mmconfig-shared.c 18.4 KB
Newer Older
1 2 3 4 5
/*
 * mmconfig-shared.c - Low-level direct PCI config space access via
 *                     MMCONFIG - common code between i386 and x86-64.
 *
 * This code does:
6
 * - known chipset handling
7 8 9 10 11 12 13 14 15
 * - ACPI decoding and validation
 *
 * Per-architecture code takes care of the mappings and accesses
 * themselves.
 */

#include <linux/pci.h>
#include <linux/init.h>
#include <linux/acpi.h>
F
Feng Tang 已提交
16
#include <linux/sfi_acpi.h>
17
#include <linux/bitmap.h>
B
Bjorn Helgaas 已提交
18
#include <linux/dmi.h>
19
#include <linux/slab.h>
20 21
#include <linux/mutex.h>
#include <linux/rculist.h>
22
#include <asm/e820.h>
23
#include <asm/pci_x86.h>
F
Feng Tang 已提交
24
#include <asm/acpi.h>
25

26
#define PREFIX "PCI: "
27

28
/* Indicate if the mmcfg resources have been placed into the resource table. */
29
static bool pci_mmcfg_running_state;
30
static bool pci_mmcfg_arch_init_failed;
31
static DEFINE_MUTEX(pci_mmcfg_lock);
32

33 34
LIST_HEAD(pci_mmcfg_list);

35 36 37 38 39 40 41 42
static __init void pci_mmconfig_remove(struct pci_mmcfg_region *cfg)
{
	if (cfg->res.parent)
		release_resource(&cfg->res);
	list_del(&cfg->list);
	kfree(cfg);
}

43 44
static __init void free_all_mmcfg(void)
{
45
	struct pci_mmcfg_region *cfg, *tmp;
46

47
	pci_mmcfg_arch_free();
48 49
	list_for_each_entry_safe(cfg, tmp, &pci_mmcfg_list, list)
		pci_mmconfig_remove(cfg);
50 51
}

52
static void list_add_sorted(struct pci_mmcfg_region *new)
53 54 55 56
{
	struct pci_mmcfg_region *cfg;

	/* keep list sorted by segment and starting bus number */
57
	list_for_each_entry_rcu(cfg, &pci_mmcfg_list, list) {
58 59 60
		if (cfg->segment > new->segment ||
		    (cfg->segment == new->segment &&
		     cfg->start_bus >= new->start_bus)) {
61
			list_add_tail_rcu(&new->list, &cfg->list);
62 63 64
			return;
		}
	}
65
	list_add_tail_rcu(&new->list, &pci_mmcfg_list);
66 67
}

68 69
static struct pci_mmcfg_region *pci_mmconfig_alloc(int segment, int start,
						   int end, u64 addr)
70
{
71
	struct pci_mmcfg_region *new;
72
	struct resource *res;
73

74 75 76
	if (addr == 0)
		return NULL;

77
	new = kzalloc(sizeof(*new), GFP_KERNEL);
78
	if (!new)
79
		return NULL;
80

81 82 83 84
	new->address = addr;
	new->segment = segment;
	new->start_bus = start;
	new->end_bus = end;
85

86 87
	res = &new->res;
	res->start = addr + PCI_MMCFG_BUS_OFFSET(start);
88
	res->end = addr + PCI_MMCFG_BUS_OFFSET(end + 1) - 1;
89 90 91 92 93
	res->flags = IORESOURCE_MEM | IORESOURCE_BUSY;
	snprintf(new->name, PCI_MMCFG_RESOURCE_NAME_LEN,
		 "PCI MMCONFIG %04x [bus %02x-%02x]", segment, start, end);
	res->name = new->name;

94
	return new;
95 96
}

97 98 99 100 101 102
static __init struct pci_mmcfg_region *pci_mmconfig_add(int segment, int start,
							int end, u64 addr)
{
	struct pci_mmcfg_region *new;

	new = pci_mmconfig_alloc(segment, start, end, addr);
103 104
	if (new) {
		mutex_lock(&pci_mmcfg_lock);
105
		list_add_sorted(new);
106
		mutex_unlock(&pci_mmcfg_lock);
107

