probe.c 64.0 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8
/*
 * probe.c - PCI detection and setup code
 */

#include <linux/kernel.h>
#include <linux/delay.h>
#include <linux/init.h>
#include <linux/pci.h>
9
#include <linux/of_device.h>
10
#include <linux/of_pci.h>
11
#include <linux/pci_hotplug.h>
L
Linus Torvalds 已提交
12 13 14
#include <linux/slab.h>
#include <linux/module.h>
#include <linux/cpumask.h>
S
Shaohua Li 已提交
15
#include <linux/pci-aspm.h>
16
#include <linux/aer.h>
17
#include <linux/acpi.h>
18
#include <linux/irqdomain.h>
19
#include <linux/pm_runtime.h>
20
#include "pci.h"
L
Linus Torvalds 已提交
21 22 23 24

#define CARDBUS_LATENCY_TIMER	176	/* secondary latency timer */
#define CARDBUS_RESERVE_BUSNR	3

25
static struct resource busn_resource = {
Y
Yinghai Lu 已提交
26 27 28 29 30 31
	.name	= "PCI busn",
	.start	= 0,
	.end	= 255,
	.flags	= IORESOURCE_BUS,
};

L
Linus Torvalds 已提交
32 33 34 35
/* Ugh.  Need to stop exporting this to modules. */
LIST_HEAD(pci_root_buses);
EXPORT_SYMBOL(pci_root_buses);

36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
static LIST_HEAD(pci_domain_busn_res_list);

struct pci_domain_busn_res {
	struct list_head list;
	struct resource res;
	int domain_nr;
};

static struct resource *get_pci_domain_busn_res(int domain_nr)
{
	struct pci_domain_busn_res *r;

	list_for_each_entry(r, &pci_domain_busn_res_list, list)
		if (r->domain_nr == domain_nr)
			return &r->res;

	r = kzalloc(sizeof(*r), GFP_KERNEL);
	if (!r)
		return NULL;

	r->domain_nr = domain_nr;
	r->res.start = 0;
	r->res.end = 0xff;
	r->res.flags = IORESOURCE_BUS | IORESOURCE_PCI_FIXED;

	list_add_tail(&r->list, &pci_domain_busn_res_list);

	return &r->res;
}

66 67 68 69
static int find_anything(struct device *dev, void *data)
{
	return 1;
}
L
Linus Torvalds 已提交
70

Z
Zhang, Yanmin 已提交
71 72 73
/*
 * Some device drivers need know if pci is initiated.
 * Basically, we think pci is not initiated when there
74
 * is no device to be found on the pci_bus_type.
Z
Zhang, Yanmin 已提交
75 76 77
 */
int no_pci_devices(void)
{
78 79
	struct device *dev;
	int no_devices;
Z
Zhang, Yanmin 已提交
80

81 82 83 84 85
	dev = bus_find_device(&pci_bus_type, NULL, NULL, find_anything);
	no_devices = (dev == NULL);
	put_device(dev);
	return no_devices;
}
Z
Zhang, Yanmin 已提交
86 87
EXPORT_SYMBOL(no_pci_devices);

L
Linus Torvalds 已提交
88 89 90
/*
 * PCI Bus Class
 */
91
static void release_pcibus_dev(struct device *dev)
L
Linus Torvalds 已提交
92
{
93
	struct pci_bus *pci_bus = to_pci_bus(dev);
L
Linus Torvalds 已提交
94

95
	put_device(pci_bus->bridge);
96
	pci_bus_remove_resources(pci_bus);
97
	pci_release_bus_of_node(pci_bus);
L
Linus Torvalds 已提交
98 99 100 101 102
	kfree(pci_bus);
}

static struct class pcibus_class = {
	.name		= "pci_bus",
103
	.dev_release	= &release_pcibus_dev,
104
	.dev_groups	= pcibus_groups,
L
Linus Torvalds 已提交
105 106 107 108 109 110 111 112
};

static int __init pcibus_class_init(void)
{
	return class_register(&pcibus_class);
}
postcore_initcall(pcibus_class_init);

113
static u64 pci_size(u64 base, u64 maxbase, u64 mask)
L
Linus Torvalds 已提交
114
{
115
	u64 size = mask & maxbase;	/* Find the significant bits */
L
Linus Torvalds 已提交
116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
	if (!size)
		return 0;

	/* Get the lowest of them to find the decode size, and
	   from that the extent.  */
	size = (size & ~(size-1)) - 1;

	/* base == maxbase can be valid only if the BAR has
	   already been programmed with all 1s.  */
	if (base == maxbase && ((base | size) & mask) != mask)
		return 0;

	return size;
}

131
static inline unsigned long decode_bar(struct pci_dev *dev, u32 bar)
132
{
133
	u32 mem_type;
134
	unsigned long flags;
135

136
	if ((bar & PCI_BASE_ADDRESS_SPACE) == PCI_BASE_ADDRESS_SPACE_IO) {
137 138 139
		flags = bar & ~PCI_BASE_ADDRESS_IO_MASK;
		flags |= IORESOURCE_IO;
		return flags;
140
	}
141

142 143 144 145
	flags = bar & ~PCI_BASE_ADDRESS_MEM_MASK;
	flags |= IORESOURCE_MEM;
	if (flags & PCI_BASE_ADDRESS_MEM_PREFETCH)
		flags |= IORESOURCE_PREFETCH;
146

147 148 149 150 151
	mem_type = bar & PCI_BASE_ADDRESS_MEM_TYPE_MASK;
	switch (mem_type) {
	case PCI_BASE_ADDRESS_MEM_TYPE_32:
		break;
	case PCI_BASE_ADDRESS_MEM_TYPE_1M:
152
		/* 1M mem BAR treated as 32-bit BAR */
153 154
		break;
	case PCI_BASE_ADDRESS_MEM_TYPE_64:
155 156
		flags |= IORESOURCE_MEM_64;
		break;
157
	default:
158
		/* mem unknown type treated as 32-bit BAR */
159 160
		break;
	}
161
	return flags;
162 163
}

164 165
#define PCI_COMMAND_DECODE_ENABLE	(PCI_COMMAND_MEMORY | PCI_COMMAND_IO)

Y
Yu Zhao 已提交
166 167 168 169 170 171 172 173
/**
 * pci_read_base - read a PCI BAR
 * @dev: the PCI device
 * @type: type of the BAR
 * @res: resource buffer to be filled in
 * @pos: BAR position in the config space
 *
 * Returns 1 if the BAR is 64-bit, or 0 if 32-bit.
174
 */
Y
Yu Zhao 已提交
175
int __pci_read_base(struct pci_dev *dev, enum pci_bar_type type,
R
Ryan Desfosses 已提交
176
		    struct resource *res, unsigned int pos)
177
{
178
	u32 l, sz, mask;
179
	u64 l64, sz64, mask64;
J
Jacob Pan 已提交
180
	u16 orig_cmd;
181
	struct pci_bus_region region, inverted_region;
182

183
	mask = type ? PCI_ROM_ADDRESS_MASK : ~0;
184

185
	/* No printks while decoding is disabled! */
J
Jacob Pan 已提交
186 187
	if (!dev->mmio_always_on) {
		pci_read_config_word(dev, PCI_COMMAND, &orig_cmd);
188 189 190 191
		if (orig_cmd & PCI_COMMAND_DECODE_ENABLE) {
			pci_write_config_word(dev, PCI_COMMAND,
				orig_cmd & ~PCI_COMMAND_DECODE_ENABLE);
		}
J
Jacob Pan 已提交
192 193
	}

194 195 196
	res->name = pci_name(dev);

	pci_read_config_dword(dev, pos, &l);
197
	pci_write_config_dword(dev, pos, l | mask);
198 199 200 201 202
	pci_read_config_dword(dev, pos, &sz);
	pci_write_config_dword(dev, pos, l);

	/*
	 * All bits set in sz means the device isn't working properly.
203 204 205
	 * If the BAR isn't implemented, all bits must be 0.  If it's a
	 * memory BAR or a ROM, bit 0 must be clear; if it's an io BAR, bit
	 * 1 must be clear.
206
	 */
207 208
	if (sz == 0xffffffff)
		sz = 0;
209 210 211 212 213 214 215 216 217

	/*
	 * I don't know how l can have all bits set.  Copied from old code.
	 * Maybe it fixes a bug on some ancient platform.
	 */
	if (l == 0xffffffff)
		l = 0;

	if (type == pci_bar_unknown) {
218 219 220
		res->flags = decode_bar(dev, l);
		res->flags |= IORESOURCE_SIZEALIGN;
		if (res->flags & IORESOURCE_IO) {
221 222 223
			l64 = l & PCI_BASE_ADDRESS_IO_MASK;
			sz64 = sz & PCI_BASE_ADDRESS_IO_MASK;
			mask64 = PCI_BASE_ADDRESS_IO_MASK & (u32)IO_SPACE_LIMIT;
224
		} else {
225 226 227
			l64 = l & PCI_BASE_ADDRESS_MEM_MASK;
			sz64 = sz & PCI_BASE_ADDRESS_MEM_MASK;
			mask64 = (u32)PCI_BASE_ADDRESS_MEM_MASK;
228 229 230
		}
	} else {
		res->flags |= (l & IORESOURCE_ROM_ENABLE);
231 232 233
		l64 = l & PCI_ROM_ADDRESS_MASK;
		sz64 = sz & PCI_ROM_ADDRESS_MASK;
		mask64 = (u32)PCI_ROM_ADDRESS_MASK;
234 235
	}

236
	if (res->flags & IORESOURCE_MEM_64) {
237 238 239 240 241 242 243
		pci_read_config_dword(dev, pos + 4, &l);
		pci_write_config_dword(dev, pos + 4, ~0);
		pci_read_config_dword(dev, pos + 4, &sz);
		pci_write_config_dword(dev, pos + 4, l);

		l64 |= ((u64)l << 32);
		sz64 |= ((u64)sz << 32);
244 245
		mask64 |= ((u64)~0 << 32);
	}
246

247 248
	if (!dev->mmio_always_on && (orig_cmd & PCI_COMMAND_DECODE_ENABLE))
		pci_write_config_word(dev, PCI_COMMAND, orig_cmd);
249

250 251
	if (!sz64)
		goto fail;
252

253
	sz64 = pci_size(l64, sz64, mask64);
254 255 256
	if (!sz64) {
		dev_info(&dev->dev, FW_BUG "reg 0x%x: invalid BAR (can't size)\n",
			 pos);
257
		goto fail;
258
	}
259 260

	if (res->flags & IORESOURCE_MEM_64) {
Y
Yinghai Lu 已提交
261 262
		if ((sizeof(pci_bus_addr_t) < 8 || sizeof(resource_size_t) < 8)
		    && sz64 > 0x100000000ULL) {
263 264 265
			res->flags |= IORESOURCE_UNSET | IORESOURCE_DISABLED;
			res->start = 0;
			res->end = 0;
266 267
			dev_err(&dev->dev, "reg 0x%x: can't handle BAR larger than 4GB (size %#010llx)\n",
				pos, (unsigned long long)sz64);
268
			goto out;
269 270
		}

Y
Yinghai Lu 已提交
271
		if ((sizeof(pci_bus_addr_t) < 8) && l) {
272
			/* Above 32-bit boundary; try to reallocate */
273
			res->flags |= IORESOURCE_UNSET;
274 275
			res->start = 0;
			res->end = sz64;
276 277
			dev_info(&dev->dev, "reg 0x%x: can't handle BAR above 4GB (bus address %#010llx)\n",
				 pos, (unsigned long long)l64);
278
			goto out;
279 280 281
		}
	}

282 283 284
	region.start = l64;
	region.end = l64 + sz64;

285 286
	pcibios_bus_to_resource(dev->bus, res, &region);
	pcibios_resource_to_bus(dev->bus, &inverted_region, res);
287 288 289 290 291 292 293 294 295 296 297 298 299 300 301

	/*
	 * If "A" is a BAR value (a bus address), "bus_to_resource(A)" is
	 * the corresponding resource address (the physical address used by
	 * the CPU.  Converting that resource address back to a bus address
	 * should yield the original BAR value:
	 *
	 *     resource_to_bus(bus_to_resource(A)) == A
	 *
	 * If it doesn't, CPU accesses to "bus_to_resource(A)" will not
	 * be claimed by the device.
	 */
	if (inverted_region.start != region.start) {
		res->flags |= IORESOURCE_UNSET;
		res->start = 0;
302
		res->end = region.end - region.start;
303 304
		dev_info(&dev->dev, "reg 0x%x: initial BAR value %#010llx invalid\n",
			 pos, (unsigned long long)region.start);
305
	}
306

307 308 309 310 311 312
	goto out;


fail:
	res->flags = 0;
out:
313
	if (res->flags)
314
		dev_printk(KERN_DEBUG, &dev->dev, "reg 0x%x: %pR\n", pos, res);
315

316
	return (res->flags & IORESOURCE_MEM_64) ? 1 : 0;
317 318
}

L
Linus Torvalds 已提交
319 320
static void pci_read_bases(struct pci_dev *dev, unsigned int howmany, int rom)
{
321
	unsigned int pos, reg;
322

