setup.c 28.5 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
/*
 *  Copyright (C) 1995  Linus Torvalds
 *
 *  Support of BIGMEM added by Gerhard Wichert, Siemens AG, July 1999
 *
 *  Memory region support
 *	David Parsons <orc@pell.chi.il.us>, July-August 1999
 *
 *  Added E820 sanitization routine (removes overlapping memory regions);
 *  Brian Moyle <bmoyle@mvista.com>, February 2001
 *
 * Moved CPU detection code to cpu/${cpu}.c
 *    Patrick Mochel <mochel@osdl.org>, March 2002
 *
 *  Provisions for empty E820 memory regions (reported by certain BIOSes).
 *  Alex Achenbach <xela@slit.de>, December 2002.
 *
 */

/*
 * This file handles the architecture-dependent parts of initialization
 */

#include <linux/sched.h>
#include <linux/mm.h>
26
#include <linux/mmzone.h>
27
#include <linux/screen_info.h>
L
Linus Torvalds 已提交
28 29
#include <linux/ioport.h>
#include <linux/acpi.h>
F
Feng Tang 已提交
30
#include <linux/sfi.h>
L
Linus Torvalds 已提交
31 32 33
#include <linux/apm_bios.h>
#include <linux/initrd.h>
#include <linux/bootmem.h>
34
#include <linux/memblock.h>
L
Linus Torvalds 已提交
35 36 37 38 39 40 41 42
#include <linux/seq_file.h>
#include <linux/console.h>
#include <linux/root_dev.h>
#include <linux/highmem.h>
#include <linux/module.h>
#include <linux/efi.h>
#include <linux/init.h>
#include <linux/edd.h>
43
#include <linux/iscsi_ibft.h>
L
Linus Torvalds 已提交
44
#include <linux/nodemask.h>
45
#include <linux/kexec.h>
46
#include <linux/dmi.h>
D
Dave Hansen 已提交
47
#include <linux/pfn.h>
48
#include <linux/pci.h>
49
#include <asm/pci-direct.h>
50
#include <linux/init_ohci1394_dma.h>
51
#include <linux/kvm_para.h>
52
#include <linux/dma-contiguous.h>
53

54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69
#include <linux/errno.h>
#include <linux/kernel.h>
#include <linux/stddef.h>
#include <linux/unistd.h>
#include <linux/ptrace.h>
#include <linux/user.h>
#include <linux/delay.h>

#include <linux/kallsyms.h>
#include <linux/cpufreq.h>
#include <linux/dma-mapping.h>
#include <linux/ctype.h>
#include <linux/uaccess.h>

#include <linux/percpu.h>
#include <linux/crash_dump.h>
70
#include <linux/tboot.h>
71
#include <linux/jiffies.h>
72

L
Linus Torvalds 已提交
73
#include <video/edid.h>
74

75
#include <asm/mtrr.h>
76
#include <asm/apic.h>
77
#include <asm/realmode.h>
L
Linus Torvalds 已提交
78 79 80
#include <asm/e820.h>
#include <asm/mpspec.h>
#include <asm/setup.h>
Y
Yinghai Lu 已提交
81
#include <asm/efi.h>
I
Ingo Molnar 已提交
82 83
#include <asm/timer.h>
#include <asm/i8259.h>
L
Linus Torvalds 已提交
84
#include <asm/sections.h>
85
#include <asm/dmi.h>
L
Linus Torvalds 已提交
86 87
#include <asm/io_apic.h>
#include <asm/ist.h>
88
#include <asm/setup_arch.h>
89
#include <asm/bios_ebda.h>
B
Bernhard Walle 已提交
90
#include <asm/cacheflush.h>
91
#include <asm/processor.h>
Y
Yinghai Lu 已提交
92
#include <asm/bugs.h>
L
Linus Torvalds 已提交
93

94
#include <asm/vsyscall.h>
95
#include <asm/cpu.h>
96 97
#include <asm/desc.h>
#include <asm/dma.h>
98
#include <asm/iommu.h>
99
#include <asm/gart.h>
100 101 102 103
#include <asm/mmu_context.h>
#include <asm/proto.h>

