown,
	.remove =		vmbus_remove,
	.probe =		vmbus_probe,
	.uevent =		vmbus_uevent,
985 986
	.dev_groups =		vmbus_dev_groups,
	.drv_groups =		vmbus_drv_groups,
987
	.pm =			&vmbus_pm,
988 989
};

990 991 992 993 994 995 996 997 998
struct onmessage_work_context {
	struct work_struct work;
	struct hv_message msg;
};

static void vmbus_onmessage_work(struct work_struct *work)
{
	struct onmessage_work_context *ctx;

999 1000 1001 1002
	/* Do not process messages if we're in DISCONNECTED state */
	if (vmbus_connection.conn_state == DISCONNECTED)
		return;

1003 1004 1005 1006 1007 1008
	ctx = container_of(work, struct onmessage_work_context,
			   work);
	vmbus_onmessage(&ctx->msg);
	kfree(ctx);
}

1009
void vmbus_on_msg_dpc(unsigned long data)
G
Greg Kroah-Hartman 已提交
1010
{
1011 1012
	struct hv_per_cpu_context *hv_cpu = (void *)data;
	void *page_addr = hv_cpu->synic_message_page;
G
Greg Kroah-Hartman 已提交
1013 1014
	struct hv_message *msg = (struct hv_message *)page_addr +
				  VMBUS_MESSAGE_SINT;
1015
	struct vmbus_channel_message_header *hdr;
1016
	const struct vmbus_channel_message_table_entry *entry;
1017
	struct onmessage_work_context *ctx;
1018
	u32 message_type = msg->header.message_type;
G
Greg Kroah-Hartman 已提交
1019

1020
	if (message_type == HVMSG_NONE)
1021 1022
		/* no msg */
		return;
1023

1024
	hdr = (struct vmbus_channel_message_header *)msg->u.payload;
1025

1026 1027
	trace_vmbus_on_msg_dpc(hdr);

1028 1029 1030 1031
	if (hdr->msgtype >= CHANNELMSG_COUNT) {
		WARN_ONCE(1, "unknown msgtype=%d\n", hdr->msgtype);
		goto msg_handled;
	}
1032

1033 1034 1035 1036 1037
	entry = &channel_message_table[hdr->msgtype];
	if (entry->handler_type	== VMHT_BLOCKING) {
		ctx = kmalloc(sizeof(*ctx), GFP_ATOMIC);
		if (ctx == NULL)
			return;
1038

1039 1040
		INIT_WORK(&ctx->work, vmbus_onmessage_work);
		memcpy(&ctx->msg, msg, sizeof(*msg));
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
		/*
		 * The host can generate a rescind message while we
		 * may still be handling the original offer. We deal with
		 * this condition by ensuring the processing is done on the
		 * same CPU.
		 */
		switch (hdr->msgtype) {
		case CHANNELMSG_RESCIND_CHANNELOFFER:
			/*
			 * If we are handling the rescind message;
			 * schedule the work on the global work queue.
			 */
			schedule_work_on(vmbus_connection.connect_cpu,
					 &ctx->work);
			break;

		case CHANNELMSG_OFFERCHANNEL:
			atomic_inc(&vmbus_connection.offer_in_progress);
			queue_work_on(vmbus_connection.connect_cpu,
				      vmbus_connection.work_queue,
				      &ctx->work);
			break;

		default:
			queue_work(vmbus_connection.work_queue, &ctx->work);
		}
1068 1069
	} else
		entry->message_handler(hdr);
G
Greg Kroah-Hartman 已提交
1070

1071
msg_handled:
1072
	vmbus_signal_eom(msg, message_type);
G
Greg Kroah-Hartman 已提交
1073 1074
}

1075
#ifdef CONFIG_PM_SLEEP
1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110
/*
 * Fake RESCIND_CHANNEL messages to clean up hv_sock channels by force for
 * hibernation, because hv_sock connections can not persist across hibernation.
 */
static void vmbus_force_channel_rescinded(struct vmbus_channel *channel)
{
	struct onmessage_work_context *ctx;
	struct vmbus_channel_rescind_offer *rescind;

	WARN_ON(!is_hvsock_channel(channel));

	/*
	 * sizeof(*ctx) is small and the allocation should really not fail,
	 * otherwise the state of the hv_sock connections ends up in limbo.
	 */
	ctx = kzalloc(sizeof(*ctx), GFP_KERNEL | __GFP_NOFAIL);

	/*
	 * So far, these are not really used by Linux. Just set them to the
	 * reasonable values conforming to the definitions of the fields.
	 */
	ctx->msg.header.message_type = 1;
	ctx->msg.header.payload_size = sizeof(*rescind);

	/* These values are actually used by Linux. */
	rescind = (struct vmbus_channel_rescind_offer *)ctx->msg.u.payload;
	rescind->header.msgtype = CHANNELMSG_RESCIND_CHANNELOFFER;
	rescind->child_relid = channel->offermsg.child_relid;

	INIT_WORK(&ctx->work, vmbus_onmessage_work);

	queue_work_on(vmbus_connection.connect_cpu,
		      vmbus_connection.work_queue,
		      &ctx->work);
}
1111
#endif /* CONFIG_PM_SLEEP */
1112

1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124
/*
 * Direct callback for channels using other deferred processing
 */
static void vmbus_channel_isr(struct vmbus_channel *channel)
{
	void (*callback_fn)(void *);

	callback_fn = READ_ONCE(channel->onchannel_callback);
	if (likely(callback_fn != NULL))
		(*callback_fn)(channel->channel_callback_context);
}

1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163
/*
 * Schedule all channels with events pending
 */
static void vmbus_chan_sched(struct hv_per_cpu_context *hv_cpu)
{
	unsigned long *recv_int_page;
	u32 maxbits, relid;

	if (vmbus_proto_version < VERSION_WIN8) {
		maxbits = MAX_NUM_CHANNELS_SUPPORTED;
		recv_int_page = vmbus_connection.recv_int_page;
	} else {
		/*
		 * When the host is win8 and beyond, the event page
		 * can be directly checked to get the id of the channel
		 * that has the interrupt pending.
		 */
		void *page_addr = hv_cpu->synic_event_page;
		union hv_synic_event_flags *event
			= (union hv_synic_event_flags *)page_addr +
						 VMBUS_MESSAGE_SINT;

		maxbits = HV_EVENT_FLAGS_COUNT;
		recv_int_page = event->flags;
	}

	if (unlikely(!recv_int_page))
		return;

	for_each_set_bit(relid, recv_int_page, maxbits) {
		struct vmbus_channel *channel;

		if (!sync_test_and_clear_bit(relid, recv_int_page))
			continue;

		/* Special case - vmbus channel protocol msg */
		if (relid == 0)
			continue;

1164 1165
		rcu_read_lock();