323 324 325
	if (dev->non_compliant_bars)
		return;

326 327
	for (pos = 0; pos < howmany; pos++) {
		struct resource *res = &dev->resource[pos];
L
Linus Torvalds 已提交
328
		reg = PCI_BASE_ADDRESS_0 + (pos << 2);
329
		pos += __pci_read_base(dev, pci_bar_unknown, res, reg);
L
Linus Torvalds 已提交
330
	}
331

L
Linus Torvalds 已提交
332
	if (rom) {
333
		struct resource *res = &dev->resource[PCI_ROM_RESOURCE];
L
Linus Torvalds 已提交
334
		dev->rom_base_reg = rom;
335
		res->flags = IORESOURCE_MEM | IORESOURCE_PREFETCH |
336
				IORESOURCE_READONLY | IORESOURCE_SIZEALIGN;
337
		__pci_read_base(dev, pci_bar_mem32, res, rom);
L
Linus Torvalds 已提交
338 339 340
	}
}

B
Bill Pemberton 已提交
341
static void pci_read_bridge_io(struct pci_bus *child)
L
Linus Torvalds 已提交
342 343 344
{
	struct pci_dev *dev = child->self;
	u8 io_base_lo, io_limit_lo;
345
	unsigned long io_mask, io_granularity, base, limit;
346
	struct pci_bus_region region;
347 348 349 350 351 352 353 354 355
	struct resource *res;

	io_mask = PCI_IO_RANGE_MASK;
	io_granularity = 0x1000;
	if (dev->io_window_1k) {
		/* Support 1K I/O space granularity */
		io_mask = PCI_IO_1K_RANGE_MASK;
		io_granularity = 0x400;
	}
L
Linus Torvalds 已提交
356 357 358 359

	res = child->resource[0];
	pci_read_config_byte(dev, PCI_IO_BASE, &io_base_lo);
	pci_read_config_byte(dev, PCI_IO_LIMIT, &io_limit_lo);
360 361
	base = (io_base_lo & io_mask) << 8;
	limit = (io_limit_lo & io_mask) << 8;
L
Linus Torvalds 已提交
362 363 364

	if ((io_base_lo & PCI_IO_RANGE_TYPE_MASK) == PCI_IO_RANGE_TYPE_32) {
		u16 io_base_hi, io_limit_hi;
365

L
Linus Torvalds 已提交
366 367
		pci_read_config_word(dev, PCI_IO_BASE_UPPER16, &io_base_hi);
		pci_read_config_word(dev, PCI_IO_LIMIT_UPPER16, &io_limit_hi);
368 369
		base |= ((unsigned long) io_base_hi << 16);
		limit |= ((unsigned long) io_limit_hi << 16);
L
Linus Torvalds 已提交
370 371
	}

372
	if (base <= limit) {
L
Linus Torvalds 已提交
373
		res->flags = (io_base_lo & PCI_IO_RANGE_TYPE_MASK) | IORESOURCE_IO;
374
		region.start = base;
375
		region.end = limit + io_granularity - 1;
376
		pcibios_bus_to_resource(dev->bus, res, &region);
377
		dev_printk(KERN_DEBUG, &dev->dev, "  bridge window %pR\n", res);
L
Linus Torvalds 已提交
378
	}
379 380
}

B
Bill Pemberton 已提交
381
static void pci_read_bridge_mmio(struct pci_bus *child)
382 383 384 385
{
	struct pci_dev *dev = child->self;
	u16 mem_base_lo, mem_limit_lo;
	unsigned long base, limit;
386
	struct pci_bus_region region;
387
	struct resource *res;
L
Linus Torvalds 已提交
388 389 390 391

	res = child->resource[1];
	pci_read_config_word(dev, PCI_MEMORY_BASE, &mem_base_lo);
	pci_read_config_word(dev, PCI_MEMORY_LIMIT, &mem_limit_lo);
392 393
	base = ((unsigned long) mem_base_lo & PCI_MEMORY_RANGE_MASK) << 16;
	limit = ((unsigned long) mem_limit_lo & PCI_MEMORY_RANGE_MASK) << 16;
394
	if (base <= limit) {
L
Linus Torvalds 已提交
395
		res->flags = (mem_base_lo & PCI_MEMORY_RANGE_TYPE_MASK) | IORESOURCE_MEM;
396 397
		region.start = base;
		region.end = limit + 0xfffff;
398
		pcibios_bus_to_resource(dev->bus, res, &region);
399
		dev_printk(KERN_DEBUG, &dev->dev, "  bridge window %pR\n", res);
L
Linus Torvalds 已提交
400
	}
401 402
}

B
Bill Pemberton 已提交
403
static void pci_read_bridge_mmio_pref(struct pci_bus *child)
404 405 406
{
	struct pci_dev *dev = child->self;
	u16 mem_base_lo, mem_limit_lo;
407
	u64 base64, limit64;
Y
Yinghai Lu 已提交
408
	pci_bus_addr_t base, limit;
409
	struct pci_bus_region region;
410
	struct resource *res;
L
Linus Torvalds 已提交
411 412 413 414

	res = child->resource[2];
	pci_read_config_word(dev, PCI_PREF_MEMORY_BASE, &mem_base_lo);
	pci_read_config_word(dev, PCI_PREF_MEMORY_LIMIT, &mem_limit_lo);
415 416
	base64 = (mem_base_lo & PCI_PREF_RANGE_MASK) << 16;
	limit64 = (mem_limit_lo & PCI_PREF_RANGE_MASK) << 16;
L
Linus Torvalds 已提交
417 418 419

	if ((mem_base_lo & PCI_PREF_RANGE_TYPE_MASK) == PCI_PREF_RANGE_TYPE_64) {
		u32 mem_base_hi, mem_limit_hi;
420

L
Linus Torvalds 已提交
421 422 423 424 425 426 427 428 429
		pci_read_config_dword(dev, PCI_PREF_BASE_UPPER32, &mem_base_hi);
		pci_read_config_dword(dev, PCI_PREF_LIMIT_UPPER32, &mem_limit_hi);

		/*
		 * Some bridges set the base > limit by default, and some
		 * (broken) BIOSes do not initialize them.  If we find
		 * this, just assume they are not being used.
		 */
		if (mem_base_hi <= mem_limit_hi) {
430 431
			base64 |= (u64) mem_base_hi << 32;
			limit64 |= (u64) mem_limit_hi << 32;
L
Linus Torvalds 已提交
432 433
		}
	}
434

Y
Yinghai Lu 已提交
435 436
	base = (pci_bus_addr_t) base64;
	limit = (pci_bus_addr_t) limit64;
437 438 439 440 441 442 443

	if (base != base64) {
		dev_err(&dev->dev, "can't handle bridge window above 4GB (bus address %#010llx)\n",
			(unsigned long long) base64);
		return;
	}

444
	if (base <= limit) {
445 446 447 448
		res->flags = (mem_base_lo & PCI_PREF_RANGE_TYPE_MASK) |
					 IORESOURCE_MEM | IORESOURCE_PREFETCH;
		if (res->flags & PCI_PREF_RANGE_TYPE_64)
			res->flags |= IORESOURCE_MEM_64;
449 450
		region.start = base;
		region.end = limit + 0xfffff;
451
		pcibios_bus_to_resource(dev->bus, res, &region);
452
		dev_printk(KERN_DEBUG, &dev->dev, "  bridge window %pR\n", res);
L
Linus Torvalds 已提交
453 454 455
	}
}

B
Bill Pemberton 已提交
456
void pci_read_bridge_bases(struct pci_bus *child)
457 458
{
	struct pci_dev *dev = child->self;
459
	struct resource *res;
460 461 462 463 464
	int i;

	if (pci_is_root_bus(child))	/* It's a host bus, nothing to read */
		return;

465 466
	dev_info(&dev->dev, "PCI bridge to %pR%s\n",
		 &child->busn_res,
467 468
		 dev->transparent ? " (subtractive decode)" : "");

469 470 471 472
	pci_bus_remove_resources(child);
	for (i = 0; i < PCI_BRIDGE_RESOURCE_NUM; i++)
		child->resource[i] = &dev->resource[PCI_BRIDGE_RESOURCES+i];

473 474 475
	pci_read_bridge_io(child);
	pci_read_bridge_mmio(child);
	pci_read_bridge_mmio_pref(child);
476 477

	if (dev->transparent) {
478
		pci_bus_for_each_resource(child->parent, res, i) {
479
			if (res && res->flags) {
480 481
				pci_bus_add_resource(child, res,
						     PCI_SUBTRACTIVE_DECODE);
482 483
				dev_printk(KERN_DEBUG, &dev->dev,
					   "  bridge window %pR (subtractive decode)\n",
484 485
					   res);
			}
486 487
		}
	}
488 489
}

490
static struct pci_bus *pci_alloc_bus(struct pci_bus *parent)
L
Linus Torvalds 已提交
491 492 493
{
	struct pci_bus *b;

494
	b = kzalloc(sizeof(*b), GFP_KERNEL);
495 496 497 498 499 500 501 502 503 504
	if (!b)
		return NULL;

	INIT_LIST_HEAD(&b->node);
	INIT_LIST_HEAD(&b->children);
	INIT_LIST_HEAD(&b->devices);
	INIT_LIST_HEAD(&b->slots);
	INIT_LIST_HEAD(&b->resources);
	b->max_bus_speed = PCI_SPEED_UNKNOWN;
	b->cur_bus_speed = PCI_SPEED_UNKNOWN;
505 506 507 508
#ifdef CONFIG_PCI_DOMAINS_GENERIC
	if (parent)
		b->domain_nr = parent->domain_nr;
#endif
L
Linus Torvalds 已提交
509 510 511
	return b;
}

512 513 514 515 516 517 518 519 520 521 522 523
static void pci_release_host_bridge_dev(struct device *dev)
{
	struct pci_host_bridge *bridge = to_pci_host_bridge(dev);

	if (bridge->release_fn)
		bridge->release_fn(bridge);

	pci_free_resource_list(&bridge->windows);

	kfree(bridge);
}

524
static struct pci_host_bridge *pci_alloc_host_bridge(void)
525 526 527 528
{
	struct pci_host_bridge *bridge;

	bridge = kzalloc(sizeof(*bridge), GFP_KERNEL);
529 530
	if (!bridge)
		return NULL;
531

532
	INIT_LIST_HEAD(&bridge->windows);
533

534 535 536
	return bridge;
}

537
static const unsigned char pcix_bus_speed[] = {
538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555
	PCI_SPEED_UNKNOWN,		/* 0 */
	PCI_SPEED_66MHz_PCIX,		/* 1 */
	PCI_SPEED_100MHz_PCIX,		/* 2 */
	PCI_SPEED_133MHz_PCIX,		/* 3 */
	PCI_SPEED_UNKNOWN,		/* 4 */
	PCI_SPEED_66MHz_PCIX_ECC,	/* 5 */
	PCI_SPEED_100MHz_PCIX_ECC,	/* 6 */
	PCI_SPEED_133MHz_PCIX_ECC,	/* 7 */
	PCI_SPEED_UNKNOWN,		/* 8 */
	PCI_SPEED_66MHz_PCIX_266,	/* 9 */
	PCI_SPEED_100MHz_PCIX_266,	/* A */
	PCI_SPEED_133MHz_PCIX_266,	/* B */
	PCI_SPEED_UNKNOWN,		/* C */
	PCI_SPEED_66MHz_PCIX_533,	/* D */
	PCI_SPEED_100MHz_PCIX_533,	/* E */
	PCI_SPEED_133MHz_PCIX_533	/* F */
};

556
const unsigned char pcie_link_speed[] = {
557 558 559
	PCI_SPEED_UNKNOWN,		/* 0 */
	PCIE_SPEED_2_5GT,		/* 1 */
	PCIE_SPEED_5_0GT,		/* 2 */
560
	PCIE_SPEED_8_0GT,		/* 3 */
561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576
	PCI_SPEED_UNKNOWN,		/* 4 */
	PCI_SPEED_UNKNOWN,		/* 5 */
	PCI_SPEED_UNKNOWN,		/* 6 */
	PCI_SPEED_UNKNOWN,		/* 7 */
	PCI_SPEED_UNKNOWN,		/* 8 */
	PCI_SPEED_UNKNOWN,		/* 9 */
	PCI_SPEED_UNKNOWN,		/* A */
	PCI_SPEED_UNKNOWN,		/* B */
	PCI_SPEED_UNKNOWN,		/* C */
	PCI_SPEED_UNKNOWN,		/* D */
	PCI_SPEED_UNKNOWN,		/* E */
	PCI_SPEED_UNKNOWN		/* F */
};

void pcie_update_link_speed(struct pci_bus *bus, u16 linksta)
{
577
	bus->cur_bus_speed = pcie_link_speed[linksta & PCI_EXP_LNKSTA_CLS];
578 579 580
}
EXPORT_SYMBOL_GPL(pcie_update_link_speed);

581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600
static unsigned char agp_speeds[] = {
	AGP_UNKNOWN,
	AGP_1X,
	AGP_2X,
	AGP_4X,
	AGP_8X
};

static enum pci_bus_speed agp_speed(int agp3, int agpstat)
{
	int index = 0;

	if (agpstat & 4)
		index = 3;
	else if (agpstat & 2)
		index = 2;
	else if (agpstat & 1)
		index = 1;
	else
		goto out;
601