#include <asm/paravirt.h>
104
#include <asm/hypervisor.h>
105
#include <asm/olpc_ofw.h>
106 107 108 109

#include <asm/percpu.h>
#include <asm/topology.h>
#include <asm/apicdef.h>
110
#include <asm/amd_nb.h>
111
#include <asm/mce.h>
112
#include <asm/alternative.h>
113
#include <asm/prom.h>
114

115
/*
116 117 118 119 120
 * max_low_pfn_mapped: highest direct mapped pfn under 4GB
 * max_pfn_mapped:     highest direct mapped pfn over 4GB
 *
 * The direct mapping only covers E820_RAM regions, so the ranges and gaps are
 * represented by pfn_mapped
121 122 123 124
 */
unsigned long max_low_pfn_mapped;
unsigned long max_pfn_mapped;

125
#ifdef CONFIG_DMI
126
RESERVE_BRK(dmi_alloc, 65536);
127
#endif
128

I
Ingo Molnar 已提交
129

130 131 132
static __initdata unsigned long _brk_start = (unsigned long)__brk_base;
unsigned long _brk_end = (unsigned long)__brk_base;

I
Ingo Molnar 已提交
133 134 135 136 137 138
#ifdef CONFIG_X86_64
int default_cpu_present_to_apicid(int mps_cpu)
{
	return __default_cpu_present_to_apicid(mps_cpu);
}

T
Thomas Gleixner 已提交
139
int default_check_phys_apicid_present(int phys_apicid)
I
Ingo Molnar 已提交
140
{
T
Thomas Gleixner 已提交
141
	return __default_check_phys_apicid_present(phys_apicid);
I
Ingo Molnar 已提交
142 143 144
}
#endif

145 146
struct boot_params boot_params;

L
Linus Torvalds 已提交
147 148 149
/*
 * Machine setup..
 */
150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170
static struct resource data_resource = {
	.name	= "Kernel data",
	.start	= 0,
	.end	= 0,
	.flags	= IORESOURCE_BUSY | IORESOURCE_MEM
};

static struct resource code_resource = {
	.name	= "Kernel code",
	.start	= 0,
	.end	= 0,
	.flags	= IORESOURCE_BUSY | IORESOURCE_MEM
};

static struct resource bss_resource = {
	.name	= "Kernel bss",
	.start	= 0,
	.end	= 0,
	.flags	= IORESOURCE_BUSY | IORESOURCE_MEM
};

171 172

#ifdef CONFIG_X86_32
L
Linus Torvalds 已提交
173
/* cpu data as detected by the assembly code in head.S */
174
struct cpuinfo_x86 new_cpu_data __cpuinitdata = {0, 0, 0, 0, -1, 1, 0, 0, -1};
L
Linus Torvalds 已提交
175
/* common cpu data for all cpus */
176
struct cpuinfo_x86 boot_cpu_data __read_mostly = {0, 0, 0, 0, -1, 1, 0, 0, -1};
177
EXPORT_SYMBOL(boot_cpu_data);
L
Linus Torvalds 已提交
178

179 180
unsigned int def_to_bigsmp;

L
Linus Torvalds 已提交
181 182 183 184 185
/* for MCA, but anyone else can use it if they want */
unsigned int machine_id;
unsigned int machine_submodel_id;
unsigned int BIOS_revision;

186 187 188 189 190 191 192 193 194 195 196 197
struct apm_info apm_info;
EXPORT_SYMBOL(apm_info);

#if defined(CONFIG_X86_SPEEDSTEP_SMI) || \
	defined(CONFIG_X86_SPEEDSTEP_SMI_MODULE)
struct ist_info ist_info;
EXPORT_SYMBOL(ist_info);
#else
struct ist_info ist_info;
#endif

#else
198 199 200
struct cpuinfo_x86 boot_cpu_data __read_mostly = {
	.x86_phys_bits = MAX_PHYSMEM_BITS,
};
201 202 203 204 205 206 207 208 209 210
EXPORT_SYMBOL(boot_cpu_data);
#endif