1166
		/* Find channel based on relid */
1167
		list_for_each_entry_rcu(channel, &hv_cpu->chan_list, percpu_list) {
1168 1169 1170
			if (channel->offermsg.child_relid != relid)
				continue;

1171 1172 1173
			if (channel->rescind)
				continue;

V
Vitaly Kuznetsov 已提交
1174 1175
			trace_vmbus_chan_sched(channel);

1176 1177
			++channel->interrupts;

1178 1179 1180
			switch (channel->callback_mode) {
			case HV_CALL_ISR:
				vmbus_channel_isr(channel);
1181
				break;
1182 1183 1184 1185 1186 1187

			case HV_CALL_BATCHED:
				hv_begin_read(&channel->inbound);
				/* fallthrough */
			case HV_CALL_DIRECT:
				tasklet_schedule(&channel->callback_event);
1188 1189
			}
		}
1190 1191

		rcu_read_unlock();
1192 1193 1194
	}
}

1195
static void vmbus_isr(void)
G
Greg Kroah-Hartman 已提交
1196
{
1197 1198 1199
	struct hv_per_cpu_context *hv_cpu
		= this_cpu_ptr(hv_context.cpu_context);
	void *page_addr = hv_cpu->synic_event_page;
G
Greg Kroah-Hartman 已提交
1200 1201
	struct hv_message *msg;
	union hv_synic_event_flags *event;
1202
	bool handled = false;
G
Greg Kroah-Hartman 已提交
1203

1204
	if (unlikely(page_addr == NULL))
1205
		return;
1206 1207 1208

	event = (union hv_synic_event_flags *)page_addr +
					 VMBUS_MESSAGE_SINT;
1209 1210 1211 1212 1213
	/*
	 * Check for events before checking for messages. This is the order
	 * in which events and messages are checked in Windows guests on
	 * Hyper-V, and the Windows team suggested we do the same.
	 */
G
Greg Kroah-Hartman 已提交
1214

1215 1216
	if ((vmbus_proto_version == VERSION_WS2008) ||
		(vmbus_proto_version == VERSION_WIN7)) {
G
Greg Kroah-Hartman 已提交
1217

1218
		/* Since we are a child, we only need to check bit 0 */
1219
		if (sync_test_and_clear_bit(0, event->flags))
1220 1221 1222 1223 1224 1225 1226 1227
			handled = true;
	} else {
		/*
		 * Our host is win8 or above. The signaling mechanism
		 * has changed and we can directly look at the event page.
		 * If bit n is set then we have an interrup on the channel
		 * whose id is n.
		 */
1228 1229
		handled = true;
	}
1230

1231
	if (handled)
1232
		vmbus_chan_sched(hv_cpu);
1233

1234
	page_addr = hv_cpu->synic_message_page;
1235 1236 1237
	msg = (struct hv_message *)page_addr + VMBUS_MESSAGE_SINT;

	/* Check if there are actual msgs to be processed */
1238
	if (msg->header.message_type != HVMSG_NONE) {
1239 1240 1241 1242
		if (msg->header.message_type == HVMSG_TIMER_EXPIRED) {
			hv_stimer0_isr();
			vmbus_signal_eom(msg, HVMSG_TIMER_EXPIRED);
		} else
1243
			tasklet_schedule(&hv_cpu->msg_dpc);
1244
	}
1245 1246

	add_interrupt_randomness(HYPERVISOR_CALLBACK_VECTOR, 0);
1247 1248
}

1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275
/*
 * Boolean to control whether to report panic messages over Hyper-V.
 *
 * It can be set via /proc/sys/kernel/hyperv/record_panic_msg
 */
static int sysctl_record_panic_msg = 1;

/*
 * Callback from kmsg_dump. Grab as much as possible from the end of the kmsg
 * buffer and call into Hyper-V to transfer the data.
 */
static void hv_kmsg_dump(struct kmsg_dumper *dumper,
			 enum kmsg_dump_reason reason)
{
	size_t bytes_written;
	phys_addr_t panic_pa;

	/* We are only interested in panics. */
	if ((reason != KMSG_DUMP_PANIC) || (!sysctl_record_panic_msg))
		return;

	panic_pa = virt_to_phys(hv_panic_page);

	/*
	 * Write dump contents to the page. No need to synchronize; panic should
	 * be single-threaded.
	 */
1276 1277 1278 1279
	kmsg_dump_get_buffer(dumper, true, hv_panic_page, PAGE_SIZE,
			     &bytes_written);
	if (bytes_written)
		hyperv_report_panic_msg(panic_pa, bytes_written);
1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298
}

static struct kmsg_dumper hv_kmsg_dumper = {
	.dump = hv_kmsg_dump,
};

static struct ctl_table_header *hv_ctl_table_hdr;

/*
 * sysctl option to allow the user to control whether kmsg data should be
 * reported to Hyper-V on panic.
 */
static struct ctl_table hv_ctl_table[] = {
	{
		.procname       = "hyperv_record_panic_msg",
		.data           = &sysctl_record_panic_msg,
		.maxlen         = sizeof(int),
		.mode           = 0644,
		.proc_handler   = proc_dointvec_minmax,
1299 1300
		.extra1		= SYSCTL_ZERO,
		.extra2		= SYSCTL_ONE
1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312
	},
	{}
};

static struct ctl_table hv_root_table[] = {
	{
		.procname	= "kernel",
		.mode		= 0555,
		.child		= hv_ctl_table
	},
	{}
};
1313

1314
/*
1315 1316 1317
 * vmbus_bus_init -Main vmbus driver initialization routine.
 *
 * Here, we
1318 1319 1320
 *	- initialize the vmbus driver context
 *	- invoke the vmbus hv main init routine
 *	- retrieve the channel offers
1321
 */
1322
static int vmbus_bus_init(void)
1323
{
1324
	int ret;
1325

1326 1327
	/* Hypervisor initialization...setup hypercall page..etc */
	ret = hv_init();
1328
	if (ret != 0) {
1329
		pr_err("Unable to initialize the hypervisor - 0x%x\n", ret);
1330
		return ret;
1331 1332
	}

1333
	ret = bus_register(&hv_bus);
1334
	if (ret)
1335
		return ret;
1336

1337
	hv_setup_vmbus_irq(vmbus_isr);
1338

1339 1340 1341
	ret = hv_synic_alloc();
	if (ret)
		goto err_alloc;
1342 1343 1344 1345 1346

	ret = hv_stimer_alloc(VMBUS_MESSAGE_SINT);
	if (ret < 0)
		goto err_alloc;

1347
	/*
1348 1349
	 * Initialize the per-cpu interrupt state and stimer state.
	 * Then connect to the host.
1350
	 */
1351
	ret = cpuhp_setup_state(CPUHP_AP_ONLINE_DYN, "hyperv/vmbus:online",
1352 1353
				hv_synic_init, hv_synic_cleanup);
	if (ret < 0)
1354
		goto err_cpuhp;
1355 1356
	hyperv_cpuhp_online = ret;