602 603 604 605 606 607 608 609 610 611
	if (agp3) {
		index += 2;
		if (index == 5)
			index = 0;
	}

 out:
	return agp_speeds[index];
}

612 613 614 615 616
static void pci_set_bus_speed(struct pci_bus *bus)
{
	struct pci_dev *bridge = bus->self;
	int pos;

617 618 619 620 621 622 623 624 625 626 627 628 629
	pos = pci_find_capability(bridge, PCI_CAP_ID_AGP);
	if (!pos)
		pos = pci_find_capability(bridge, PCI_CAP_ID_AGP3);
	if (pos) {
		u32 agpstat, agpcmd;

		pci_read_config_dword(bridge, pos + PCI_AGP_STATUS, &agpstat);
		bus->max_bus_speed = agp_speed(agpstat & 8, agpstat & 7);

		pci_read_config_dword(bridge, pos + PCI_AGP_COMMAND, &agpcmd);
		bus->cur_bus_speed = agp_speed(agpstat & 8, agpcmd & 7);
	}

630 631 632 633 634
	pos = pci_find_capability(bridge, PCI_CAP_ID_PCIX);
	if (pos) {
		u16 status;
		enum pci_bus_speed max;

635 636 637 638
		pci_read_config_word(bridge, pos + PCI_X_BRIDGE_SSTATUS,
				     &status);

		if (status & PCI_X_SSTATUS_533MHZ) {
639
			max = PCI_SPEED_133MHz_PCIX_533;
640
		} else if (status & PCI_X_SSTATUS_266MHZ) {
641
			max = PCI_SPEED_133MHz_PCIX_266;
642
		} else if (status & PCI_X_SSTATUS_133MHZ) {
R
Ryan Desfosses 已提交
643
			if ((status & PCI_X_SSTATUS_VERS) == PCI_X_SSTATUS_V2)
644
				max = PCI_SPEED_133MHz_PCIX_ECC;
R
Ryan Desfosses 已提交
645
			else
646 647 648 649 650 651
				max = PCI_SPEED_133MHz_PCIX;
		} else {
			max = PCI_SPEED_66MHz_PCIX;
		}

		bus->max_bus_speed = max;
652 653
		bus->cur_bus_speed = pcix_bus_speed[
			(status & PCI_X_SSTATUS_FREQ) >> 6];
654 655 656 657

		return;
	}

658
	if (pci_is_pcie(bridge)) {
659 660 661
		u32 linkcap;
		u16 linksta;

662
		pcie_capability_read_dword(bridge, PCI_EXP_LNKCAP, &linkcap);
663
		bus->max_bus_speed = pcie_link_speed[linkcap & PCI_EXP_LNKCAP_SLS];
664

665
		pcie_capability_read_word(bridge, PCI_EXP_LNKSTA, &linksta);
666 667 668 669
		pcie_update_link_speed(bus, linksta);
	}
}

670 671
static struct irq_domain *pci_host_bridge_msi_domain(struct pci_bus *bus)
{
672 673
	struct irq_domain *d;

674 675 676 677
	/*
	 * Any firmware interface that can resolve the msi_domain
	 * should be called from here.
	 */
678
	d = pci_host_bridge_of_msi_domain(bus);
679 680
	if (!d)
		d = pci_host_bridge_acpi_msi_domain(bus);
681

682 683 684 685 686 687 688 689 690 691 692 693 694 695
#ifdef CONFIG_PCI_MSI_IRQ_DOMAIN
	/*
	 * If no IRQ domain was found via the OF tree, try looking it up
	 * directly through the fwnode_handle.
	 */
	if (!d) {
		struct fwnode_handle *fwnode = pci_root_bus_fwnode(bus);

		if (fwnode)
			d = irq_find_matching_fwnode(fwnode,
						     DOMAIN_BUS_PCI_MSI);
	}
#endif

696
	return d;
697 698 699 700 701
}

static void pci_set_bus_msi_domain(struct pci_bus *bus)
{
	struct irq_domain *d;
702
	struct pci_bus *b;
703 704

	/*
705 706 707
	 * The bus can be a root bus, a subordinate bus, or a virtual bus
	 * created by an SR-IOV device.  Walk up to the first bridge device
	 * found or derive the domain from the host bridge.
708
	 */
709 710 711 712 713 714 715
	for (b = bus, d = NULL; !d && !pci_is_root_bus(b); b = b->parent) {
		if (b->self)
			d = dev_get_msi_domain(&b->self->dev);
	}

	if (!d)
		d = pci_host_bridge_msi_domain(b);
716 717 718 719

	dev_set_msi_domain(&bus->dev, d);
}

720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835
static int pci_register_host_bridge(struct pci_host_bridge *bridge)
{
	struct device *parent = bridge->dev.parent;
	struct resource_entry *window, *n;
	struct pci_bus *bus, *b;
	resource_size_t offset;
	LIST_HEAD(resources);
	struct resource *res;
	char addr[64], *fmt;
	const char *name;
	int err;

	bus = pci_alloc_bus(NULL);
	if (!bus)
		return -ENOMEM;

	bridge->bus = bus;

	/* temporarily move resources off the list */
	list_splice_init(&bridge->windows, &resources);
	bus->sysdata = bridge->sysdata;
	bus->msi = bridge->msi;
	bus->ops = bridge->ops;
	bus->number = bus->busn_res.start = bridge->busnr;
#ifdef CONFIG_PCI_DOMAINS_GENERIC
	bus->domain_nr = pci_bus_find_domain_nr(bus, parent);
#endif

	b = pci_find_bus(pci_domain_nr(bus), bridge->busnr);
	if (b) {
		/* If we already got to this bus through a different bridge, ignore it */
		dev_dbg(&b->dev, "bus already known\n");
		err = -EEXIST;
		goto free;
	}

	dev_set_name(&bridge->dev, "pci%04x:%02x", pci_domain_nr(bus),
		     bridge->busnr);

	err = pcibios_root_bridge_prepare(bridge);
	if (err)
		goto free;

	err = device_register(&bridge->dev);
	if (err)
		put_device(&bridge->dev);

	bus->bridge = get_device(&bridge->dev);
	device_enable_async_suspend(bus->bridge);
	pci_set_bus_of_node(bus);
	pci_set_bus_msi_domain(bus);

	if (!parent)
		set_dev_node(bus->bridge, pcibus_to_node(bus));

	bus->dev.class = &pcibus_class;
	bus->dev.parent = bus->bridge;

	dev_set_name(&bus->dev, "%04x:%02x", pci_domain_nr(bus), bus->number);
	name = dev_name(&bus->dev);

	err = device_register(&bus->dev);
	if (err)
		goto unregister;

	pcibios_add_bus(bus);

	/* Create legacy_io and legacy_mem files for this bus */
	pci_create_legacy_files(bus);

	if (parent)
		dev_info(parent, "PCI host bridge to bus %s\n", name);
	else
		pr_info("PCI host bridge to bus %s\n", name);

	/* Add initial resources to the bus */
	resource_list_for_each_entry_safe(window, n, &resources) {
		list_move_tail(&window->node, &bridge->windows);
		offset = window->offset;
		res = window->res;

		if (res->flags & IORESOURCE_BUS)
			pci_bus_insert_busn_res(bus, bus->number, res->end);
		else
			pci_bus_add_resource(bus, res, 0);

		if (offset) {
			if (resource_type(res) == IORESOURCE_IO)
				fmt = " (bus address [%#06llx-%#06llx])";
			else
				fmt = " (bus address [%#010llx-%#010llx])";

			snprintf(addr, sizeof(addr), fmt,
				 (unsigned long long)(res->start - offset),
				 (unsigned long long)(res->end - offset));
		} else
			addr[0] = '\0';

		dev_info(&bus->dev, "root bus resource %pR%s\n", res, addr);
	}

	down_write(&pci_bus_sem);
	list_add_tail(&bus->node, &pci_root_buses);
	up_write(&pci_bus_sem);

	return 0;

unregister:
	put_device(&bridge->dev);
	device_unregister(&bridge->dev);

free:
	kfree(bus);
	return err;
}

836 837
static struct pci_bus *pci_alloc_child_bus(struct pci_bus *parent,
					   struct pci_dev *bridge, int busnr)
L
Linus Torvalds 已提交
838 839 840
{
	struct pci_bus *child;
	int i;
841
	int ret;
L
Linus Torvalds 已提交
842 843 844 845

	/*
	 * Allocate a new bus, and inherit stuff from the parent..
	 */
846
	child = pci_alloc_bus(parent);
L
Linus Torvalds 已提交
847 848 849 850 851
	if (!child)
		return NULL;

	child->parent = parent;
	child->ops = parent->ops;
852
	child->msi = parent->msi;
L
Linus Torvalds 已提交
853
	child->sysdata = parent->sysdata;
854
	child->bus_flags = parent->bus_flags;
L
Linus Torvalds 已提交
855

856
	/* initialize some portions of the bus device, but don't register it
857
	 * now as the parent is not properly set up yet.
858 859
	 */
	child->dev.class = &pcibus_class;
860
	dev_set_name(&child->dev, "%04x:%02x", pci_domain_nr(child), busnr);
L
Linus Torvalds 已提交
861 862 863 864 865

	/*
	 * Set up the primary, secondary and subordinate
	 * bus numbers.
	 */
866 867 868
	child->number = child->busn_res.start = busnr;
	child->primary = parent->busn_res.start;
	child->busn_res.end = 0xff;
L
Linus Torvalds 已提交
869

870 871 872 873
	if (!bridge) {
		child->dev.parent = parent->bridge;
		goto add_dev;
	}
874 875 876

	child->self = bridge;
	child->bridge = get_device(&bridge->dev);
877
	child->dev.parent = child->bridge;
878
	pci_set_bus_of_node(child);
879 880
	pci_set_bus_speed(child);

L
Linus Torvalds 已提交
881
	/* Set up default resource pointers and names.. */
882
	for (i = 0; i < PCI_BRIDGE_RESOURCE_NUM; i++) {
L
Linus Torvalds 已提交
883 884 885 886 887
		child->resource[i] = &bridge->resource[PCI_BRIDGE_RESOURCES+i];
		child->resource[i]->name = child->name;
	}
	bridge->subordinate = child;

888
add_dev:
889
	pci_set_bus_msi_domain(child);
890 891 892
	ret = device_register(&child->dev);
	WARN_ON(ret < 0);

893 894
	pcibios_add_bus(child);

895 896 897 898 899 900
	if (child->ops->add_bus) {
		ret = child->ops->add_bus(child);
		if (WARN_ON(ret < 0))
			dev_err(&child->dev, "failed to add bus: %d\n", ret);
	}

901 902 903
	/* Create legacy_io and legacy_mem files for this bus */
	pci_create_legacy_files(child);

L
Linus Torvalds 已提交
904 905 906
	return child;
}

R
Ryan Desfosses 已提交
907 908
struct pci_bus *pci_add_new_bus(struct pci_bus *parent, struct pci_dev *dev,
				int busnr)
L
Linus Torvalds 已提交
909 910 911 912
{
	struct pci_bus *child;

	child = pci_alloc_child_bus(parent, dev, busnr);
913
	if (child) {
914
		down_write(&pci_bus_sem);
L
Linus Torvalds 已提交
915
		list_add_tail(&child->node, &parent->children);
916
		up_write(&pci_bus_sem);
917
	}
L
Linus Torvalds 已提交
918 919
	return child;
}
920
EXPORT_SYMBOL(pci_add_new_bus);
L
Linus Torvalds 已提交
921

922 923 924 925 926 927 928 929 930 931 932
static void pci_enable_crs(struct pci_dev *pdev)
{
	u16 root_cap = 0;

	/* Enable CRS Software Visibility if supported */
	pcie_capability_read_word(pdev, PCI_EXP_RTCAP, &root_cap);
	if (root_cap & PCI_EXP_RTCAP_CRSVIS)
		pcie_capability_set_word(pdev, PCI_EXP_RTCTL,
					 PCI_EXP_RTCTL_CRSSVE);
}

L
Linus Torvalds 已提交
933 934 935 936 937 938 939 940 941 942
/*
 * If it's a bridge, configure it and scan the bus behind it.
 * For CardBus bridges, we don't scan behind as the devices will
 * be handled by the bridge driver itself.
 *
 * We need to process bridges in two passes -- first we scan those
 * already configured by the BIOS and after we are done with all of
 * them, we proceed to assigning numbers to the remaining buses in
 * order to avoid overlaps between old and new bus numbers.
 */
B
Bill Pemberton 已提交
943
int pci_scan_bridge(struct pci_bus *bus, struct pci_dev *dev, int max, int pass)
L
Linus Torvalds 已提交
944 945 946
{
	struct pci_bus *child;
	int is_cardbus = (dev->hdr_type == PCI_HEADER_TYPE_CARDBUS);
947
	u32 buses, i, j = 0;
L
Linus Torvalds 已提交
948
	u16 bctl;
949
	u8 primary, secondary, subordinate;
950
	int broken = 0;
L
Linus Torvalds 已提交
951