108
		pr_info(PREFIX
109 110 111
		       "MMCONFIG for domain %04x [bus %02x-%02x] at %pR "
		       "(base %#lx)\n",
		       segment, start, end, &new->res, (unsigned long)addr);
112
	}
113 114 115 116

	return new;
}

117 118 119 120
struct pci_mmcfg_region *pci_mmconfig_lookup(int segment, int bus)
{
	struct pci_mmcfg_region *cfg;

121
	list_for_each_entry_rcu(cfg, &pci_mmcfg_list, list)
122 123 124 125 126 127 128
		if (cfg->segment == segment &&
		    cfg->start_bus <= bus && bus <= cfg->end_bus)
			return cfg;

	return NULL;
}

129
static const char __init *pci_mmcfg_e7520(void)
130 131
{
	u32 win;
132
	raw_pci_ops->read(0, 0, PCI_DEVFN(0, 0), 0xce, 2, &win);
133

134
	win = win & 0xf000;
135 136 137
	if (win == 0x0000 || win == 0xf000)
		return NULL;

138
	if (pci_mmconfig_add(0, 0, 255, win << 16) == NULL)
139 140
		return NULL;

141 142 143
	return "Intel Corporation E7520 Memory Controller Hub";
}

144
static const char __init *pci_mmcfg_intel_945(void)
145 146 147
{
	u32 pciexbar, mask = 0, len = 0;

148
	raw_pci_ops->read(0, 0, PCI_DEVFN(0, 0), 0x48, 4, &pciexbar);
149 150 151

	/* Enable bit */
	if (!(pciexbar & 1))
152
		return NULL;
153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168

	/* Size bits */
	switch ((pciexbar >> 1) & 3) {
	case 0:
		mask = 0xf0000000U;
		len  = 0x10000000U;
		break;
	case 1:
		mask = 0xf8000000U;
		len  = 0x08000000U;
		break;
	case 2:
		mask = 0xfc000000U;
		len  = 0x04000000U;
		break;
	default:
169
		return NULL;
170 171 172 173 174 175
	}

	/* Errata #2, things break when not aligned on a 256Mb boundary */
	/* Can only happen in 64M/128M mode */

	if ((pciexbar & mask) & 0x0fffffffU)
176
		return NULL;
177

178 179
	/* Don't hit the APIC registers and their friends */
	if ((pciexbar & mask) >= 0xf0000000U)
180 181
		return NULL;

182
	if (pci_mmconfig_add(0, 0, (len >> 20) - 1, pciexbar & mask) == NULL)
183 184
		return NULL;

185 186 187
	return "Intel Corporation 945G/GZ/P/PL Express Memory Controller Hub";
}

188 189 190 191 192
static const char __init *pci_mmcfg_amd_fam10h(void)
{
	u32 low, high, address;
	u64 base, msr;
	int i;
193
	unsigned segnbits = 0, busnbits, end_bus;
194

195 196 197
	if (!(pci_probe & PCI_CHECK_ENABLE_AMD_MMCONF))
		return NULL;

198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226
	address = MSR_FAM10H_MMIO_CONF_BASE;
	if (rdmsr_safe(address, &low, &high))
		return NULL;

	msr = high;
	msr <<= 32;
	msr |= low;

	/* mmconfig is not enable */
	if (!(msr & FAM10H_MMIO_CONF_ENABLE))
		return NULL;

	base = msr & (FAM10H_MMIO_CONF_BASE_MASK<<FAM10H_MMIO_CONF_BASE_SHIFT);

	busnbits = (msr >> FAM10H_MMIO_CONF_BUSRANGE_SHIFT) &
			 FAM10H_MMIO_CONF_BUSRANGE_MASK;

	/*
	 * only handle bus 0 ?
	 * need to skip it
	 */
	if (!busnbits)
		return NULL;

	if (busnbits > 8) {
		segnbits = busnbits - 8;
		busnbits = 8;
	}

227
	end_bus = (1 << busnbits) - 1;
228
	for (i = 0; i < (1 << segnbits); i++)
229 230 231 232 233
		if (pci_mmconfig_add(i, 0, end_bus,
				     base + (1<<28) * i) == NULL) {
			free_all_mmcfg();
			return NULL;
		}
234 235 236 237

	return "AMD Family 10h NB";
}

238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257
static bool __initdata mcp55_checked;
static const char __init *pci_mmcfg_nvidia_mcp55(void)
{
	int bus;
	int mcp55_mmconf_found = 0;