1357
	ret = vmbus_connect();
1358
	if (ret)
1359
		goto err_connect;
1360

1361 1362 1363
	/*
	 * Only register if the crash MSRs are available
	 */
1364
	if (ms_hyperv.misc_features & HV_FEATURE_GUEST_CRASH_MSR_AVAILABLE) {
1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377
		u64 hyperv_crash_ctl;
		/*
		 * Sysctl registration is not fatal, since by default
		 * reporting is enabled.
		 */
		hv_ctl_table_hdr = register_sysctl_table(hv_root_table);
		if (!hv_ctl_table_hdr)
			pr_err("Hyper-V: sysctl table register error");

		/*
		 * Register for panic kmsg callback only if the right
		 * capability is supported by the hypervisor.
		 */
1378
		hv_get_crash_ctl(hyperv_crash_ctl);
1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390
		if (hyperv_crash_ctl & HV_CRASH_CTL_CRASH_NOTIFY_MSG) {
			hv_panic_page = (void *)get_zeroed_page(GFP_KERNEL);
			if (hv_panic_page) {
				ret = kmsg_dump_register(&hv_kmsg_dumper);
				if (ret)
					pr_err("Hyper-V: kmsg dump register "
						"error 0x%x\n", ret);
			} else
				pr_err("Hyper-V: panic message page memory "
					"allocation failed");
		}

1391
		register_die_notifier(&hyperv_die_block);
1392 1393 1394 1395
		atomic_notifier_chain_register(&panic_notifier_list,
					       &hyperv_panic_block);
	}

1396
	vmbus_request_offers();
1397

1398
	return 0;
1399

1400
err_connect:
1401
	cpuhp_remove_state(hyperv_cpuhp_online);
1402 1403
err_cpuhp:
	hv_stimer_free();
1404 1405
err_alloc:
	hv_synic_free();
1406
	hv_remove_vmbus_irq();
1407 1408

	bus_unregister(&hv_bus);
1409
	free_page((unsigned long)hv_panic_page);
1410 1411
	unregister_sysctl_table(hv_ctl_table_hdr);
	hv_ctl_table_hdr = NULL;
1412
	return ret;
1413 1414
}

1415
/**
1416 1417
 * __vmbus_child_driver_register() - Register a vmbus's driver
 * @hv_driver: Pointer to driver structure you want to register
1418 1419
 * @owner: owner module of the drv
 * @mod_name: module name string
1420 1421
 *
 * Registers the given driver with Linux through the 'driver_register()' call
1422
 * and sets up the hyper-v vmbus handling for this driver.
1423 1424
 * It will return the state of the 'driver_register()' call.
 *
1425
 */
1426
int __vmbus_driver_register(struct hv_driver *hv_driver, struct module *owner, const char *mod_name)
1427
{
1428
	int ret;
1429

1430
	pr_info("registering driver %s\n", hv_driver->name);
1431

1432 1433 1434 1435
	ret = vmbus_exists();
	if (ret < 0)
		return ret;

1436 1437 1438 1439
	hv_driver->driver.name = hv_driver->name;
	hv_driver->driver.owner = owner;
	hv_driver->driver.mod_name = mod_name;
	hv_driver->driver.bus = &hv_bus;
1440

1441 1442 1443
	spin_lock_init(&hv_driver->dynids.lock);
	INIT_LIST_HEAD(&hv_driver->dynids.list);

1444
	ret = driver_register(&hv_driver->driver);
1445

1446
	return ret;
1447
}
1448
EXPORT_SYMBOL_GPL(__vmbus_driver_register);
1449

1450
/**
1451
 * vmbus_driver_unregister() - Unregister a vmbus's driver
1452 1453
 * @hv_driver: Pointer to driver structure you want to
 *             un-register
1454
 *
1455 1456
 * Un-register the given driver that was previous registered with a call to
 * vmbus_driver_register()
1457
 */
1458
void vmbus_driver_unregister(struct hv_driver *hv_driver)
1459
{
1460
	pr_info("unregistering driver %s\n", hv_driver->name);
1461

1462
	if (!vmbus_exists()) {
1463
		driver_unregister(&hv_driver->driver);
1464 1465
		vmbus_free_dynids(hv_driver);
	}
1466
}
1467
EXPORT_SYMBOL_GPL(vmbus_driver_unregister);
1468

1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482

/*
 * Called when last reference to channel is gone.
 */
static void vmbus_chan_release(struct kobject *kobj)
{
	struct vmbus_channel *channel
		= container_of(kobj, struct vmbus_channel, kobj);

	kfree_rcu(channel, rcu);
}

struct vmbus_chan_attribute {
	struct attribute attr;
1483
	ssize_t (*show)(struct vmbus_channel *chan, char *buf);
1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501
	ssize_t (*store)(struct vmbus_channel *chan,
			 const char *buf, size_t count);
};
#define VMBUS_CHAN_ATTR(_name, _mode, _show, _store) \
	struct vmbus_chan_attribute chan_attr_##_name \
		= __ATTR(_name, _mode, _show, _store)
#define VMBUS_CHAN_ATTR_RW(_name) \
	struct vmbus_chan_attribute chan_attr_##_name = __ATTR_RW(_name)
#define VMBUS_CHAN_ATTR_RO(_name) \
	struct vmbus_chan_attribute chan_attr_##_name = __ATTR_RO(_name)
#define VMBUS_CHAN_ATTR_WO(_name) \
	struct vmbus_chan_attribute chan_attr_##_name = __ATTR_WO(_name)

static ssize_t vmbus_chan_attr_show(struct kobject *kobj,
				    struct attribute *attr, char *buf)
{
	const struct vmbus_chan_attribute *attribute
		= container_of(attr, struct vmbus_chan_attribute, attr);
1502
	struct vmbus_channel *chan
1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514
		= container_of(kobj, struct vmbus_channel, kobj);

	if (!attribute->show)
		return -EIO;

	return attribute->show(chan, buf);
}

static const struct sysfs_ops vmbus_chan_sysfs_ops = {
	.show = vmbus_chan_attr_show,
};

1515
static ssize_t out_mask_show(struct vmbus_channel *channel, char *buf)
1516
{
1517 1518
	struct hv_ring_buffer_info *rbi = &channel->outbound;
	ssize_t ret;
1519

1520 1521 1522
	mutex_lock(&rbi->ring_buffer_mutex);
	if (!rbi->ring_buffer) {
		mutex_unlock(&rbi->ring_buffer_mutex);
1523
		return -EINVAL;
1524
	}
1525

1526 1527 1528
	ret = sprintf(buf, "%u\n", rbi->ring_buffer->interrupt_mask);
	mutex_unlock(&rbi->ring_buffer_mutex);
	return ret;
1529
}
1530
static VMBUS_CHAN_ATTR_RO(out_mask);
1531