952 953 954 955 956 957
	/*
	 * Make sure the bridge is powered on to be able to access config
	 * space of devices below it.
	 */
	pm_runtime_get_sync(&dev->dev);

L
Linus Torvalds 已提交
958
	pci_read_config_dword(dev, PCI_PRIMARY_BUS, &buses);
959 960 961
	primary = buses & 0xFF;
	secondary = (buses >> 8) & 0xFF;
	subordinate = (buses >> 16) & 0xFF;
L
Linus Torvalds 已提交
962

963 964
	dev_dbg(&dev->dev, "scanning [bus %02x-%02x] behind bridge, pass %d\n",
		secondary, subordinate, pass);
L
Linus Torvalds 已提交
965

966 967 968 969 970
	if (!primary && (primary != bus->number) && secondary && subordinate) {
		dev_warn(&dev->dev, "Primary bus is hard wired to 0\n");
		primary = bus->number;
	}

971 972
	/* Check if setup is sensible at all */
	if (!pass &&
973
	    (primary != bus->number || secondary <= bus->number ||
974
	     secondary > subordinate)) {
975 976
		dev_info(&dev->dev, "bridge configuration invalid ([bus %02x-%02x]), reconfiguring\n",
			 secondary, subordinate);
977 978 979
		broken = 1;
	}

L
Linus Torvalds 已提交
980
	/* Disable MasterAbortMode during probing to avoid reporting
981
	   of bus errors (in some architectures) */
L
Linus Torvalds 已提交
982 983 984 985
	pci_read_config_word(dev, PCI_BRIDGE_CONTROL, &bctl);
	pci_write_config_word(dev, PCI_BRIDGE_CONTROL,
			      bctl & ~PCI_BRIDGE_CTL_MASTER_ABORT);

986 987
	pci_enable_crs(dev);

988 989 990
	if ((secondary || subordinate) && !pcibios_assign_all_busses() &&
	    !is_cardbus && !broken) {
		unsigned int cmax;
L
Linus Torvalds 已提交
991 992 993 994 995
		/*
		 * Bus already configured by firmware, process it in the first
		 * pass and just note the configuration.
		 */
		if (pass)
996
			goto out;
L
Linus Torvalds 已提交
997 998

		/*
999 1000 1001 1002
		 * The bus might already exist for two reasons: Either we are
		 * rescanning the bus or the bus is reachable through more than
		 * one bridge. The second case can happen with the i450NX
		 * chipset.
L
Linus Torvalds 已提交
1003
		 */
1004
		child = pci_find_bus(pci_domain_nr(bus), secondary);
A
Alex Chiang 已提交
1005
		if (!child) {
1006
			child = pci_add_new_bus(bus, dev, secondary);
A
Alex Chiang 已提交
1007 1008
			if (!child)
				goto out;
1009
			child->primary = primary;
Y
Yinghai Lu 已提交
1010
			pci_bus_insert_busn_res(child, secondary, subordinate);
A
Alex Chiang 已提交
1011
			child->bridge_ctl = bctl;
L
Linus Torvalds 已提交
1012 1013 1014
		}

		cmax = pci_scan_child_bus(child);
1015 1016 1017 1018 1019 1020
		if (cmax > subordinate)
			dev_warn(&dev->dev, "bridge has subordinate %02x but max busn %02x\n",
				 subordinate, cmax);
		/* subordinate should equal child->busn_res.end */
		if (subordinate > max)
			max = subordinate;
L
Linus Torvalds 已提交
1021 1022 1023 1024 1025
	} else {
		/*
		 * We need to assign a number to this bus which we always
		 * do in the second pass.
		 */
1026
		if (!pass) {
1027
			if (pcibios_assign_all_busses() || broken || is_cardbus)
1028 1029 1030 1031 1032 1033 1034 1035
				/* Temporarily disable forwarding of the
				   configuration cycles on all bridges in
				   this bus segment to avoid possible
				   conflicts in the second pass between two
				   bridges programmed with overlapping
				   bus ranges. */
				pci_write_config_dword(dev, PCI_PRIMARY_BUS,
						       buses & ~0xffffff);
1036
			goto out;
1037
		}
L
Linus Torvalds 已提交
1038 1039 1040 1041

		/* Clear errors */
		pci_write_config_word(dev, PCI_STATUS, 0xffff);

1042 1043 1044
		/* Prevent assigning a bus number that already exists.
		 * This can happen when a bridge is hot-plugged, so in
		 * this case we only re-scan this bus. */
1045 1046
		child = pci_find_bus(pci_domain_nr(bus), max+1);
		if (!child) {
1047
			child = pci_add_new_bus(bus, dev, max+1);
1048 1049
			if (!child)
				goto out;
1050
			pci_bus_insert_busn_res(child, max+1, 0xff);
1051
		}
1052
		max++;
L
Linus Torvalds 已提交
1053 1054
		buses = (buses & 0xff000000)
		      | ((unsigned int)(child->primary)     <<  0)
1055 1056
		      | ((unsigned int)(child->busn_res.start)   <<  8)
		      | ((unsigned int)(child->busn_res.end) << 16);
L
Linus Torvalds 已提交
1057 1058 1059 1060 1061 1062 1063 1064 1065

		/*
		 * yenta.c forces a secondary latency timer of 176.
		 * Copy that behaviour here.
		 */
		if (is_cardbus) {
			buses &= ~0xff000000;
			buses |= CARDBUS_LATENCY_TIMER << 24;
		}
1066

L
Linus Torvalds 已提交
1067 1068 1069 1070 1071 1072
		/*
		 * We need to blast all three values with a single write.
		 */
		pci_write_config_dword(dev, PCI_PRIMARY_BUS, buses);

		if (!is_cardbus) {
1073
			child->bridge_ctl = bctl;
L
Linus Torvalds 已提交
1074 1075 1076 1077 1078 1079 1080
			max = pci_scan_child_bus(child);
		} else {
			/*
			 * For CardBus bridges, we leave 4 bus numbers
			 * as cards with a PCI-to-PCI bridge can be
			 * inserted later.
			 */
R
Ryan Desfosses 已提交
1081
			for (i = 0; i < CARDBUS_RESERVE_BUSNR; i++) {
1082
				struct pci_bus *parent = bus;
1083 1084 1085
				if (pci_find_bus(pci_domain_nr(bus),
							max+i+1))
					break;
1086 1087
				while (parent->parent) {
					if ((!pcibios_assign_all_busses()) &&
1088 1089
					    (parent->busn_res.end > max) &&
					    (parent->busn_res.end <= max+i)) {
1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103
						j = 1;
					}
					parent = parent->parent;
				}
				if (j) {
					/*
					 * Often, there are two cardbus bridges
					 * -- try to leave one valid bus number
					 * for each one.
					 */
					i /= 2;
					break;
				}
			}
1104
			max += i;
L
Linus Torvalds 已提交
1105 1106 1107 1108
		}
		/*
		 * Set the subordinate bus number to its real value.
		 */
Y
Yinghai Lu 已提交
1109
		pci_bus_update_busn_res_end(child, max);
L
Linus Torvalds 已提交
1110 1111 1112
		pci_write_config_byte(dev, PCI_SUBORDINATE_BUS, max);
	}

1113 1114 1115
	sprintf(child->name,
		(is_cardbus ? "PCI CardBus %04x:%02x" : "PCI Bus %04x:%02x"),
		pci_domain_nr(bus), child->number);
L
Linus Torvalds 已提交
1116

1117
	/* Has only triggered on CardBus, fixup is in yenta_socket */
1118
	while (bus->parent) {
1119 1120
		if ((child->busn_res.end > bus->busn_res.end) ||
		    (child->number > bus->busn_res.end) ||
1121
		    (child->number < bus->number) ||
1122
		    (child->busn_res.end < bus->number)) {
1123
			dev_info(&child->dev, "%pR %s hidden behind%s bridge %s %pR\n",
1124 1125 1126
				&child->busn_res,
				(bus->number > child->busn_res.end &&
				 bus->busn_res.end < child->number) ?
1127 1128
					"wholly" : "partially",
				bus->self->transparent ? " transparent" : "",
1129
				dev_name(&bus->dev),
1130
				&bus->busn_res);
1131 1132 1133 1134
		}
		bus = bus->parent;
	}

1135 1136 1137
out:
	pci_write_config_word(dev, PCI_BRIDGE_CONTROL, bctl);

1138 1139
	pm_runtime_put(&dev->dev);

L
Linus Torvalds 已提交
1140 1141
	return max;
}
1142
EXPORT_SYMBOL(pci_scan_bridge);
L
Linus Torvalds 已提交
1143 1144 1145 1146 1147 1148 1149 1150 1151 1152

/*
 * Read interrupt line and base address registers.
 * The architecture-dependent code can tweak these, of course.
 */
static void pci_read_irq(struct pci_dev *dev)
{
	unsigned char irq;

	pci_read_config_byte(dev, PCI_INTERRUPT_PIN, &irq);
1153
	dev->pin = irq;
L
Linus Torvalds 已提交
1154 1155 1156 1157 1158
	if (irq)
		pci_read_config_byte(dev, PCI_INTERRUPT_LINE, &irq);
	dev->irq = irq;
}

1159
void set_pcie_port_type(struct pci_dev *pdev)
Y
Yu Zhao 已提交
1160 1161 1162
{
	int pos;
	u16 reg16;
1163 1164
	int type;
	struct pci_dev *parent;
Y
Yu Zhao 已提交
1165 1166 1167 1168

	pos = pci_find_capability(pdev, PCI_CAP_ID_EXP);
	if (!pos)
		return;
1169
	pdev->pcie_cap = pos;
Y
Yu Zhao 已提交
1170
	pci_read_config_word(pdev, pos + PCI_EXP_FLAGS, &reg16);
1171
	pdev->pcie_flags_reg = reg16;
1172 1173
	pci_read_config_word(pdev, pos + PCI_EXP_DEVCAP, &reg16);
	pdev->pcie_mpss = reg16 & PCI_EXP_DEVCAP_PAYLOAD;
1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186

	/*
	 * A Root Port is always the upstream end of a Link.  No PCIe
	 * component has two Links.  Two Links are connected by a Switch
	 * that has a Port on each Link and internal logic to connect the
	 * two Ports.
	 */
	type = pci_pcie_type(pdev);
	if (type == PCI_EXP_TYPE_ROOT_PORT)
		pdev->has_secondary_link = 1;
	else if (type == PCI_EXP_TYPE_UPSTREAM ||
		 type == PCI_EXP_TYPE_DOWNSTREAM) {
		parent = pci_upstream_bridge(pdev);
1187 1188 1189 1190 1191 1192

		/*
		 * Usually there's an upstream device (Root Port or Switch
		 * Downstream Port), but we can't assume one exists.
		 */
		if (parent && !parent->has_secondary_link)
1193 1194
			pdev->has_secondary_link = 1;
	}
Y
Yu Zhao 已提交
1195 1196
}

1197
void set_pcie_hotplug_bridge(struct pci_dev *pdev)
1198 1199 1200
{
	u32 reg32;

1201
	pcie_capability_read_dword(pdev, PCI_EXP_SLTCAP, &reg32);
1202 1203 1204 1205
	if (reg32 & PCI_EXP_SLTCAP_HPC)
		pdev->is_hotplug_bridge = 1;
}

1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242
/**
 * pci_ext_cfg_is_aliased - is ext config space just an alias of std config?
 * @dev: PCI device
 *
 * PCI Express to PCI/PCI-X Bridge Specification, rev 1.0, 4.1.4 says that
 * when forwarding a type1 configuration request the bridge must check that
 * the extended register address field is zero.  The bridge is not permitted
 * to forward the transactions and must handle it as an Unsupported Request.
 * Some bridges do not follow this rule and simply drop the extended register
 * bits, resulting in the standard config space being aliased, every 256
 * bytes across the entire configuration space.  Test for this condition by
 * comparing the first dword of each potential alias to the vendor/device ID.
 * Known offenders:
 *   ASM1083/1085 PCIe-to-PCI Reversible Bridge (1b21:1080, rev 01 & 03)
 *   AMD/ATI SBx00 PCI to PCI Bridge (1002:4384, rev 40)
 */
static bool pci_ext_cfg_is_aliased(struct pci_dev *dev)
{
#ifdef CONFIG_PCI_QUIRKS
	int pos;
	u32 header, tmp;

	pci_read_config_dword(dev, PCI_VENDOR_ID, &header);

	for (pos = PCI_CFG_SPACE_SIZE;
	     pos < PCI_CFG_SPACE_EXP_SIZE; pos += PCI_CFG_SPACE_SIZE) {
		if (pci_read_config_dword(dev, pos, &tmp) != PCIBIOS_SUCCESSFUL
		    || header != tmp)
			return false;
	}

	return true;
#else
	return false;
#endif
}

1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259
/**
 * pci_cfg_space_size - get the configuration space size of the PCI device.
 * @dev: PCI device
 *
 * Regular PCI devices have 256 bytes, but PCI-X 2 and PCI Express devices
 * have 4096 bytes.  Even if the device is capable, that doesn't mean we can
 * access it.  Maybe we don't have a way to generate extended config space
 * accesses, or the device is behind a reverse Express bridge.  So we try
 * reading the dword at 0x100 which must either be 0 or a valid extended
 * capability header.
 */
static int pci_cfg_space_size_ext(struct pci_dev *dev)
{
	u32 status;
	int pos = PCI_CFG_SPACE_SIZE;

	if (pci_read_config_dword(dev, pos, &status) != PCIBIOS_SUCCESSFUL)
1260
		return PCI_CFG_SPACE_SIZE;
1261
	if (status == 0xffffffff || pci_ext_cfg_is_aliased(dev))
1262
		return PCI_CFG_SPACE_SIZE;
1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276

	return PCI_CFG_SPACE_EXP_SIZE;
}

int pci_cfg_space_size(struct pci_dev *dev)
{
	int pos;
	u32 status;
	u16 class;

	class = dev->class >> 8;
	if (class == PCI_CLASS_BRIDGE_HOST)
		return pci_cfg_space_size_ext(dev);