	static const u32 extcfg_regnum		= 0x90;
	static const u32 extcfg_regsize		= 4;
	static const u32 extcfg_enable_mask	= 1<<31;
	static const u32 extcfg_start_mask	= 0xff<<16;
	static const int extcfg_start_shift	= 16;
	static const u32 extcfg_size_mask	= 0x3<<28;
	static const int extcfg_size_shift	= 28;
	static const int extcfg_sizebus[]	= {0x100, 0x80, 0x40, 0x20};
	static const u32 extcfg_base_mask[]	= {0x7ff8, 0x7ffc, 0x7ffe, 0x7fff};
	static const int extcfg_base_lshift	= 25;

	/*
	 * do check if amd fam10h already took over
	 */
258
	if (!acpi_disabled || !list_empty(&pci_mmcfg_list) || mcp55_checked)
259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286
		return NULL;

	mcp55_checked = true;
	for (bus = 0; bus < 256; bus++) {
		u64 base;
		u32 l, extcfg;
		u16 vendor, device;
		int start, size_index, end;

		raw_pci_ops->read(0, bus, PCI_DEVFN(0, 0), 0, 4, &l);
		vendor = l & 0xffff;
		device = (l >> 16) & 0xffff;

		if (PCI_VENDOR_ID_NVIDIA != vendor || 0x0369 != device)
			continue;

		raw_pci_ops->read(0, bus, PCI_DEVFN(0, 0), extcfg_regnum,
				  extcfg_regsize, &extcfg);

		if (!(extcfg & extcfg_enable_mask))
			continue;

		size_index = (extcfg & extcfg_size_mask) >> extcfg_size_shift;
		base = extcfg & extcfg_base_mask[size_index];
		/* base could > 4G */
		base <<= extcfg_base_lshift;
		start = (extcfg & extcfg_start_mask) >> extcfg_start_shift;
		end = start + extcfg_sizebus[size_index] - 1;
287 288
		if (pci_mmconfig_add(0, start, end, base) == NULL)
			continue;
289 290 291 292 293 294 295 296 297
		mcp55_mmconf_found++;
	}

	if (!mcp55_mmconf_found)
		return NULL;

	return "nVidia MCP55";
}

298
struct pci_mmcfg_hostbridge_probe {
299 300
	u32 bus;
	u32 devfn;
301 302 303 304 305
	u32 vendor;
	u32 device;
	const char *(*probe)(void);
};

306
static struct pci_mmcfg_hostbridge_probe pci_mmcfg_probes[] __initdata = {
307 308 309 310 311 312 313 314
	{ 0, PCI_DEVFN(0, 0), PCI_VENDOR_ID_INTEL,
	  PCI_DEVICE_ID_INTEL_E7520_MCH, pci_mmcfg_e7520 },
	{ 0, PCI_DEVFN(0, 0), PCI_VENDOR_ID_INTEL,
	  PCI_DEVICE_ID_INTEL_82945G_HB, pci_mmcfg_intel_945 },
	{ 0, PCI_DEVFN(0x18, 0), PCI_VENDOR_ID_AMD,
	  0x1200, pci_mmcfg_amd_fam10h },
	{ 0xff, PCI_DEVFN(0, 0), PCI_VENDOR_ID_AMD,
	  0x1200, pci_mmcfg_amd_fam10h },
315 316
	{ 0, PCI_DEVFN(0, 0), PCI_VENDOR_ID_NVIDIA,
	  0x0369, pci_mmcfg_nvidia_mcp55 },
317 318
};

319 320
static void __init pci_mmcfg_check_end_bus_number(void)
{
321
	struct pci_mmcfg_region *cfg, *cfgx;
322

323
	/* Fixup overlaps */
324
	list_for_each_entry(cfg, &pci_mmcfg_list, list) {
325 326
		if (cfg->end_bus < cfg->start_bus)
			cfg->end_bus = 255;
327

328 329 330 331
		/* Don't access the list head ! */
		if (cfg->list.next == &pci_mmcfg_list)
			break;

332
		cfgx = list_entry(cfg->list.next, typeof(*cfg), list);
333
		if (cfg->end_bus >= cfgx->start_bus)
334
			cfg->end_bus = cfgx->start_bus - 1;
335 336 337
	}
}