1532
static ssize_t in_mask_show(struct vmbus_channel *channel, char *buf)
1533
{
1534 1535
	struct hv_ring_buffer_info *rbi = &channel->inbound;
	ssize_t ret;
1536

1537 1538 1539
	mutex_lock(&rbi->ring_buffer_mutex);
	if (!rbi->ring_buffer) {
		mutex_unlock(&rbi->ring_buffer_mutex);
1540
		return -EINVAL;
1541
	}
1542

1543 1544 1545
	ret = sprintf(buf, "%u\n", rbi->ring_buffer->interrupt_mask);
	mutex_unlock(&rbi->ring_buffer_mutex);
	return ret;
1546
}
1547
static VMBUS_CHAN_ATTR_RO(in_mask);
1548

1549
static ssize_t read_avail_show(struct vmbus_channel *channel, char *buf)
1550
{
1551 1552
	struct hv_ring_buffer_info *rbi = &channel->inbound;
	ssize_t ret;
1553

1554 1555 1556
	mutex_lock(&rbi->ring_buffer_mutex);
	if (!rbi->ring_buffer) {
		mutex_unlock(&rbi->ring_buffer_mutex);
1557
		return -EINVAL;
1558
	}
1559

1560 1561 1562
	ret = sprintf(buf, "%u\n", hv_get_bytes_to_read(rbi));
	mutex_unlock(&rbi->ring_buffer_mutex);
	return ret;
1563
}
1564
static VMBUS_CHAN_ATTR_RO(read_avail);
1565

1566
static ssize_t write_avail_show(struct vmbus_channel *channel, char *buf)
1567
{
1568 1569
	struct hv_ring_buffer_info *rbi = &channel->outbound;
	ssize_t ret;
1570

1571 1572 1573
	mutex_lock(&rbi->ring_buffer_mutex);
	if (!rbi->ring_buffer) {
		mutex_unlock(&rbi->ring_buffer_mutex);
1574
		return -EINVAL;
1575
	}
1576

1577 1578 1579
	ret = sprintf(buf, "%u\n", hv_get_bytes_to_write(rbi));
	mutex_unlock(&rbi->ring_buffer_mutex);
	return ret;
1580
}
1581
static VMBUS_CHAN_ATTR_RO(write_avail);
1582

1583
static ssize_t show_target_cpu(struct vmbus_channel *channel, char *buf)
1584 1585 1586
{
	return sprintf(buf, "%u\n", channel->target_cpu);
}
1587
static VMBUS_CHAN_ATTR(cpu, S_IRUGO, show_target_cpu, NULL);
1588

1589
static ssize_t channel_pending_show(struct vmbus_channel *channel,
1590 1591 1592 1593 1594 1595
				    char *buf)
{
	return sprintf(buf, "%d\n",
		       channel_pending(channel,
				       vmbus_connection.monitor_pages[1]));
}
1596
static VMBUS_CHAN_ATTR(pending, S_IRUGO, channel_pending_show, NULL);
1597

1598
static ssize_t channel_latency_show(struct vmbus_channel *channel,
1599 1600 1601 1602 1603 1604
				    char *buf)
{
	return sprintf(buf, "%d\n",
		       channel_latency(channel,
				       vmbus_connection.monitor_pages[1]));
}
1605
static VMBUS_CHAN_ATTR(latency, S_IRUGO, channel_latency_show, NULL);
1606

1607
static ssize_t channel_interrupts_show(struct vmbus_channel *channel, char *buf)
1608 1609 1610
{
	return sprintf(buf, "%llu\n", channel->interrupts);
}
1611
static VMBUS_CHAN_ATTR(interrupts, S_IRUGO, channel_interrupts_show, NULL);
1612

1613
static ssize_t channel_events_show(struct vmbus_channel *channel, char *buf)
1614 1615 1616
{
	return sprintf(buf, "%llu\n", channel->sig_events);
}
1617
static VMBUS_CHAN_ATTR(events, S_IRUGO, channel_events_show, NULL);
1618

1619
static ssize_t channel_intr_in_full_show(struct vmbus_channel *channel,
1620 1621 1622 1623 1624 1625 1626
					 char *buf)
{
	return sprintf(buf, "%llu\n",
		       (unsigned long long)channel->intr_in_full);
}
static VMBUS_CHAN_ATTR(intr_in_full, 0444, channel_intr_in_full_show, NULL);

1627
static ssize_t channel_intr_out_empty_show(struct vmbus_channel *channel,
1628 1629 1630 1631 1632 1633 1634
					   char *buf)
{
	return sprintf(buf, "%llu\n",
		       (unsigned long long)channel->intr_out_empty);
}
static VMBUS_CHAN_ATTR(intr_out_empty, 0444, channel_intr_out_empty_show, NULL);

1635
static ssize_t channel_out_full_first_show(struct vmbus_channel *channel,
1636 1637 1638 1639 1640 1641 1642
					   char *buf)
{
	return sprintf(buf, "%llu\n",
		       (unsigned long long)channel->out_full_first);
}
static VMBUS_CHAN_ATTR(out_full_first, 0444, channel_out_full_first_show, NULL);

1643
static ssize_t channel_out_full_total_show(struct vmbus_channel *channel,
1644 1645 1646 1647 1648 1649 1650
					   char *buf)
{
	return sprintf(buf, "%llu\n",
		       (unsigned long long)channel->out_full_total);
}
static VMBUS_CHAN_ATTR(out_full_total, 0444, channel_out_full_total_show, NULL);

1651
static ssize_t subchannel_monitor_id_show(struct vmbus_channel *channel,
1652 1653 1654 1655 1656 1657
					  char *buf)
{
	return sprintf(buf, "%u\n", channel->offermsg.monitorid);
}
static VMBUS_CHAN_ATTR(monitor_id, S_IRUGO, subchannel_monitor_id_show, NULL);

1658
static ssize_t subchannel_id_show(struct vmbus_channel *channel,
1659 1660 1661 1662 1663 1664 1665
				  char *buf)
{
	return sprintf(buf, "%u\n",
		       channel->offermsg.offer.sub_channel_index);
}
static VMBUS_CHAN_ATTR_RO(subchannel_id);

1666 1667 1668 1669 1670 1671 1672 1673
static struct attribute *vmbus_chan_attrs[] = {
	&chan_attr_out_mask.attr,
	&chan_attr_in_mask.attr,
	&chan_attr_read_avail.attr,
	&chan_attr_write_avail.attr,
	&chan_attr_cpu.attr,
	&chan_attr_pending.attr,
	&chan_attr_latency.attr,
1674 1675
	&chan_attr_interrupts.attr,
	&chan_attr_events.attr,
1676 1677 1678 1679
	&chan_attr_intr_in_full.attr,
	&chan_attr_intr_out_empty.attr,
	&chan_attr_out_full_first.attr,
	&chan_attr_out_full_total.attr,
1680 1681
	&chan_attr_monitor_id.attr,
	&chan_attr_subchannel_id.attr,
1682 1683 1684
	NULL
};