1277 1278
	if (pci_is_pcie(dev))
		return pci_cfg_space_size_ext(dev);
1279

1280 1281 1282
	pos = pci_find_capability(dev, PCI_CAP_ID_PCIX);
	if (!pos)
		return PCI_CFG_SPACE_SIZE;
1283

1284 1285 1286
	pci_read_config_dword(dev, pos + PCI_X_STATUS, &status);
	if (status & (PCI_X_STATUS_266MHZ | PCI_X_STATUS_533MHZ))
		return pci_cfg_space_size_ext(dev);
1287 1288 1289 1290

	return PCI_CFG_SPACE_SIZE;
}

1291
#define LEGACY_IO_RESOURCE	(IORESOURCE_IO | IORESOURCE_PCI_FIXED)
1292

1293
static void pci_msi_setup_pci_dev(struct pci_dev *dev)
1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308
{
	/*
	 * Disable the MSI hardware to avoid screaming interrupts
	 * during boot.  This is the power on reset default so
	 * usually this should be a noop.
	 */
	dev->msi_cap = pci_find_capability(dev, PCI_CAP_ID_MSI);
	if (dev->msi_cap)
		pci_msi_set_enable(dev, 0);

	dev->msix_cap = pci_find_capability(dev, PCI_CAP_ID_MSIX);
	if (dev->msix_cap)
		pci_msix_clear_and_set_ctrl(dev, PCI_MSIX_FLAGS_ENABLE, 0);
}

L
Linus Torvalds 已提交
1309 1310 1311 1312
/**
 * pci_setup_device - fill in class and map information of a device
 * @dev: the device structure to fill
 *
1313
 * Initialize the device structure with information about the device's
L
Linus Torvalds 已提交
1314 1315
 * vendor,class,memory and IO-space addresses,IRQ lines etc.
 * Called at initialisation of the PCI subsystem and by CardBus services.
Y
Yu Zhao 已提交
1316 1317
 * Returns 0 on success and negative if unknown type of device (not normal,
 * bridge or CardBus).
L
Linus Torvalds 已提交
1318
 */
Y
Yu Zhao 已提交
1319
int pci_setup_device(struct pci_dev *dev)
L
Linus Torvalds 已提交
1320 1321
{
	u32 class;
1322
	u16 cmd;
Y
Yu Zhao 已提交
1323
	u8 hdr_type;
1324
	int pos = 0;
1325 1326
	struct pci_bus_region region;
	struct resource *res;
Y
Yu Zhao 已提交
1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338

	if (pci_read_config_byte(dev, PCI_HEADER_TYPE, &hdr_type))
		return -EIO;

	dev->sysdata = dev->bus->sysdata;
	dev->dev.parent = dev->bus->bridge;
	dev->dev.bus = &pci_bus_type;
	dev->hdr_type = hdr_type & 0x7f;
	dev->multifunction = !!(hdr_type & 0x80);
	dev->error_state = pci_channel_io_normal;
	set_pcie_port_type(dev);

1339
	pci_dev_assign_slot(dev);
Y
Yu Zhao 已提交
1340 1341 1342
	/* Assume 32-bit PCI; let 64-bit PCI cards (which are far rarer)
	   set this higher, assuming the system even supports it.  */
	dev->dma_mask = 0xffffffff;
L
Linus Torvalds 已提交
1343

1344 1345 1346
	dev_set_name(&dev->dev, "%04x:%02x:%02x.%d", pci_domain_nr(dev->bus),
		     dev->bus->number, PCI_SLOT(dev->devfn),
		     PCI_FUNC(dev->devfn));
L
Linus Torvalds 已提交
1347 1348

	pci_read_config_dword(dev, PCI_CLASS_REVISION, &class);
A
Auke Kok 已提交
1349
	dev->revision = class & 0xff;
Y
Yinghai Lu 已提交
1350
	dev->class = class >> 8;		    /* upper 3 bytes */
L
Linus Torvalds 已提交
1351

Y
Yinghai Lu 已提交
1352 1353
	dev_printk(KERN_DEBUG, &dev->dev, "[%04x:%04x] type %02x class %#08x\n",
		   dev->vendor, dev->device, dev->hdr_type, dev->class);
L
Linus Torvalds 已提交
1354

1355 1356 1357
	/* need to have dev->class ready */
	dev->cfg_size = pci_cfg_space_size(dev);

L
Linus Torvalds 已提交
1358
	/* "Unknown power state" */
1359
	dev->current_state = PCI_UNKNOWN;
L
Linus Torvalds 已提交
1360 1361 1362

	/* Early fixups, before probing the BARs */
	pci_fixup_device(pci_fixup_early, dev);
1363 1364
	/* device class may be changed after fixup */
	class = dev->class >> 8;
L
Linus Torvalds 已提交
1365

1366 1367 1368 1369 1370 1371 1372 1373 1374 1375
	if (dev->non_compliant_bars) {
		pci_read_config_word(dev, PCI_COMMAND, &cmd);
		if (cmd & (PCI_COMMAND_IO | PCI_COMMAND_MEMORY)) {
			dev_info(&dev->dev, "device has non-compliant BARs; disabling IO/MEM decoding\n");
			cmd &= ~PCI_COMMAND_IO;
			cmd &= ~PCI_COMMAND_MEMORY;
			pci_write_config_word(dev, PCI_COMMAND, cmd);
		}
	}

L
Linus Torvalds 已提交
1376 1377 1378 1379 1380 1381 1382 1383
	switch (dev->hdr_type) {		    /* header type */
	case PCI_HEADER_TYPE_NORMAL:		    /* standard header */
		if (class == PCI_CLASS_BRIDGE_PCI)
			goto bad;
		pci_read_irq(dev);
		pci_read_bases(dev, 6, PCI_ROM_ADDRESS);
		pci_read_config_word(dev, PCI_SUBSYSTEM_VENDOR_ID, &dev->subsystem_vendor);
		pci_read_config_word(dev, PCI_SUBSYSTEM_ID, &dev->subsystem_device);
1384 1385

		/*
1386 1387 1388 1389
		 * Do the ugly legacy mode stuff here rather than broken chip
		 * quirk code. Legacy mode ATA controllers have fixed
		 * addresses. These are not always echoed in BAR0-3, and
		 * BAR0-3 in a few cases contain junk!
1390 1391 1392 1393 1394
		 */
		if (class == PCI_CLASS_STORAGE_IDE) {
			u8 progif;
			pci_read_config_byte(dev, PCI_CLASS_PROG, &progif);
			if ((progif & 1) == 0) {
1395 1396 1397 1398
				region.start = 0x1F0;
				region.end = 0x1F7;
				res = &dev->resource[0];
				res->flags = LEGACY_IO_RESOURCE;
1399
				pcibios_bus_to_resource(dev->bus, res, &region);
1400 1401
				dev_info(&dev->dev, "legacy IDE quirk: reg 0x10: %pR\n",
					 res);
1402 1403 1404 1405
				region.start = 0x3F6;
				region.end = 0x3F6;
				res = &dev->resource[1];
				res->flags = LEGACY_IO_RESOURCE;
1406
				pcibios_bus_to_resource(dev->bus, res, &region);
1407 1408
				dev_info(&dev->dev, "legacy IDE quirk: reg 0x14: %pR\n",
					 res);
1409 1410
			}
			if ((progif & 4) == 0) {
1411 1412 1413 1414
				region.start = 0x170;
				region.end = 0x177;
				res = &dev->resource[2];
				res->flags = LEGACY_IO_RESOURCE;
1415
				pcibios_bus_to_resource(dev->bus, res, &region);
1416 1417
				dev_info(&dev->dev, "legacy IDE quirk: reg 0x18: %pR\n",
					 res);
1418 1419 1420 1421
				region.start = 0x376;
				region.end = 0x376;
				res = &dev->resource[3];
				res->flags = LEGACY_IO_RESOURCE;
1422
				pcibios_bus_to_resource(dev->bus, res, &region);
1423 1424
				dev_info(&dev->dev, "legacy IDE quirk: reg 0x1c: %pR\n",
					 res);
1425 1426
			}
		}
L
Linus Torvalds 已提交
1427 1428 1429 1430 1431 1432 1433
		break;

	case PCI_HEADER_TYPE_BRIDGE:		    /* bridge header */
		if (class != PCI_CLASS_BRIDGE_PCI)
			goto bad;
		/* The PCI-to-PCI bridge spec requires that subtractive
		   decoding (i.e. transparent) bridge must have programming
1434
		   interface code of 0x01. */
1435
		pci_read_irq(dev);
L
Linus Torvalds 已提交
1436 1437
		dev->transparent = ((dev->class & 0xff) == 1);
		pci_read_bases(dev, 2, PCI_ROM_ADDRESS1);
1438
		set_pcie_hotplug_bridge(dev);
1439 1440 1441 1442 1443
		pos = pci_find_capability(dev, PCI_CAP_ID_SSVID);
		if (pos) {
			pci_read_config_word(dev, pos + PCI_SSVID_VENDOR_ID, &dev->subsystem_vendor);
			pci_read_config_word(dev, pos + PCI_SSVID_DEVICE_ID, &dev->subsystem_device);
		}
L
Linus Torvalds 已提交
1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455
		break;

	case PCI_HEADER_TYPE_CARDBUS:		    /* CardBus bridge header */
		if (class != PCI_CLASS_BRIDGE_CARDBUS)
			goto bad;
		pci_read_irq(dev);
		pci_read_bases(dev, 1, 0);
		pci_read_config_word(dev, PCI_CB_SUBSYSTEM_VENDOR_ID, &dev->subsystem_vendor);
		pci_read_config_word(dev, PCI_CB_SUBSYSTEM_ID, &dev->subsystem_device);
		break;

	default:				    /* unknown header */
1456 1457
		dev_err(&dev->dev, "unknown header type %02x, ignoring device\n",
			dev->hdr_type);
Y
Yu Zhao 已提交
1458
		return -EIO;
L
Linus Torvalds 已提交
1459 1460

	bad:
1461 1462
		dev_err(&dev->dev, "ignoring class %#08x (doesn't match header type %02x)\n",
			dev->class, dev->hdr_type);
1463
		dev->class = PCI_CLASS_NOT_DEFINED << 8;
L
Linus Torvalds 已提交
1464 1465 1466 1467 1468 1469
	}

	/* We found a fine healthy device, go go go... */
	return 0;
}

1470 1471 1472
static void pci_configure_mps(struct pci_dev *dev)
{
	struct pci_dev *bridge = pci_upstream_bridge(dev);
1473
	int mps, p_mps, rc;
1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488

	if (!pci_is_pcie(dev) || !bridge || !pci_is_pcie(bridge))
		return;

	mps = pcie_get_mps(dev);
	p_mps = pcie_get_mps(bridge);

	if (mps == p_mps)
		return;

	if (pcie_bus_config == PCIE_BUS_TUNE_OFF) {
		dev_warn(&dev->dev, "Max Payload Size %d, but upstream %s set to %d; if necessary, use \"pci=pcie_bus_safe\" and report a bug\n",
			 mps, pci_name(bridge), p_mps);
		return;
	}
1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505

	/*
	 * Fancier MPS configuration is done later by
	 * pcie_bus_configure_settings()
	 */
	if (pcie_bus_config != PCIE_BUS_DEFAULT)
		return;

	rc = pcie_set_mps(dev, p_mps);
	if (rc) {
		dev_warn(&dev->dev, "can't set Max Payload Size to %d; if necessary, use \"pci=pcie_bus_safe\" and report a bug\n",
			 p_mps);
		return;
	}

	dev_info(&dev->dev, "Max Payload Size set to %d (was %d, max %d)\n",
		 p_mps, mps, 128 << dev->pcie_mpss);
1506 1507
}

1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519
static struct hpp_type0 pci_default_type0 = {
	.revision = 1,
	.cache_line_size = 8,
	.latency_timer = 0x40,
	.enable_serr = 0,
	.enable_perr = 0,
};

static void program_hpp_type0(struct pci_dev *dev, struct hpp_type0 *hpp)
{
	u16 pci_cmd, pci_bctl;