338 339 340
static int __init pci_mmcfg_check_hostbridge(void)
{
	u32 l;
341
	u32 bus, devfn;
342 343 344 345
	u16 vendor, device;
	int i;
	const char *name;

346 347 348
	if (!raw_pci_ops)
		return 0;

349
	free_all_mmcfg();
350

351
	for (i = 0; i < ARRAY_SIZE(pci_mmcfg_probes); i++) {
352 353
		bus =  pci_mmcfg_probes[i].bus;
		devfn = pci_mmcfg_probes[i].devfn;
354
		raw_pci_ops->read(0, bus, devfn, 0, 4, &l);
355 356 357
		vendor = l & 0xffff;
		device = (l >> 16) & 0xffff;

358
		name = NULL;
359 360
		if (pci_mmcfg_probes[i].vendor == vendor &&
		    pci_mmcfg_probes[i].device == device)
361 362
			name = pci_mmcfg_probes[i].probe();

363
		if (name)
364
			pr_info(PREFIX "%s with MMCONFIG support\n", name);
365 366
	}

367 368 369
	/* some end_bus_number is crazy, fix it */
	pci_mmcfg_check_end_bus_number();

370
	return !list_empty(&pci_mmcfg_list);
371 372
}

373
static acpi_status check_mcfg_resource(struct acpi_resource *res, void *data)
374 375 376 377 378 379 380 381 382 383 384
{
	struct resource *mcfg_res = data;
	struct acpi_resource_address64 address;
	acpi_status status;

	if (res->type == ACPI_RESOURCE_TYPE_FIXED_MEMORY32) {
		struct acpi_resource_fixed_memory32 *fixmem32 =
			&res->data.fixed_memory32;
		if (!fixmem32)
			return AE_OK;
		if ((mcfg_res->start >= fixmem32->address) &&
385
		    (mcfg_res->end < (fixmem32->address +
386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401
				      fixmem32->address_length))) {
			mcfg_res->flags = 1;
			return AE_CTRL_TERMINATE;
		}
	}
	if ((res->type != ACPI_RESOURCE_TYPE_ADDRESS32) &&
	    (res->type != ACPI_RESOURCE_TYPE_ADDRESS64))
		return AE_OK;

	status = acpi_resource_to_address64(res, &address);
	if (ACPI_FAILURE(status) ||
	   (address.address_length <= 0) ||
	   (address.resource_type != ACPI_MEMORY_RANGE))
		return AE_OK;

	if ((mcfg_res->start >= address.minimum) &&
402
	    (mcfg_res->end < (address.minimum + address.address_length))) {
403 404 405 406 407 408
		mcfg_res->flags = 1;
		return AE_CTRL_TERMINATE;
	}
	return AE_OK;
}

409 410
static acpi_status find_mboard_resource(acpi_handle handle, u32 lvl,
					void *context, void **rv)
411 412 413 414 415 416 417 418 419 420 421 422
{
	struct resource *mcfg_res = context;

	acpi_walk_resources(handle, METHOD_NAME__CRS,
			    check_mcfg_resource, context);

	if (mcfg_res->flags)
		return AE_CTRL_TERMINATE;

	return AE_OK;
}

423
static int is_acpi_reserved(u64 start, u64 end, unsigned not_used)
424 425 426 427
{
	struct resource mcfg_res;

	mcfg_res.start = start;
428
	mcfg_res.end = end - 1;
429 430 431 432 433 434 435 436 437 438 439
	mcfg_res.flags = 0;

	acpi_get_devices("PNP0C01", find_mboard_resource, &mcfg_res, NULL);

	if (!mcfg_res.flags)
		acpi_get_devices("PNP0C02", find_mboard_resource, &mcfg_res,
				 NULL);

	return mcfg_res.flags;
}

440 441
typedef int (*check_reserved_t)(u64 start, u64 end, unsigned type);

442 443 444
static int __ref is_mmconf_reserved(check_reserved_t is_reserved,
				    struct pci_mmcfg_region *cfg,
				    struct device *dev, int with_e820)
445
{
446 447
	u64 addr = cfg->res.start;
	u64 size = resource_size(&cfg->res);
448
	u64 old_size = size;
449 450
	int num_buses;
	char *method = with_e820 ? "E820" : "ACPI motherboard resources";
451