1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709
/*
 * Channel-level attribute_group callback function. Returns the permission for
 * each attribute, and returns 0 if an attribute is not visible.
 */
static umode_t vmbus_chan_attr_is_visible(struct kobject *kobj,
					  struct attribute *attr, int idx)
{
	const struct vmbus_channel *channel =
		container_of(kobj, struct vmbus_channel, kobj);

	/* Hide the monitor attributes if the monitor mechanism is not used. */
	if (!channel->offermsg.monitor_allocated &&
	    (attr == &chan_attr_pending.attr ||
	     attr == &chan_attr_latency.attr ||
	     attr == &chan_attr_monitor_id.attr))
		return 0;

	return attr->mode;
}

static struct attribute_group vmbus_chan_group = {
	.attrs = vmbus_chan_attrs,
	.is_visible = vmbus_chan_attr_is_visible
};

1710 1711 1712 1713 1714 1715 1716 1717 1718 1719
static struct kobj_type vmbus_chan_ktype = {
	.sysfs_ops = &vmbus_chan_sysfs_ops,
	.release = vmbus_chan_release,
};

/*
 * vmbus_add_channel_kobj - setup a sub-directory under device/channels
 */
int vmbus_add_channel_kobj(struct hv_device *dev, struct vmbus_channel *channel)
{
1720
	const struct device *device = &dev->device;
1721 1722 1723 1724 1725 1726 1727 1728 1729 1730
	struct kobject *kobj = &channel->kobj;
	u32 relid = channel->offermsg.child_relid;
	int ret;

	kobj->kset = dev->channels_kset;
	ret = kobject_init_and_add(kobj, &vmbus_chan_ktype, NULL,
				   "%u", relid);
	if (ret)
		return ret;

1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741
	ret = sysfs_create_group(kobj, &vmbus_chan_group);

	if (ret) {
		/*
		 * The calling functions' error handling paths will cleanup the
		 * empty channel directory.
		 */
		dev_err(device, "Unable to set up channel sysfs files\n");
		return ret;
	}

1742 1743 1744 1745 1746
	kobject_uevent(kobj, KOBJ_ADD);

	return 0;
}

1747 1748 1749 1750 1751 1752 1753 1754
/*
 * vmbus_remove_channel_attr_group - remove the channel's attribute group
 */
void vmbus_remove_channel_attr_group(struct vmbus_channel *channel)
{
	sysfs_remove_group(&channel->kobj, &vmbus_chan_group);
}

1755
/*
1756
 * vmbus_device_create - Creates and registers a new child device
1757
 * on the vmbus.
1758
 */
1759 1760
struct hv_device *vmbus_device_create(const guid_t *type,
				      const guid_t *instance,
S
stephen hemminger 已提交
1761
				      struct vmbus_channel *channel)
1762
{
1763
	struct hv_device *child_device_obj;
1764

1765 1766
	child_device_obj = kzalloc(sizeof(struct hv_device), GFP_KERNEL);
	if (!child_device_obj) {
1767
		pr_err("Unable to allocate device object for child device\n");
1768 1769 1770
		return NULL;
	}

1771
	child_device_obj->channel = channel;
1772 1773
	guid_copy(&child_device_obj->dev_type, type);
	guid_copy(&child_device_obj->dev_instance, instance);
1774
	child_device_obj->vendor_id = 0x1414; /* MSFT vendor ID */
1775 1776 1777 1778

	return child_device_obj;
}

1779
/*
1780
 * vmbus_device_register - Register the child device
1781
 */
1782
int vmbus_device_register(struct hv_device *child_device_obj)
1783
{
1784 1785
	struct kobject *kobj = &child_device_obj->device.kobj;
	int ret;
1786

1787
	dev_set_name(&child_device_obj->device, "%pUl",
1788
		     child_device_obj->channel->offermsg.offer.if_instance.b);
1789

1790
	child_device_obj->device.bus = &hv_bus;
1791
	child_device_obj->device.parent = &hv_acpi_dev->dev;
1792
	child_device_obj->device.release = vmbus_device_release;
1793

1794 1795 1796 1797
	/*
	 * Register with the LDM. This will kick off the driver/device
	 * binding...which will eventually call vmbus_match() and vmbus_probe()
	 */
1798
	ret = device_register(&child_device_obj->device);
1799
	if (ret) {
1800
		pr_err("Unable to register child device\n");
1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821
		return ret;
	}

	child_device_obj->channels_kset = kset_create_and_add("channels",
							      NULL, kobj);
	if (!child_device_obj->channels_kset) {
		ret = -ENOMEM;
		goto err_dev_unregister;
	}

	ret = vmbus_add_channel_kobj(child_device_obj,
				     child_device_obj->channel);
	if (ret) {
		pr_err("Unable to register primary channeln");
		goto err_kset_unregister;
	}

	return 0;

err_kset_unregister:
	kset_unregister(child_device_obj->channels_kset);
1822

1823 1824
err_dev_unregister:
	device_unregister(&child_device_obj->device);
1825 1826 1827
	return ret;
}

1828
/*
1829
 * vmbus_device_unregister - Remove the specified child device
1830
 * from the vmbus.
1831
 */
1832
void vmbus_device_unregister(struct hv_device *device_obj)
1833
{
1834 1835 1836
	pr_debug("child device %s unregistered\n",
		dev_name(&device_obj->device));

1837 1838
	kset_unregister(device_obj->channels_kset);

1839 1840 1841 1842
	/*
	 * Kick off the process of unregistering the device.
	 * This will call vmbus_remove() and eventually vmbus_device_release()
	 */
1843
	device_unregister(&device_obj->device);
1844 1845 1846
}


1847
/*
1848
 * VMBUS is an acpi enumerated device. Get the information we
1849
 * need from DSDT.
1850
 */
1851
#define VTPM_BASE_ADDRESS 0xfed40000
1852
static acpi_status vmbus_walk_resources(struct acpi_resource *res, void *ctx)
1853
{
1854 1855 1856 1857 1858 1859
	resource_size_t start = 0;
	resource_size_t end = 0;
	struct resource *new_res;
	struct resource **old_res = &hyperv_mmio;
	struct resource **prev_res = NULL;

1860
	switch (res->type) {
1861 1862 1863 1864 1865 1866 1867 1868 1869

	/*
	 * "Address" descriptors are for bus windows. Ignore
	 * "memory" descriptors, which are for registers on
	 * devices.
	 */
	case ACPI_RESOURCE_TYPE_ADDRESS32:
		start = res->data.address32.address.minimum;
		end = res->data.address32.address.maximum;
G
Gerd Hoffmann 已提交
1870
		break;
1871