1520
	if (!hpp)
1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571
		hpp = &pci_default_type0;

	if (hpp->revision > 1) {
		dev_warn(&dev->dev,
			 "PCI settings rev %d not supported; using defaults\n",
			 hpp->revision);
		hpp = &pci_default_type0;
	}

	pci_write_config_byte(dev, PCI_CACHE_LINE_SIZE, hpp->cache_line_size);
	pci_write_config_byte(dev, PCI_LATENCY_TIMER, hpp->latency_timer);
	pci_read_config_word(dev, PCI_COMMAND, &pci_cmd);
	if (hpp->enable_serr)
		pci_cmd |= PCI_COMMAND_SERR;
	if (hpp->enable_perr)
		pci_cmd |= PCI_COMMAND_PARITY;
	pci_write_config_word(dev, PCI_COMMAND, pci_cmd);

	/* Program bridge control value */
	if ((dev->class >> 8) == PCI_CLASS_BRIDGE_PCI) {
		pci_write_config_byte(dev, PCI_SEC_LATENCY_TIMER,
				      hpp->latency_timer);
		pci_read_config_word(dev, PCI_BRIDGE_CONTROL, &pci_bctl);
		if (hpp->enable_serr)
			pci_bctl |= PCI_BRIDGE_CTL_SERR;
		if (hpp->enable_perr)
			pci_bctl |= PCI_BRIDGE_CTL_PARITY;
		pci_write_config_word(dev, PCI_BRIDGE_CONTROL, pci_bctl);
	}
}

static void program_hpp_type1(struct pci_dev *dev, struct hpp_type1 *hpp)
{
	if (hpp)
		dev_warn(&dev->dev, "PCI-X settings not supported\n");
}

static void program_hpp_type2(struct pci_dev *dev, struct hpp_type2 *hpp)
{
	int pos;
	u32 reg32;

	if (!hpp)
		return;

	if (hpp->revision > 1) {
		dev_warn(&dev->dev, "PCIe settings rev %d not supported\n",
			 hpp->revision);
		return;
	}

1572 1573 1574 1575 1576 1577 1578 1579 1580 1581
	/*
	 * Don't allow _HPX to change MPS or MRRS settings.  We manage
	 * those to make sure they're consistent with the rest of the
	 * platform.
	 */
	hpp->pci_exp_devctl_and |= PCI_EXP_DEVCTL_PAYLOAD |
				    PCI_EXP_DEVCTL_READRQ;
	hpp->pci_exp_devctl_or &= ~(PCI_EXP_DEVCTL_PAYLOAD |
				    PCI_EXP_DEVCTL_READRQ);

1582 1583 1584 1585 1586
	/* Initialize Device Control Register */
	pcie_capability_clear_and_set_word(dev, PCI_EXP_DEVCTL,
			~hpp->pci_exp_devctl_and, hpp->pci_exp_devctl_or);

	/* Initialize Link Control Register */
1587
	if (pcie_cap_has_lnkctl(dev))
1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623
		pcie_capability_clear_and_set_word(dev, PCI_EXP_LNKCTL,
			~hpp->pci_exp_lnkctl_and, hpp->pci_exp_lnkctl_or);

	/* Find Advanced Error Reporting Enhanced Capability */
	pos = pci_find_ext_capability(dev, PCI_EXT_CAP_ID_ERR);
	if (!pos)
		return;

	/* Initialize Uncorrectable Error Mask Register */
	pci_read_config_dword(dev, pos + PCI_ERR_UNCOR_MASK, &reg32);
	reg32 = (reg32 & hpp->unc_err_mask_and) | hpp->unc_err_mask_or;
	pci_write_config_dword(dev, pos + PCI_ERR_UNCOR_MASK, reg32);

	/* Initialize Uncorrectable Error Severity Register */
	pci_read_config_dword(dev, pos + PCI_ERR_UNCOR_SEVER, &reg32);
	reg32 = (reg32 & hpp->unc_err_sever_and) | hpp->unc_err_sever_or;
	pci_write_config_dword(dev, pos + PCI_ERR_UNCOR_SEVER, reg32);

	/* Initialize Correctable Error Mask Register */
	pci_read_config_dword(dev, pos + PCI_ERR_COR_MASK, &reg32);
	reg32 = (reg32 & hpp->cor_err_mask_and) | hpp->cor_err_mask_or;
	pci_write_config_dword(dev, pos + PCI_ERR_COR_MASK, reg32);

	/* Initialize Advanced Error Capabilities and Control Register */
	pci_read_config_dword(dev, pos + PCI_ERR_CAP, &reg32);
	reg32 = (reg32 & hpp->adv_err_cap_and) | hpp->adv_err_cap_or;
	pci_write_config_dword(dev, pos + PCI_ERR_CAP, reg32);

	/*
	 * FIXME: The following two registers are not supported yet.
	 *
	 *   o Secondary Uncorrectable Error Severity Register
	 *   o Secondary Uncorrectable Error Mask Register
	 */
}

1624 1625 1626 1627 1628
static void pci_configure_device(struct pci_dev *dev)
{
	struct hotplug_params hpp;
	int ret;

1629 1630
	pci_configure_mps(dev);

1631 1632 1633 1634 1635 1636 1637 1638 1639 1640
	memset(&hpp, 0, sizeof(hpp));
	ret = pci_get_hp_params(dev, &hpp);
	if (ret)
		return;

	program_hpp_type2(dev, hpp.t2);
	program_hpp_type1(dev, hpp.t1);
	program_hpp_type0(dev, hpp.t0);
}

1641 1642 1643
static void pci_release_capabilities(struct pci_dev *dev)
{
	pci_vpd_release(dev);
1644
	pci_iov_release(dev);
1645
	pci_free_cap_save_buffers(dev);
1646 1647
}

L
Linus Torvalds 已提交
1648 1649 1650 1651 1652 1653 1654 1655 1656
/**
 * pci_release_dev - free a pci device structure when all users of it are finished.
 * @dev: device that's been disconnected
 *
 * Will be called only by the device core when all users of this pci device are
 * done.
 */
static void pci_release_dev(struct device *dev)
{
1657
	struct pci_dev *pci_dev;
L
Linus Torvalds 已提交
1658

1659
	pci_dev = to_pci_dev(dev);
1660
	pci_release_capabilities(pci_dev);
1661
	pci_release_of_node(pci_dev);
1662
	pcibios_release_device(pci_dev);
1663
	pci_bus_put(pci_dev->bus);
1664
	kfree(pci_dev->driver_override);
1665
	kfree(pci_dev->dma_alias_mask);
L
Linus Torvalds 已提交
1666 1667 1668
	kfree(pci_dev);
}

1669
struct pci_dev *pci_alloc_dev(struct pci_bus *bus)
1670 1671 1672 1673 1674 1675 1676 1677
{
	struct pci_dev *dev;

	dev = kzalloc(sizeof(struct pci_dev), GFP_KERNEL);
	if (!dev)
		return NULL;

	INIT_LIST_HEAD(&dev->bus_list);
1678
	dev->dev.type = &pci_dev_type;
1679
	dev->bus = pci_bus_get(bus);
1680 1681 1682

	return dev;
}
1683 1684
EXPORT_SYMBOL(pci_alloc_dev);

1685
bool pci_bus_read_dev_vendor_id(struct pci_bus *bus, int devfn, u32 *l,
R
Ryan Desfosses 已提交
1686
				int crs_timeout)
L
Linus Torvalds 已提交
1687 1688 1689
{
	int delay = 1;

1690 1691
	if (pci_bus_read_config_dword(bus, devfn, PCI_VENDOR_ID, l))
		return false;
L
Linus Torvalds 已提交
1692 1693

	/* some broken boards return 0 or ~0 if a slot is empty: */
1694 1695 1696
	if (*l == 0xffffffff || *l == 0x00000000 ||
	    *l == 0x0000ffff || *l == 0xffff0000)
		return false;
L
Linus Torvalds 已提交
1697

1698 1699 1700 1701 1702 1703 1704
	/*
	 * Configuration Request Retry Status.  Some root ports return the
	 * actual device ID instead of the synthetic ID (0xFFFF) required
	 * by the PCIe spec.  Ignore the device ID and only check for
	 * (vendor id == 1).
	 */
	while ((*l & 0xffff) == 0x0001) {
1705 1706 1707
		if (!crs_timeout)
			return false;

L
Linus Torvalds 已提交
1708 1709
		msleep(delay);
		delay *= 2;
1710 1711
		if (pci_bus_read_config_dword(bus, devfn, PCI_VENDOR_ID, l))
			return false;
L
Linus Torvalds 已提交
1712
		/* Card hasn't responded in 60 seconds?  Must be stuck. */
1713
		if (delay > crs_timeout) {
1714 1715 1716
			printk(KERN_WARNING "pci %04x:%02x:%02x.%d: not responding\n",
			       pci_domain_nr(bus), bus->number, PCI_SLOT(devfn),
			       PCI_FUNC(devfn));
1717
			return false;
L
Linus Torvalds 已提交
1718 1719 1720
		}
	}

1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736
	return true;
}
EXPORT_SYMBOL(pci_bus_read_dev_vendor_id);

/*
 * Read the config data for a PCI device, sanity-check it
 * and fill in the dev structure...
 */
static struct pci_dev *pci_scan_device(struct pci_bus *bus, int devfn)
{
	struct pci_dev *dev;
	u32 l;

	if (!pci_bus_read_dev_vendor_id(bus, devfn, &l, 60*1000))
		return NULL;

1737
	dev = pci_alloc_dev(bus);
L
Linus Torvalds 已提交
1738 1739 1740 1741 1742 1743
	if (!dev)
		return NULL;

	dev->devfn = devfn;
	dev->vendor = l & 0xffff;
	dev->device = (l >> 16) & 0xffff;
1744

1745 1746
	pci_set_of_node(dev);

Y
Yu Zhao 已提交
1747
	if (pci_setup_device(dev)) {
1748
		pci_bus_put(dev->bus);
L
Linus Torvalds 已提交
1749 1750 1751 1752 1753 1754 1755
		kfree(dev);
		return NULL;
	}

	return dev;
}

1756 1757
static void pci_init_capabilities(struct pci_dev *dev)
{
1758 1759 1760
	/* Enhanced Allocation */
	pci_ea_init(dev);

1761 1762
	/* Setup MSI caps & disable MSI/MSI-X interrupts */
	pci_msi_setup_pci_dev(dev);
1763

1764 1765 1766
	/* Buffers for saving PCIe and PCI-X capabilities */
	pci_allocate_cap_save_buffers(dev);

1767 1768 1769 1770
	/* Power Management */
	pci_pm_init(dev);

	/* Vital Product Data */
1771
	pci_vpd_init(dev);
Y
Yu Zhao 已提交
1772 1773

	/* Alternative Routing-ID Forwarding */
1774
	pci_configure_ari(dev);
1775 1776 1777

	/* Single Root I/O Virtualization */
	pci_iov_init(dev);
1778

1779 1780 1781
	/* Address Translation Services */
	pci_ats_init(dev);

1782
	/* Enable ACS P2P upstream forwarding */
C
Chris Wright 已提交
1783
	pci_enable_acs(dev);
1784

1785 1786
	/* Precision Time Measurement */
	pci_ptm_init(dev);
B
Bjorn Helgaas 已提交
1787

K
Keith Busch 已提交
1788 1789
	/* Advanced Error Reporting */
	pci_aer_init(dev);
1790 1791
}

M
Marc Zyngier 已提交
1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808
/*
 * This is the equivalent of pci_host_bridge_msi_domain that acts on
 * devices. Firmware interfaces that can select the MSI domain on a
 * per-device basis should be called from here.
 */
static struct irq_domain *pci_dev_msi_domain(struct pci_dev *dev)
{
	struct irq_domain *d;

	/*
	 * If a domain has been set through the pcibios_add_device
	 * callback, then this is the one (platform code knows best).
	 */
	d = dev_get_msi_domain(&dev->dev);
	if (d)
		return d;

1809 1810 1811 1812 1813 1814 1815 1816
	/*
	 * Let's see if we have a firmware interface able to provide
	 * the domain.
	 */
	d = pci_msi_get_device_domain(dev);
	if (d)
		return d;

M
Marc Zyngier 已提交
1817 1818 1819
	return NULL;
}

1820 1821
static void pci_set_msi_domain(struct pci_dev *dev)
{
M
Marc Zyngier 已提交
1822 1823
	struct irq_domain *d;

1824
	/*
M
Marc Zyngier 已提交
1825 1826 1827
	 * If the platform or firmware interfaces cannot supply a
	 * device-specific MSI domain, then inherit the default domain
	 * from the host bridge itself.
1828
	 */
M
Marc Zyngier 已提交
1829 1830 1831 1832 1833
	d = pci_dev_msi_domain(dev);
	if (!d)
		d = dev_get_msi_domain(&dev->bus->dev);

	dev_set_msi_domain(&dev->dev, d);
1834 1835
}

1836 1837 1838 1839 1840
/**
 * pci_dma_configure - Setup DMA configuration
 * @dev: ptr to pci_dev struct of the PCI device
 *
 * Function to update PCI devices's DMA configuration using the same
1841
 * info from the OF node or ACPI node of host bridge's parent (if any).
1842 1843 1844 1845 1846
 */
static void pci_dma_configure(struct pci_dev *dev)
{
	struct device *bridge = pci_get_host_bridge_device(dev);