452
	while (!is_reserved(addr, addr + size, E820_RESERVED)) {
453 454 455 456 457
		size >>= 1;
		if (size < (16UL<<20))
			break;
	}

458 459 460 461 462 463 464
	if (size < (16UL<<20) && size != old_size)
		return 0;

	if (dev)
		dev_info(dev, "MMCONFIG at %pR reserved in %s\n",
			 &cfg->res, method);
	else
465
		pr_info(PREFIX "MMCONFIG at %pR reserved in %s\n",
466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483
		       &cfg->res, method);

	if (old_size != size) {
		/* update end_bus */
		cfg->end_bus = cfg->start_bus + ((size>>20) - 1);
		num_buses = cfg->end_bus - cfg->start_bus + 1;
		cfg->res.end = cfg->res.start +
		    PCI_MMCFG_BUS_OFFSET(num_buses) - 1;
		snprintf(cfg->name, PCI_MMCFG_RESOURCE_NAME_LEN,
			 "PCI MMCONFIG %04x [bus %02x-%02x]",
			 cfg->segment, cfg->start_bus, cfg->end_bus);

		if (dev)
			dev_info(dev,
				"MMCONFIG "
				"at %pR (base %#lx) (size reduced!)\n",
				&cfg->res, (unsigned long) cfg->address);
		else
484
			pr_info(PREFIX
485 486 487 488
				"MMCONFIG for %04x [bus%02x-%02x] "
				"at %pR (base %#lx) (size reduced!)\n",
				cfg->segment, cfg->start_bus, cfg->end_bus,
				&cfg->res, (unsigned long) cfg->address);
489 490
	}

491
	return 1;
492 493
}

494 495
static int __ref pci_mmcfg_check_reserved(struct device *dev,
		  struct pci_mmcfg_region *cfg, int early)
496 497
{
	if (!early && !acpi_disabled) {
498
		if (is_mmconf_reserved(is_acpi_reserved, cfg, dev, 0))
499
			return 1;
500 501 502 503 504 505

		if (dev)
			dev_info(dev, FW_INFO
				 "MMCONFIG at %pR not reserved in "
				 "ACPI motherboard resources\n",
				 &cfg->res);
506
		else
507
			pr_info(FW_INFO PREFIX
508 509 510 511 512
			       "MMCONFIG at %pR not reserved in "
			       "ACPI motherboard resources\n",
			       &cfg->res);
	}

513 514 515 516 517 518 519 520 521
	/*
	 * e820_all_mapped() is marked as __init.
	 * All entries from ACPI MCFG table have been checked at boot time.
	 * For MCFG information constructed from hotpluggable host bridge's
	 * _CBA method, just assume it's reserved.
	 */
	if (pci_mmcfg_running_state)
		return 1;

522 523 524
	/* Don't try to do this check unless configuration
	   type 1 is available. how about type 2 ?*/
	if (raw_pci_ops)
525
		return is_mmconf_reserved(e820_all_mapped, cfg, dev, 1);
526 527 528 529

	return 0;
}

530
static void __init pci_mmcfg_reject_broken(int early)
531
{
532
	struct pci_mmcfg_region *cfg;
533

534
	list_for_each_entry(cfg, &pci_mmcfg_list, list) {
535
		if (pci_mmcfg_check_reserved(NULL, cfg, early) == 0) {
536
			pr_info(PREFIX "not using MMCONFIG\n");
537 538
			free_all_mmcfg();
			return;
539
		}
540 541 542
	}
}

B
Bjorn Helgaas 已提交
543 544
static int __init acpi_mcfg_check_entry(struct acpi_table_mcfg *mcfg,
					struct acpi_mcfg_allocation *cfg)
545
{
B
Bjorn Helgaas 已提交
546 547 548 549 550
	int year;

	if (cfg->address < 0xFFFFFFFF)
		return 0;

551
	if (!strncmp(mcfg->header.oem_id, "SGI", 3))
B
Bjorn Helgaas 已提交
552
		return 0;
553

B
Bjorn Helgaas 已提交
554 555 556 557 558 559
	if (mcfg->header.revision >= 1) {
		if (dmi_get_date(DMI_BIOS_DATE, &year, NULL, NULL) &&
		    year >= 2010)
			return 0;
	}