1872
	case ACPI_RESOURCE_TYPE_ADDRESS64:
1873 1874
		start = res->data.address64.address.minimum;
		end = res->data.address64.address.maximum;
G
Gerd Hoffmann 已提交
1875
		break;
1876 1877 1878 1879 1880

	default:
		/* Unused resource type */
		return AE_OK;

1881
	}
1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901
	/*
	 * Ignore ranges that are below 1MB, as they're not
	 * necessary or useful here.
	 */
	if (end < 0x100000)
		return AE_OK;

	new_res = kzalloc(sizeof(*new_res), GFP_ATOMIC);
	if (!new_res)
		return AE_NO_MEMORY;

	/* If this range overlaps the virtual TPM, truncate it. */
	if (end > VTPM_BASE_ADDRESS && start < VTPM_BASE_ADDRESS)
		end = VTPM_BASE_ADDRESS;

	new_res->name = "hyperv mmio";
	new_res->flags = IORESOURCE_MEM;
	new_res->start = start;
	new_res->end = end;

1902 1903 1904
	/*
	 * If two ranges are adjacent, merge them.
	 */
1905 1906 1907 1908 1909 1910
	do {
		if (!*old_res) {
			*old_res = new_res;
			break;
		}

1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922
		if (((*old_res)->end + 1) == new_res->start) {
			(*old_res)->end = new_res->end;
			kfree(new_res);
			break;
		}

		if ((*old_res)->start == new_res->end + 1) {
			(*old_res)->start = new_res->start;
			kfree(new_res);
			break;
		}

1923
		if ((*old_res)->start > new_res->end) {
1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934
			new_res->sibling = *old_res;
			if (prev_res)
				(*prev_res)->sibling = new_res;
			*old_res = new_res;
			break;
		}

		prev_res = old_res;
		old_res = &(*old_res)->sibling;

	} while (1);
1935 1936 1937 1938

	return AE_OK;
}

1939 1940 1941 1942 1943 1944
static int vmbus_acpi_remove(struct acpi_device *device)
{
	struct resource *cur_res;
	struct resource *next_res;

	if (hyperv_mmio) {
1945 1946 1947 1948 1949 1950
		if (fb_mmio) {
			__release_region(hyperv_mmio, fb_mmio->start,
					 resource_size(fb_mmio));
			fb_mmio = NULL;
		}

1951 1952 1953 1954 1955 1956 1957 1958 1959
		for (cur_res = hyperv_mmio; cur_res; cur_res = next_res) {
			next_res = cur_res->sibling;
			kfree(cur_res);
		}
	}

	return 0;
}

1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983
static void vmbus_reserve_fb(void)
{
	int size;
	/*
	 * Make a claim for the frame buffer in the resource tree under the
	 * first node, which will be the one below 4GB.  The length seems to
	 * be underreported, particularly in a Generation 1 VM.  So start out
	 * reserving a larger area and make it smaller until it succeeds.
	 */

	if (screen_info.lfb_base) {
		if (efi_enabled(EFI_BOOT))
			size = max_t(__u32, screen_info.lfb_size, 0x800000);
		else
			size = max_t(__u32, screen_info.lfb_size, 0x4000000);

		for (; !fb_mmio && (size >= 0x100000); size >>= 1) {
			fb_mmio = __request_region(hyperv_mmio,
						   screen_info.lfb_base, size,
						   fb_mmio_name, 0);
		}
	}
}

1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
/**
 * vmbus_allocate_mmio() - Pick a memory-mapped I/O range.
 * @new:		If successful, supplied a pointer to the
 *			allocated MMIO space.
 * @device_obj:		Identifies the caller
 * @min:		Minimum guest physical address of the
 *			allocation
 * @max:		Maximum guest physical address
 * @size:		Size of the range to be allocated
 * @align:		Alignment of the range to be allocated
 * @fb_overlap_ok:	Whether this allocation can be allowed
 *			to overlap the video frame buffer.
 *
 * This function walks the resources granted to VMBus by the
 * _CRS object in the ACPI namespace underneath the parent
 * "bridge" whether that's a root PCI bus in the Generation 1
 * case or a Module Device in the Generation 2 case.  It then
 * attempts to allocate from the global MMIO pool in a way that
 * matches the constraints supplied in these parameters and by
 * that _CRS.
 *
 * Return: 0 on success, -errno on failure
 */
int vmbus_allocate_mmio(struct resource **new, struct hv_device *device_obj,
			resource_size_t min, resource_size_t max,
			resource_size_t size, resource_size_t align,
			bool fb_overlap_ok)
{
2012
	struct resource *iter, *shadow;
2013
	resource_size_t range_min, range_max, start;
2014
	const char *dev_n = dev_name(&device_obj->device);
2015
	int retval;
2016 2017 2018

	retval = -ENXIO;
	down(&hyperv_mmio_lock);
2019

2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039
	/*
	 * If overlaps with frame buffers are allowed, then first attempt to
	 * make the allocation from within the reserved region.  Because it
	 * is already reserved, no shadow allocation is necessary.
	 */
	if (fb_overlap_ok && fb_mmio && !(min > fb_mmio->end) &&
	    !(max < fb_mmio->start)) {

		range_min = fb_mmio->start;
		range_max = fb_mmio->end;
		start = (range_min + align - 1) & ~(align - 1);
		for (; start + size - 1 <= range_max; start += align) {
			*new = request_mem_region_exclusive(start, size, dev_n);
			if (*new) {
				retval = 0;
				goto exit;
			}
		}
	}

2040 2041 2042 2043 2044 2045
	for (iter = hyperv_mmio; iter; iter = iter->sibling) {
		if ((iter->start >= max) || (iter->end <= min))
			continue;

		range_min = iter->start;
		range_max = iter->end;
2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057
		start = (range_min + align - 1) & ~(align - 1);
		for (; start + size - 1 <= range_max; start += align) {
			shadow = __request_region(iter, start, size, NULL,
						  IORESOURCE_BUSY);
			if (!shadow)
				continue;

			*new = request_mem_region_exclusive(start, size, dev_n);
			if (*new) {
				shadow->name = (char *)*new;
				retval = 0;
				goto exit;
2058 2059
			}

2060
			__release_region(iter, start, size);
2061 2062 2063
		}
	}

2064 2065 2066
exit:
	up(&hyperv_mmio_lock);
	return retval;
2067 2068 2069
}
EXPORT_SYMBOL_GPL(vmbus_allocate_mmio);

2070 2071 2072 2073 2074 2075 2076 2077 2078 2079
/**
 * vmbus_free_mmio() - Free a memory-mapped I/O range.
 * @start:		Base address of region to release.
 * @size:		Size of the range to be allocated
 *
 * This function releases anything requested by
 * vmbus_mmio_allocate().
 */
void vmbus_free_mmio(resource_size_t start, resource_size_t size)
{
2080 2081 2082 2083 2084 2085 2086 2087 2088
	struct resource *iter;

	down(&hyperv_mmio_lock);
	for (iter = hyperv_mmio; iter; iter = iter->sibling) {
		if ((iter->start >= start + size) || (iter->end <= start))
			continue;