1847 1848
	if (IS_ENABLED(CONFIG_OF) &&
		bridge->parent && bridge->parent->of_node) {
1849
			of_dma_configure(&dev->dev, bridge->parent->of_node);
1850 1851 1852 1853 1854 1855 1856 1857 1858
	} else if (has_acpi_companion(bridge)) {
		struct acpi_device *adev = to_acpi_device_node(bridge->fwnode);
		enum dev_dma_attr attr = acpi_get_dma_attr(adev);

		if (attr == DEV_DMA_NOT_SUPPORTED)
			dev_warn(&dev->dev, "DMA not supported.\n");
		else
			arch_setup_dma_ops(&dev->dev, 0, 0, NULL,
					   attr == DEV_DMA_COHERENT);
1859 1860 1861 1862 1863
	}

	pci_put_host_bridge_device(bridge);
}

1864
void pci_device_add(struct pci_dev *dev, struct pci_bus *bus)
L
Linus Torvalds 已提交
1865
{
1866 1867
	int ret;

1868 1869
	pci_configure_device(dev);

1870 1871
	device_initialize(&dev->dev);
	dev->dev.release = pci_release_dev;
L
Linus Torvalds 已提交
1872

1873
	set_dev_node(&dev->dev, pcibus_to_node(bus));
1874
	dev->dev.dma_mask = &dev->dma_mask;
1875
	dev->dev.dma_parms = &dev->dma_parms;
1876
	dev->dev.coherent_dma_mask = 0xffffffffull;
1877
	pci_dma_configure(dev);
L
Linus Torvalds 已提交
1878

1879
	pci_set_dma_max_seg_size(dev, 65536);
1880
	pci_set_dma_seg_boundary(dev, 0xffffffff);
1881

L
Linus Torvalds 已提交
1882 1883 1884
	/* Fix up broken headers */
	pci_fixup_device(pci_fixup_header, dev);

1885 1886 1887
	/* moved out from quirk header fixup code */
	pci_reassigndev_resource_alignment(dev);

1888 1889 1890
	/* Clear the state_saved flag. */
	dev->state_saved = false;

1891 1892
	/* Initialize various capabilities */
	pci_init_capabilities(dev);
1893

L
Linus Torvalds 已提交
1894 1895 1896 1897
	/*
	 * Add the device to our list of discovered devices
	 * and the bus list for fixup functions, etc.
	 */
1898
	down_write(&pci_bus_sem);
L
Linus Torvalds 已提交
1899
	list_add_tail(&dev->bus_list, &bus->devices);
1900
	up_write(&pci_bus_sem);
1901 1902 1903 1904

	ret = pcibios_add_device(dev);
	WARN_ON(ret < 0);

1905 1906 1907
	/* Setup MSI irq domain */
	pci_set_msi_domain(dev);

1908 1909 1910 1911
	/* Notifier could use PCI capabilities */
	dev->match_driver = false;
	ret = device_add(&dev->dev);
	WARN_ON(ret < 0);
1912 1913
}

1914
struct pci_dev *pci_scan_single_device(struct pci_bus *bus, int devfn)
1915 1916 1917
{
	struct pci_dev *dev;

T
Trent Piepho 已提交
1918 1919 1920 1921 1922 1923
	dev = pci_get_slot(bus, devfn);
	if (dev) {
		pci_dev_put(dev);
		return dev;
	}

1924 1925 1926 1927 1928
	dev = pci_scan_device(bus, devfn);
	if (!dev)
		return NULL;

	pci_device_add(dev, bus);
L
Linus Torvalds 已提交
1929 1930 1931

	return dev;
}
1932
EXPORT_SYMBOL(pci_scan_single_device);
L
Linus Torvalds 已提交
1933

1934
static unsigned next_fn(struct pci_bus *bus, struct pci_dev *dev, unsigned fn)
M
Matthew Wilcox 已提交
1935
{
1936 1937 1938
	int pos;
	u16 cap = 0;
	unsigned next_fn;
1939

1940 1941 1942 1943 1944 1945
	if (pci_ari_enabled(bus)) {
		if (!dev)
			return 0;
		pos = pci_find_ext_capability(dev, PCI_EXT_CAP_ID_ARI);
		if (!pos)
			return 0;
1946

1947 1948 1949 1950
		pci_read_config_word(dev, pos + PCI_ARI_CAP, &cap);
		next_fn = PCI_ARI_CAP_NFN(cap);
		if (next_fn <= fn)
			return 0;	/* protect against malformed list */
M
Matthew Wilcox 已提交
1951

1952 1953 1954 1955 1956 1957
		return next_fn;
	}

	/* dev may be NULL for non-contiguous multifunction devices */
	if (!dev || dev->multifunction)
		return (fn + 1) % 8;
M
Matthew Wilcox 已提交
1958 1959 1960 1961 1962 1963 1964

	return 0;
}

static int only_one_child(struct pci_bus *bus)
{
	struct pci_dev *parent = bus->self;
1965

M
Matthew Wilcox 已提交
1966 1967
	if (!parent || !pci_is_pcie(parent))
		return 0;
1968
	if (pci_pcie_type(parent) == PCI_EXP_TYPE_ROOT_PORT)
1969
		return 1;
1970 1971 1972 1973 1974 1975 1976

	/*
	 * PCIe downstream ports are bridges that normally lead to only a
	 * device 0, but if PCI_SCAN_ALL_PCIE_DEVS is set, scan all
	 * possible devices, not just device 0.  See PCIe spec r3.0,
	 * sec 7.3.1.
	 */
1977
	if (parent->has_secondary_link &&
1978
	    !pci_has_flag(PCI_SCAN_ALL_PCIE_DEVS))
M
Matthew Wilcox 已提交
1979 1980 1981 1982
		return 1;
	return 0;
}

L
Linus Torvalds 已提交
1983 1984 1985 1986 1987 1988 1989
/**
 * pci_scan_slot - scan a PCI slot on a bus for devices.
 * @bus: PCI bus to scan
 * @devfn: slot number to scan (must have zero function.)
 *
 * Scan a PCI slot on the specified PCI bus for devices, adding
 * discovered devices to the @bus->devices list.  New devices
1990
 * will not have is_added set.
1991 1992
 *
 * Returns the number of new devices found.
L
Linus Torvalds 已提交
1993
 */
1994
int pci_scan_slot(struct pci_bus *bus, int devfn)
L
Linus Torvalds 已提交
1995
{
M
Matthew Wilcox 已提交
1996
	unsigned fn, nr = 0;
1997
	struct pci_dev *dev;
M
Matthew Wilcox 已提交
1998 1999 2000

	if (only_one_child(bus) && (devfn > 0))
		return 0; /* Already scanned the entire slot */
L
Linus Torvalds 已提交
2001

2002
	dev = pci_scan_single_device(bus, devfn);
2003 2004 2005
	if (!dev)
		return 0;
	if (!dev->is_added)
2006 2007
		nr++;

2008
	for (fn = next_fn(bus, dev, 0); fn > 0; fn = next_fn(bus, dev, fn)) {
M
Matthew Wilcox 已提交
2009 2010 2011 2012 2013
		dev = pci_scan_single_device(bus, devfn + fn);
		if (dev) {
			if (!dev->is_added)
				nr++;
			dev->multifunction = 1;
L
Linus Torvalds 已提交
2014 2015
		}
	}
S
Shaohua Li 已提交
2016

2017 2018
	/* only one slot has pcie device */
	if (bus->self && nr)
S
Shaohua Li 已提交
2019 2020
		pcie_aspm_init_link_state(bus->self);

L
Linus Torvalds 已提交
2021 2022
	return nr;
}
2023
EXPORT_SYMBOL(pci_scan_slot);
L
Linus Torvalds 已提交
2024

2025 2026 2027 2028 2029 2030 2031
static int pcie_find_smpss(struct pci_dev *dev, void *data)
{
	u8 *smpss = data;

	if (!pci_is_pcie(dev))
		return 0;

2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045
	/*
	 * We don't have a way to change MPS settings on devices that have
	 * drivers attached.  A hot-added device might support only the minimum
	 * MPS setting (MPS=128).  Therefore, if the fabric contains a bridge
	 * where devices may be hot-added, we limit the fabric MPS to 128 so
	 * hot-added devices will work correctly.
	 *
	 * However, if we hot-add a device to a slot directly below a Root
	 * Port, it's impossible for there to be other existing devices below
	 * the port.  We don't limit the MPS in this case because we can
	 * reconfigure MPS on both the Root Port and the hot-added device,
	 * and there are no other devices involved.
	 *
	 * Note that this PCIE_BUS_SAFE path assumes no peer-to-peer DMA.
2046
	 */
2047 2048
	if (dev->is_hotplug_bridge &&
	    pci_pcie_type(dev) != PCI_EXP_TYPE_ROOT_PORT)
2049 2050 2051 2052 2053 2054 2055 2056 2057 2058
		*smpss = 0;

	if (*smpss > dev->pcie_mpss)
		*smpss = dev->pcie_mpss;

	return 0;
}

static void pcie_write_mps(struct pci_dev *dev, int mps)
{
2059
	int rc;
2060 2061

	if (pcie_bus_config == PCIE_BUS_PERFORMANCE) {
2062
		mps = 128 << dev->pcie_mpss;
2063

2064 2065
		if (pci_pcie_type(dev) != PCI_EXP_TYPE_ROOT_PORT &&
		    dev->bus->self)
2066
			/* For "Performance", the assumption is made that
2067 2068 2069 2070 2071
			 * downstream communication will never be larger than
			 * the MRRS.  So, the MPS only needs to be configured
			 * for the upstream communication.  This being the case,
			 * walk from the top down and set the MPS of the child
			 * to that of the parent bus.
2072 2073 2074 2075 2076
			 *
			 * Configure the device MPS with the smaller of the
			 * device MPSS or the bridge MPS (which is assumed to be
			 * properly configured at this point to the largest
			 * allowable MPS based on its parent bus).
2077
			 */
2078
			mps = min(mps, pcie_get_mps(dev->bus->self));
2079 2080 2081 2082 2083 2084 2085
	}

	rc = pcie_set_mps(dev, mps);
	if (rc)
		dev_err(&dev->dev, "Failed attempting to set the MPS\n");
}

2086
static void pcie_write_mrrs(struct pci_dev *dev)
2087
{
2088
	int rc, mrrs;
2089

2090 2091 2092 2093 2094 2095 2096 2097
	/* In the "safe" case, do not configure the MRRS.  There appear to be
	 * issues with setting MRRS to 0 on a number of devices.
	 */
	if (pcie_bus_config != PCIE_BUS_PERFORMANCE)
		return;

	/* For Max performance, the MRRS must be set to the largest supported
	 * value.  However, it cannot be configured larger than the MPS the
2098 2099
	 * device or the bus can support.  This should already be properly
	 * configured by a prior call to pcie_write_mps.
2100
	 */
2101
	mrrs = pcie_get_mps(dev);
2102 2103

	/* MRRS is a R/W register.  Invalid values can be written, but a
2104
	 * subsequent read will verify if the value is acceptable or not.
2105 2106
	 * If the MRRS value provided is not acceptable (e.g., too large),
	 * shrink the value until it is acceptable to the HW.
2107
	 */
2108 2109
	while (mrrs != pcie_get_readrq(dev) && mrrs >= 128) {
		rc = pcie_set_readrq(dev, mrrs);
2110 2111
		if (!rc)
			break;
2112

2113
		dev_warn(&dev->dev, "Failed attempting to set the MRRS\n");
2114 2115
		mrrs /= 2;
	}
2116 2117

	if (mrrs < 128)
2118
		dev_err(&dev->dev, "MRRS was unable to be configured with a safe value.  If problems are experienced, try running with pci=pcie_bus_safe\n");
2119 2120 2121 2122
}

static int pcie_bus_configure_set(struct pci_dev *dev, void *data)
{
J
Jon Mason 已提交
2123
	int mps, orig_mps;
2124 2125 2126 2127

	if (!pci_is_pcie(dev))
		return 0;

2128 2129
	if (pcie_bus_config == PCIE_BUS_TUNE_OFF ||
	    pcie_bus_config == PCIE_BUS_DEFAULT)
2130 2131
		return 0;

J
Jon Mason 已提交
2132 2133
	mps = 128 << *(u8 *)data;
	orig_mps = pcie_get_mps(dev);
2134 2135

	pcie_write_mps(dev, mps);
2136
	pcie_write_mrrs(dev);
2137

2138 2139
	dev_info(&dev->dev, "Max Payload Size set to %4d/%4d (was %4d), Max Read Rq %4d\n",
		 pcie_get_mps(dev), 128 << dev->pcie_mpss,
J
Jon Mason 已提交
2140
		 orig_mps, pcie_get_readrq(dev));
2141 2142 2143 2144

	return 0;
}

J
Jon Mason 已提交
2145
/* pcie_bus_configure_settings requires that pci_walk_bus work in a top-down,
2146 2147 2148
 * parents then children fashion.  If this changes, then this code will not
 * work as designed.
 */
2149
void pcie_bus_configure_settings(struct pci_bus *bus)
2150
{
2151
	u8 smpss = 0;
2152