560
	pr_err(PREFIX "MCFG region for %04x [bus %02x-%02x] at %#llx "
B
Bjorn Helgaas 已提交
561 562 563
	       "is above 4GB, ignored\n", cfg->pci_segment,
	       cfg->start_bus_number, cfg->end_bus_number, cfg->address);
	return -EINVAL;
564 565 566 567 568
}

static int __init pci_parse_mcfg(struct acpi_table_header *header)
{
	struct acpi_table_mcfg *mcfg;
569
	struct acpi_mcfg_allocation *cfg_table, *cfg;
570
	unsigned long i;
571
	int entries;
572 573 574 575 576 577 578

	if (!header)
		return -EINVAL;

	mcfg = (struct acpi_table_mcfg *)header;

	/* how many config structures do we have */
579
	free_all_mmcfg();
580
	entries = 0;
581 582
	i = header->length - sizeof(struct acpi_table_mcfg);
	while (i >= sizeof(struct acpi_mcfg_allocation)) {
583
		entries++;
584
		i -= sizeof(struct acpi_mcfg_allocation);
585
	}
586
	if (entries == 0) {
587
		pr_err(PREFIX "MMCONFIG has no entries\n");
588 589 590
		return -ENODEV;
	}

591
	cfg_table = (struct acpi_mcfg_allocation *) &mcfg[1];
592
	for (i = 0; i < entries; i++) {
593 594
		cfg = &cfg_table[i];
		if (acpi_mcfg_check_entry(mcfg, cfg)) {
595
			free_all_mmcfg();
596 597
			return -ENODEV;
		}
598 599 600

		if (pci_mmconfig_add(cfg->pci_segment, cfg->start_bus_number,
				   cfg->end_bus_number, cfg->address) == NULL) {
601
			pr_warn(PREFIX "no memory for MCFG entries\n");
602 603 604
			free_all_mmcfg();
			return -ENOMEM;
		}
605 606 607 608 609
	}

	return 0;
}

610
static void __init __pci_mmcfg_init(int early)
611
{
612
	pci_mmcfg_reject_broken(early);
613
	if (list_empty(&pci_mmcfg_list))
614 615
		return;

616 617 618 619 620 621 622 623 624 625
	if (pcibios_last_bus < 0) {
		const struct pci_mmcfg_region *cfg;

		list_for_each_entry(cfg, &pci_mmcfg_list, list) {
			if (cfg->segment)
				break;
			pcibios_last_bus = cfg->end_bus;
		}
	}

626
	if (pci_mmcfg_arch_init())
627
		pci_probe = (pci_probe & ~PCI_PROBE_MASK) | PCI_PROBE_MMCONF;
628
	else {
629
		free_all_mmcfg();
630
		pci_mmcfg_arch_init_failed = true;
631 632
	}
}
633

634 635
static int __initdata known_bridge;

636
void __init pci_mmcfg_early_init(void)
Y
Yinghai Lu 已提交
637
{
638 639 640 641 642 643 644
	if (pci_probe & PCI_PROBE_MMCONF) {
		if (pci_mmcfg_check_hostbridge())
			known_bridge = 1;
		else
			acpi_sfi_table_parse(ACPI_SIG_MCFG, pci_parse_mcfg);
		__pci_mmcfg_init(1);
	}
Y
Yinghai Lu 已提交
645 646 647 648
}

void __init pci_mmcfg_late_init(void)
{
649 650 651 652 653 654 655 656 657 658 659 660
	/* MMCONFIG disabled */
	if ((pci_probe & PCI_PROBE_MMCONF) == 0)
		return;

	if (known_bridge)
		return;

	/* MMCONFIG hasn't been enabled yet, try again */
	if (pci_probe & PCI_PROBE_MASK & ~PCI_PROBE_MMCONF) {
		acpi_sfi_table_parse(ACPI_SIG_MCFG, pci_parse_mcfg);
		__pci_mmcfg_init(0);
	}
Y
Yinghai Lu 已提交
661 662
}

663 664
static int __init pci_mmcfg_late_insert_resources(void)
{
665 666
	struct pci_mmcfg_region *cfg;

667 668
	pci_mmcfg_running_state = true;