		__release_region(iter, start, size);
	}
2089
	release_mem_region(start, size);
2090
	up(&hyperv_mmio_lock);
2091 2092 2093 2094

}
EXPORT_SYMBOL_GPL(vmbus_free_mmio);

2095 2096 2097
static int vmbus_acpi_add(struct acpi_device *device)
{
	acpi_status result;
2098
	int ret_val = -ENODEV;
2099
	struct acpi_device *ancestor;
2100

2101 2102
	hv_acpi_dev = device;

2103
	result = acpi_walk_resources(device->handle, METHOD_NAME__CRS,
2104
					vmbus_walk_resources, NULL);
2105

2106 2107 2108
	if (ACPI_FAILURE(result))
		goto acpi_walk_err;
	/*
2109 2110
	 * Some ancestor of the vmbus acpi device (Gen1 or Gen2
	 * firmware) is the VMOD that has the mmio ranges. Get that.
2111
	 */
2112 2113 2114
	for (ancestor = device->parent; ancestor; ancestor = ancestor->parent) {
		result = acpi_walk_resources(ancestor->handle, METHOD_NAME__CRS,
					     vmbus_walk_resources, NULL);
2115 2116

		if (ACPI_FAILURE(result))
2117
			continue;
2118 2119
		if (hyperv_mmio) {
			vmbus_reserve_fb();
2120
			break;
2121
		}
2122
	}
2123 2124 2125
	ret_val = 0;

acpi_walk_err:
2126
	complete(&probe_event);
2127 2128
	if (ret_val)
		vmbus_acpi_remove(device);
2129
	return ret_val;
2130 2131
}

2132
#ifdef CONFIG_PM_SLEEP
2133 2134
static int vmbus_bus_suspend(struct device *dev)
{
2135 2136
	struct vmbus_channel *channel, *sc;
	unsigned long flags;
2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154

	while (atomic_read(&vmbus_connection.offer_in_progress) != 0) {
		/*
		 * We wait here until the completion of any channel
		 * offers that are currently in progress.
		 */
		msleep(1);
	}

	mutex_lock(&vmbus_connection.channel_mutex);
	list_for_each_entry(channel, &vmbus_connection.chn_list, listentry) {
		if (!is_hvsock_channel(channel))
			continue;

		vmbus_force_channel_rescinded(channel);
	}
	mutex_unlock(&vmbus_connection.channel_mutex);

2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171
	/*
	 * Wait until all the sub-channels and hv_sock channels have been
	 * cleaned up. Sub-channels should be destroyed upon suspend, otherwise
	 * they would conflict with the new sub-channels that will be created
	 * in the resume path. hv_sock channels should also be destroyed, but
	 * a hv_sock channel of an established hv_sock connection can not be
	 * really destroyed since it may still be referenced by the userspace
	 * application, so we just force the hv_sock channel to be rescinded
	 * by vmbus_force_channel_rescinded(), and the userspace application
	 * will thoroughly destroy the channel after hibernation.
	 *
	 * Note: the counter nr_chan_close_on_suspend may never go above 0 if
	 * the VM has no sub-channel and hv_sock channel, e.g. a 1-vCPU VM.
	 */
	if (atomic_read(&vmbus_connection.nr_chan_close_on_suspend) > 0)
		wait_for_completion(&vmbus_connection.ready_for_suspend_event);

2172 2173
	WARN_ON(atomic_read(&vmbus_connection.nr_chan_fixup_on_resume) != 0);

2174 2175 2176
	mutex_lock(&vmbus_connection.channel_mutex);

	list_for_each_entry(channel, &vmbus_connection.chn_list, listentry) {
2177 2178 2179 2180 2181 2182
		/*
		 * Invalidate the field. Upon resume, vmbus_onoffer() will fix
		 * up the field, and the other fields (if necessary).
		 */
		channel->offermsg.child_relid = INVALID_RELID;

2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196
		if (is_hvsock_channel(channel)) {
			if (!channel->rescind) {
				pr_err("hv_sock channel not rescinded!\n");
				WARN_ON_ONCE(1);
			}
			continue;
		}

		spin_lock_irqsave(&channel->lock, flags);
		list_for_each_entry(sc, &channel->sc_list, sc_list) {
			pr_err("Sub-channel not deleted!\n");
			WARN_ON_ONCE(1);
		}
		spin_unlock_irqrestore(&channel->lock, flags);
2197 2198

		atomic_inc(&vmbus_connection.nr_chan_fixup_on_resume);
2199 2200 2201 2202
	}

	mutex_unlock(&vmbus_connection.channel_mutex);

2203 2204 2205 2206
	vmbus_initiate_unload(false);

	vmbus_connection.conn_state = DISCONNECTED;

2207 2208 2209
	/* Reset the event for the next resume. */
	reinit_completion(&vmbus_connection.ready_for_resume_event);

2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222
	return 0;
}

static int vmbus_bus_resume(struct device *dev)
{
	struct vmbus_channel_msginfo *msginfo;
	size_t msgsize;
	int ret;

	/*
	 * We only use the 'vmbus_proto_version', which was in use before
	 * hibernation, to re-negotiate with the host.
	 */
2223
	if (!vmbus_proto_version) {
2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242
		pr_err("Invalid proto version = 0x%x\n", vmbus_proto_version);
		return -EINVAL;
	}

	msgsize = sizeof(*msginfo) +
		  sizeof(struct vmbus_channel_initiate_contact);

	msginfo = kzalloc(msgsize, GFP_KERNEL);

	if (msginfo == NULL)
		return -ENOMEM;

	ret = vmbus_negotiate_version(msginfo, vmbus_proto_version);

	kfree(msginfo);

	if (ret != 0)
		return ret;

2243 2244
	WARN_ON(atomic_read(&vmbus_connection.nr_chan_fixup_on_resume) == 0);

2245 2246
	vmbus_request_offers();

2247 2248
	wait_for_completion(&vmbus_connection.ready_for_resume_event);

2249 2250 2251
	/* Reset the event for the next suspend. */
	reinit_completion(&vmbus_connection.ready_for_suspend_event);

2252 2253
	return 0;
}
2254
#endif /* CONFIG_PM_SLEEP */
2255

2256 2257
static const struct acpi_device_id vmbus_acpi_device_ids[] = {
	{"VMBUS", 0},
2258
	{"VMBus", 0},
2259 2260 2261 2262
	{"", 0},
};
MODULE_DEVICE_TABLE(acpi, vmbus_acpi_device_ids);