2153
	if (!bus->self)
2154 2155 2156
		return;

	if (!pci_is_pcie(bus->self))
2157 2158 2159
		return;

	/* FIXME - Peer to peer DMA is possible, though the endpoint would need
2160
	 * to be aware of the MPS of the destination.  To work around this,
2161 2162 2163 2164 2165
	 * simply force the MPS of the entire system to the smallest possible.
	 */
	if (pcie_bus_config == PCIE_BUS_PEER2PEER)
		smpss = 0;

2166
	if (pcie_bus_config == PCIE_BUS_SAFE) {
2167
		smpss = bus->self->pcie_mpss;
2168

2169 2170 2171 2172 2173 2174 2175
		pcie_find_smpss(bus->self, &smpss);
		pci_walk_bus(bus, pcie_find_smpss, &smpss);
	}

	pcie_bus_configure_set(bus->self, &smpss);
	pci_walk_bus(bus, pcie_bus_configure_set, &smpss);
}
2176
EXPORT_SYMBOL_GPL(pcie_bus_configure_settings);
2177

B
Bill Pemberton 已提交
2178
unsigned int pci_scan_child_bus(struct pci_bus *bus)
L
Linus Torvalds 已提交
2179
{
2180
	unsigned int devfn, pass, max = bus->busn_res.start;
L
Linus Torvalds 已提交
2181 2182
	struct pci_dev *dev;

B
Bjorn Helgaas 已提交
2183
	dev_dbg(&bus->dev, "scanning bus\n");
L
Linus Torvalds 已提交
2184 2185 2186 2187 2188

	/* Go find them, Rover! */
	for (devfn = 0; devfn < 0x100; devfn += 8)
		pci_scan_slot(bus, devfn);

2189 2190 2191
	/* Reserve buses for SR-IOV capability. */
	max += pci_iov_bus_range(bus);

L
Linus Torvalds 已提交
2192 2193 2194 2195
	/*
	 * After performing arch-dependent fixup of the bus, look behind
	 * all PCI-to-PCI bridges on this bus.
	 */
A
Alex Chiang 已提交
2196
	if (!bus->is_added) {
B
Bjorn Helgaas 已提交
2197
		dev_dbg(&bus->dev, "fixups for bus\n");
A
Alex Chiang 已提交
2198
		pcibios_fixup_bus(bus);
2199
		bus->is_added = 1;
A
Alex Chiang 已提交
2200 2201
	}

R
Ryan Desfosses 已提交
2202
	for (pass = 0; pass < 2; pass++)
L
Linus Torvalds 已提交
2203
		list_for_each_entry(dev, &bus->devices, bus_list) {
2204
			if (pci_is_bridge(dev))
L
Linus Torvalds 已提交
2205 2206 2207
				max = pci_scan_bridge(bus, dev, max, pass);
		}

2208 2209 2210 2211 2212 2213 2214 2215 2216
	/*
	 * Make sure a hotplug bridge has at least the minimum requested
	 * number of buses.
	 */
	if (bus->self && bus->self->is_hotplug_bridge && pci_hotplug_bus_size) {
		if (max - bus->busn_res.start < pci_hotplug_bus_size - 1)
			max = bus->busn_res.start + pci_hotplug_bus_size - 1;
	}

L
Linus Torvalds 已提交
2217 2218 2219 2220 2221 2222 2223
	/*
	 * We've scanned the bus and so we know all about what's on
	 * the other side of any bridges that may be on this bus plus
	 * any devices.
	 *
	 * Return how far we've got finding sub-buses.
	 */
B
Bjorn Helgaas 已提交
2224
	dev_dbg(&bus->dev, "bus scan returning with max=%02x\n", max);
L
Linus Torvalds 已提交
2225 2226
	return max;
}
2227
EXPORT_SYMBOL_GPL(pci_scan_child_bus);
L
Linus Torvalds 已提交
2228

2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240
/**
 * pcibios_root_bridge_prepare - Platform-specific host bridge setup.
 * @bridge: Host bridge to set up.
 *
 * Default empty implementation.  Replace with an architecture-specific setup
 * routine, if necessary.
 */
int __weak pcibios_root_bridge_prepare(struct pci_host_bridge *bridge)
{
	return 0;
}

2241 2242 2243 2244 2245 2246 2247 2248
void __weak pcibios_add_bus(struct pci_bus *bus)
{
}

void __weak pcibios_remove_bus(struct pci_bus *bus)
{
}

2249 2250 2251
static struct pci_bus *pci_create_root_bus_msi(struct device *parent,
		int bus, struct pci_ops *ops, void *sysdata,
		struct list_head *resources, struct msi_controller *msi)
L
Linus Torvalds 已提交
2252
{
2253
	int error;
2254
	struct pci_host_bridge *bridge;
L
Linus Torvalds 已提交
2255

2256
	bridge = pci_alloc_host_bridge();
2257
	if (!bridge)
2258
		return NULL;
2259 2260

	bridge->dev.parent = parent;
2261
	bridge->dev.release = pci_release_host_bridge_dev;
2262

2263 2264 2265 2266 2267
	list_splice_init(resources, &bridge->windows);
	bridge->sysdata = sysdata;
	bridge->busnr = bus;
	bridge->ops = ops;
	bridge->msi = msi;
2268

2269 2270 2271
	error = pci_register_host_bridge(bridge);
	if (error < 0)
		goto err_out;
2272

2273
	return bridge->bus;
L
Linus Torvalds 已提交
2274 2275

err_out:
2276
	kfree(bridge);
L
Linus Torvalds 已提交
2277 2278
	return NULL;
}
2279 2280 2281 2282 2283 2284 2285

struct pci_bus *pci_create_root_bus(struct device *parent, int bus,
		struct pci_ops *ops, void *sysdata, struct list_head *resources)
{
	return pci_create_root_bus_msi(parent, bus, ops, sysdata, resources,
				       NULL);
}
2286
EXPORT_SYMBOL_GPL(pci_create_root_bus);
2287

2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303
int pci_bus_insert_busn_res(struct pci_bus *b, int bus, int bus_max)
{
	struct resource *res = &b->busn_res;
	struct resource *parent_res, *conflict;

	res->start = bus;
	res->end = bus_max;
	res->flags = IORESOURCE_BUS;

	if (!pci_is_root_bus(b))
		parent_res = &b->parent->busn_res;
	else {
		parent_res = get_pci_domain_busn_res(pci_domain_nr(b));
		res->flags |= IORESOURCE_PCI_FIXED;
	}

2304
	conflict = request_resource_conflict(parent_res, res);
2305 2306 2307 2308 2309 2310 2311 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

	if (conflict)
		dev_printk(KERN_DEBUG, &b->dev,
			   "busn_res: can not insert %pR under %s%pR (conflicts with %s %pR)\n",
			    res, pci_is_root_bus(b) ? "domain " : "",
			    parent_res, conflict->name, conflict);

	return conflict == NULL;
}

int pci_bus_update_busn_res_end(struct pci_bus *b, int bus_max)
{
	struct resource *res = &b->busn_res;
	struct resource old_res = *res;
	resource_size_t size;
	int ret;

	if (res->start > bus_max)
		return -EINVAL;

	size = bus_max - res->start + 1;
	ret = adjust_resource(res, res->start, size);
	dev_printk(KERN_DEBUG, &b->dev,
			"busn_res: %pR end %s updated to %02x\n",
			&old_res, ret ? "can not be" : "is", bus_max);

	if (!ret && !res->parent)
		pci_bus_insert_busn_res(b, res->start, res->end);

	return ret;
}

void pci_bus_release_busn_res(struct pci_bus *b)
{
	struct resource *res = &b->busn_res;
	int ret;

	if (!res->flags || !res->parent)
		return;

	ret = release_resource(res);
	dev_printk(KERN_DEBUG, &b->dev,
			"busn_res: %pR %s released\n",
			res, ret ? "can not be" : "is");
}

2351 2352 2353
struct pci_bus *pci_scan_root_bus_msi(struct device *parent, int bus,
		struct pci_ops *ops, void *sysdata,
		struct list_head *resources, struct msi_controller *msi)
2354
{
2355
	struct resource_entry *window;
2356
	bool found = false;
2357
	struct pci_bus *b;
2358 2359
	int max;

2360
	resource_list_for_each_entry(window, resources)
2361 2362 2363 2364
		if (window->res->flags & IORESOURCE_BUS) {
			found = true;
			break;
		}
2365

2366
	b = pci_create_root_bus_msi(parent, bus, ops, sysdata, resources, msi);
2367 2368 2369
	if (!b)
		return NULL;

2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381
	if (!found) {
		dev_info(&b->dev,
		 "No busn resource found for root bus, will use [bus %02x-ff]\n",
			bus);
		pci_bus_insert_busn_res(b, bus, 255);
	}

	max = pci_scan_child_bus(b);

	if (!found)
		pci_bus_update_busn_res_end(b, max);

2382 2383
	return b;
}
2384 2385 2386 2387 2388 2389 2390

struct pci_bus *pci_scan_root_bus(struct device *parent, int bus,
		struct pci_ops *ops, void *sysdata, struct list_head *resources)
{
	return pci_scan_root_bus_msi(parent, bus, ops, sysdata, resources,
				     NULL);
}
2391 2392
EXPORT_SYMBOL(pci_scan_root_bus);

B
Bill Pemberton 已提交
2393
struct pci_bus *pci_scan_bus(int bus, struct pci_ops *ops,
2394 2395 2396 2397 2398 2399 2400
					void *sysdata)
{
	LIST_HEAD(resources);
	struct pci_bus *b;

	pci_add_resource(&resources, &ioport_resource);
	pci_add_resource(&resources, &iomem_resource);
2401
	pci_add_resource(&resources, &busn_resource);
2402 2403
	b = pci_create_root_bus(NULL, bus, ops, sysdata, &resources);
	if (b) {
2404
		pci_scan_child_bus(b);
2405 2406 2407 2408 2409 2410 2411
	} else {
		pci_free_resource_list(&resources);
	}
	return b;
}
EXPORT_SYMBOL(pci_scan_bus);

2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422
/**
 * pci_rescan_bus_bridge_resize - scan a PCI bus for devices.
 * @bridge: PCI bridge for the bus to scan
 *
 * Scan a PCI bus and child buses for new devices, add them,
 * and enable them, resizing bridge mmio/io resource if necessary
 * and possible.  The caller must ensure the child devices are already
 * removed for resizing to occur.
 *
 * Returns the max number of subordinate bus discovered.
 */
2423
unsigned int pci_rescan_bus_bridge_resize(struct pci_dev *bridge)
2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436
{
	unsigned int max;
	struct pci_bus *bus = bridge->subordinate;

	max = pci_scan_child_bus(bus);

	pci_assign_unassigned_bridge_resources(bridge);

	pci_bus_add_devices(bus);

	return max;
}

2437 2438 2439 2440 2441 2442 2443 2444 2445
/**
 * pci_rescan_bus - scan a PCI bus for devices.
 * @bus: PCI bus to scan
 *
 * Scan a PCI bus and child buses for new devices, adds them,
 * and enables them.
 *
 * Returns the max number of subordinate bus discovered.
 */
2446
unsigned int pci_rescan_bus(struct pci_bus *bus)
2447 2448 2449 2450 2451 2452 2453 2454 2455 2456 2457
{
	unsigned int max;

	max = pci_scan_child_bus(bus);
	pci_assign_unassigned_bus_resources(bus);
	pci_bus_add_devices(bus);

	return max;
}
EXPORT_SYMBOL_GPL(pci_rescan_bus);

2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475
/*
 * pci_rescan_bus(), pci_rescan_bus_bridge_resize() and PCI device removal
 * routines should always be executed under this mutex.
 */
static DEFINE_MUTEX(pci_rescan_remove_lock);

void pci_lock_rescan_remove(void)
{
	mutex_lock(&pci_rescan_remove_lock);
}
EXPORT_SYMBOL_GPL(pci_lock_rescan_remove);

void pci_unlock_rescan_remove(void)
{
	mutex_unlock(&pci_rescan_remove_lock);
}
EXPORT_SYMBOL_GPL(pci_unlock_rescan_remove);

R
Ryan Desfosses 已提交
2476 2477
static int __init pci_sort_bf_cmp(const struct device *d_a,
				  const struct device *d_b)
2478
{
2479 2480 2481
	const struct pci_dev *a = to_pci_dev(d_a);
	const struct pci_dev *b = to_pci_dev(d_b);

2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 2493
	if      (pci_domain_nr(a->bus) < pci_domain_nr(b->bus)) return -1;
	else if (pci_domain_nr(a->bus) > pci_domain_nr(b->bus)) return  1;

	if      (a->bus->number < b->bus->number) return -1;
	else if (a->bus->number > b->bus->number) return  1;

	if      (a->devfn < b->devfn) return -1;
	else if (a->devfn > b->devfn) return  1;

	return 0;
}

2494
void __init pci_sort_breadthfirst(void)
2495
{
2496
	bus_sort_breadthfirst(&pci_bus_type, &pci_sort_bf_cmp);
2497
}