669 670
	/* If we are not using MMCONFIG, don't insert the resources. */
	if ((pci_probe & PCI_PROBE_MMCONF) == 0)
671 672 673 674 675 676 677
		return 1;

	/*
	 * Attempt to insert the mmcfg resources but not with the busy flag
	 * marked so it won't cause request errors when __request_region is
	 * called.
	 */
678 679 680
	list_for_each_entry(cfg, &pci_mmcfg_list, list)
		if (!cfg->res.parent)
			insert_resource(&iomem_resource, &cfg->res);
681 682 683 684 685 686 687 688 689 690

	return 0;
}

/*
 * Perform MMCONFIG resource insertion after PCI initialization to allow for
 * misprogrammed MCFG tables that state larger sizes but actually conflict
 * with other system resources.
 */
late_initcall(pci_mmcfg_late_insert_resources);
691 692

/* Add MMCFG information for host bridges */
693 694
int pci_mmconfig_insert(struct device *dev, u16 seg, u8 start, u8 end,
			phys_addr_t addr)
695 696 697 698 699 700 701 702
{
	int rc;
	struct resource *tmp = NULL;
	struct pci_mmcfg_region *cfg;

	if (!(pci_probe & PCI_PROBE_MMCONF) || pci_mmcfg_arch_init_failed)
		return -ENODEV;

703
	if (start > end)
704 705 706 707 708 709 710 711 712 713 714 715 716 717 718
		return -EINVAL;

	mutex_lock(&pci_mmcfg_lock);
	cfg = pci_mmconfig_lookup(seg, start);
	if (cfg) {
		if (cfg->end_bus < end)
			dev_info(dev, FW_INFO
				 "MMCONFIG for "
				 "domain %04x [bus %02x-%02x] "
				 "only partially covers this bridge\n",
				  cfg->segment, cfg->start_bus, cfg->end_bus);
		mutex_unlock(&pci_mmcfg_lock);
		return -EEXIST;
	}

719 720 721 722 723
	if (!addr) {
		mutex_unlock(&pci_mmcfg_lock);
		return -EINVAL;
	}

724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787
	rc = -EBUSY;
	cfg = pci_mmconfig_alloc(seg, start, end, addr);
	if (cfg == NULL) {
		dev_warn(dev, "fail to add MMCONFIG (out of memory)\n");
		rc = -ENOMEM;
	} else if (!pci_mmcfg_check_reserved(dev, cfg, 0)) {
		dev_warn(dev, FW_BUG "MMCONFIG %pR isn't reserved\n",
			 &cfg->res);
	} else {
		/* Insert resource if it's not in boot stage */
		if (pci_mmcfg_running_state)
			tmp = insert_resource_conflict(&iomem_resource,
						       &cfg->res);

		if (tmp) {
			dev_warn(dev,
				 "MMCONFIG %pR conflicts with "
				 "%s %pR\n",
				 &cfg->res, tmp->name, tmp);
		} else if (pci_mmcfg_arch_map(cfg)) {
			dev_warn(dev, "fail to map MMCONFIG %pR.\n",
				 &cfg->res);
		} else {
			list_add_sorted(cfg);
			dev_info(dev, "MMCONFIG at %pR (base %#lx)\n",
				 &cfg->res, (unsigned long)addr);
			cfg = NULL;
			rc = 0;
		}
	}

	if (cfg) {
		if (cfg->res.parent)
			release_resource(&cfg->res);
		kfree(cfg);
	}

	mutex_unlock(&pci_mmcfg_lock);

	return rc;
}

/* Delete MMCFG information for host bridges */
int pci_mmconfig_delete(u16 seg, u8 start, u8 end)
{
	struct pci_mmcfg_region *cfg;

	mutex_lock(&pci_mmcfg_lock);
	list_for_each_entry_rcu(cfg, &pci_mmcfg_list, list)
		if (cfg->segment == seg && cfg->start_bus == start &&
		    cfg->end_bus == end) {
			list_del_rcu(&cfg->list);
			synchronize_rcu();
			pci_mmcfg_arch_unmap(cfg);
			if (cfg->res.parent)
				release_resource(&cfg->res);
			mutex_unlock(&pci_mmcfg_lock);
			kfree(cfg);
			return 0;
		}
	mutex_unlock(&pci_mmcfg_lock);

	return -ENOENT;
}