2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275
/*
 * Note: we must use SET_NOIRQ_SYSTEM_SLEEP_PM_OPS rather than
 * SET_SYSTEM_SLEEP_PM_OPS, otherwise NIC SR-IOV can not work, because the
 * "pci_dev_pm_ops" uses the "noirq" callbacks: in the resume path, the
 * pci "noirq" restore callback runs before "non-noirq" callbacks (see
 * resume_target_kernel() -> dpm_resume_start(), and hibernation_restore() ->
 * dpm_resume_end()). This means vmbus_bus_resume() and the pci-hyperv's
 * resume callback must also run via the "noirq" callbacks.
 */
static const struct dev_pm_ops vmbus_bus_pm = {
	SET_NOIRQ_SYSTEM_SLEEP_PM_OPS(vmbus_bus_suspend, vmbus_bus_resume)
};

2276 2277 2278 2279 2280
static struct acpi_driver vmbus_acpi_driver = {
	.name = "vmbus",
	.ids = vmbus_acpi_device_ids,
	.ops = {
		.add = vmbus_acpi_add,
2281
		.remove = vmbus_acpi_remove,
2282
	},
2283
	.drv.pm = &vmbus_bus_pm,
2284 2285
};

2286 2287
static void hv_kexec_handler(void)
{
2288
	hv_stimer_global_cleanup();
2289
	vmbus_initiate_unload(false);
2290 2291 2292
	vmbus_connection.conn_state = DISCONNECTED;
	/* Make sure conn_state is set as hv_synic_cleanup checks for it */
	mb();
2293
	cpuhp_remove_state(hyperv_cpuhp_online);
2294
	hyperv_cleanup();
2295 2296
};

2297 2298
static void hv_crash_handler(struct pt_regs *regs)
{
2299 2300
	int cpu;

2301
	vmbus_initiate_unload(true);
2302 2303 2304 2305 2306
	/*
	 * In crash handler we can't schedule synic cleanup for all CPUs,
	 * doing the cleanup for current CPU only. This should be sufficient
	 * for kdump.
	 */
2307
	vmbus_connection.conn_state = DISCONNECTED;
2308 2309 2310
	cpu = smp_processor_id();
	hv_stimer_cleanup(cpu);
	hv_synic_cleanup(cpu);
2311
	hyperv_cleanup();
2312 2313
};

2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354
static int hv_synic_suspend(void)
{
	/*
	 * When we reach here, all the non-boot CPUs have been offlined, and
	 * the stimers on them have been unbound in hv_synic_cleanup() ->
	 * hv_stimer_cleanup() -> clockevents_unbind_device().
	 *
	 * hv_synic_suspend() only runs on CPU0 with interrupts disabled. Here
	 * we do not unbind the stimer on CPU0 because: 1) it's unnecessary
	 * because the interrupts remain disabled between syscore_suspend()
	 * and syscore_resume(): see create_image() and resume_target_kernel();
	 * 2) the stimer on CPU0 is automatically disabled later by
	 * syscore_suspend() -> timekeeping_suspend() -> tick_suspend() -> ...
	 * -> clockevents_shutdown() -> ... -> hv_ce_shutdown(); 3) a warning
	 * would be triggered if we call clockevents_unbind_device(), which
	 * may sleep, in an interrupts-disabled context. So, we intentionally
	 * don't call hv_stimer_cleanup(0) here.
	 */

	hv_synic_disable_regs(0);

	return 0;
}

static void hv_synic_resume(void)
{
	hv_synic_enable_regs(0);

	/*
	 * Note: we don't need to call hv_stimer_init(0), because the timer
	 * on CPU0 is not unbound in hv_synic_suspend(), and the timer is
	 * automatically re-enabled in timekeeping_resume().
	 */
}

/* The callbacks run only on CPU0, with irqs_disabled. */
static struct syscore_ops hv_synic_syscore_ops = {
	.suspend = hv_synic_suspend,
	.resume = hv_synic_resume,
};

2355
static int __init hv_acpi_init(void)
2356
{
2357
	int ret, t;
2358

2359
	if (!hv_is_hyperv_initialized())
2360 2361
		return -ENODEV;

2362 2363 2364
	init_completion(&probe_event);

	/*
2365
	 * Get ACPI resources first.
2366
	 */
2367 2368
	ret = acpi_bus_register_driver(&vmbus_acpi_driver);

2369 2370 2371
	if (ret)
		return ret;

2372 2373 2374 2375 2376
	t = wait_for_completion_timeout(&probe_event, 5*HZ);
	if (t == 0) {
		ret = -ETIMEDOUT;
		goto cleanup;
	}
2377

2378
	ret = vmbus_bus_init();
2379
	if (ret)
2380 2381
		goto cleanup;

2382
	hv_setup_kexec_handler(hv_kexec_handler);
2383
	hv_setup_crash_handler(hv_crash_handler);
2384

2385 2386
	register_syscore_ops(&hv_synic_syscore_ops);

2387 2388 2389 2390
	return 0;

cleanup:
	acpi_bus_unregister_driver(&vmbus_acpi_driver);
2391
	hv_acpi_dev = NULL;
2392
	return ret;
2393 2394
}

2395 2396
static void __exit vmbus_exit(void)
{
2397 2398
	int cpu;

2399 2400
	unregister_syscore_ops(&hv_synic_syscore_ops);

2401
	hv_remove_kexec_handler();
2402
	hv_remove_crash_handler();
2403
	vmbus_connection.conn_state = DISCONNECTED;
2404
	hv_stimer_global_cleanup();
2405
	vmbus_disconnect();
2406
	hv_remove_vmbus_irq();
2407 2408 2409 2410 2411 2412
	for_each_online_cpu(cpu) {
		struct hv_per_cpu_context *hv_cpu
			= per_cpu_ptr(hv_context.cpu_context, cpu);

		tasklet_kill(&hv_cpu->msg_dpc);
	}
2413
	vmbus_free_channels();
2414

2415
	if (ms_hyperv.misc_features & HV_FEATURE_GUEST_CRASH_MSR_AVAILABLE) {
2416
		kmsg_dump_unregister(&hv_kmsg_dumper);
2417
		unregister_die_notifier(&hyperv_die_block);
2418 2419 2420
		atomic_notifier_chain_unregister(&panic_notifier_list,
						 &hyperv_panic_block);
	}
2421 2422

	free_page((unsigned long)hv_panic_page);
2423 2424
	unregister_sysctl_table(hv_ctl_table_hdr);
	hv_ctl_table_hdr = NULL;
2425
	bus_unregister(&hv_bus);
2426

2427
	cpuhp_remove_state(hyperv_cpuhp_online);
2428
	hv_synic_free();
2429 2430 2431
	acpi_bus_unregister_driver(&vmbus_acpi_driver);
}

2432

2433
MODULE_LICENSE("GPL");
2434
MODULE_DESCRIPTION("Microsoft Hyper-V VMBus Driver");
2435

2436
subsys_initcall(hv_acpi_init);
2437
module_exit(vmbus_exit);