#if !defined(CONFIG_X86_PAE) || defined(CONFIG_X86_64)
unsigned long mmu_cr4_features;
#else
unsigned long mmu_cr4_features = X86_CR4_PAE;
#endif

211 212
/* Boot loader ID and version as integers, for the benefit of proc_dointvec */
int bootloader_type, bootloader_version;
L
Linus Torvalds 已提交
213 214 215 216 217

/*
 * Setup options
 */
struct screen_info screen_info;
218
EXPORT_SYMBOL(screen_info);
L
Linus Torvalds 已提交
219
struct edid_info edid_info;
220
EXPORT_SYMBOL_GPL(edid_info);
L
Linus Torvalds 已提交
221 222 223

extern int root_mountflags;

224
unsigned long saved_video_mode;
L
Linus Torvalds 已提交
225

226
#define RAMDISK_IMAGE_START_MASK	0x07FF
L
Linus Torvalds 已提交
227
#define RAMDISK_PROMPT_FLAG		0x8000
228
#define RAMDISK_LOAD_FLAG		0x4000
L
Linus Torvalds 已提交
229

230
static char __initdata command_line[COMMAND_LINE_SIZE];
231 232 233
#ifdef CONFIG_CMDLINE_BOOL
static char __initdata builtin_cmdline[COMMAND_LINE_SIZE] = CONFIG_CMDLINE;
#endif
L
Linus Torvalds 已提交
234 235 236 237 238 239 240 241 242 243 244

#if defined(CONFIG_EDD) || defined(CONFIG_EDD_MODULE)
struct edd edd;
#ifdef CONFIG_EDD_MODULE
EXPORT_SYMBOL(edd);
#endif
/**
 * copy_edd() - Copy the BIOS EDD information
 *              from boot_params into a safe place.
 *
 */
245
static inline void __init copy_edd(void)
L
Linus Torvalds 已提交
246
{
247 248 249 250 251
     memcpy(edd.mbr_signature, boot_params.edd_mbr_sig_buffer,
	    sizeof(edd.mbr_signature));
     memcpy(edd.edd_info, boot_params.eddbuf, sizeof(edd.edd_info));
     edd.mbr_signature_nr = boot_params.edd_mbr_sig_buf_entries;
     edd.edd_info_nr = boot_params.eddbuf_entries;
L
Linus Torvalds 已提交
252 253
}
#else
254
static inline void __init copy_edd(void)
L
Linus Torvalds 已提交
255 256 257 258
{
}
#endif

259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277
void * __init extend_brk(size_t size, size_t align)
{
	size_t mask = align - 1;
	void *ret;

	BUG_ON(_brk_start == 0);
	BUG_ON(align & mask);

	_brk_end = (_brk_end + mask) & ~mask;
	BUG_ON((char *)(_brk_end + size) > __brk_limit);

	ret = (void *)_brk_end;
	_brk_end += size;

	memset(ret, 0, size);

	return ret;
}

278
#ifdef CONFIG_X86_32
279
static void __init cleanup_highmap(void)
280 281
{
}
282 283
#endif

284 285 286
static void __init reserve_brk(void)
{
	if (_brk_end > _brk_start)
287 288
		memblock_reserve(__pa(_brk_start),
				 __pa(_brk_end) - __pa(_brk_start));
289 290 291 292 293 294

	/* Mark brk area as locked down and no longer taking any
	   new allocations */
	_brk_start = 0;
}

295 296
#ifdef CONFIG_BLK_DEV_INITRD

Y
Yinghai Lu 已提交
297 298 299 300 301 302 303 304 305 306 307 308 309
static u64 __init get_ramdisk_image(void)
{
	u64 ramdisk_image = boot_params.hdr.ramdisk_image;

	return ramdisk_image;
}
static u64 __init get_ramdisk_size(void)
{
	u64 ramdisk_size = boot_params.hdr.ramdisk_size;

	return ramdisk_size;
}

310 311
#define MAX_MAP_CHUNK	(NR_FIX_BTMAPS << PAGE_SHIFT)
static void __init relocate_initrd(void)
312
{
313
	/* Assume only end is not page aligned */
Y
Yinghai Lu 已提交
314 315
	u64 ramdisk_image = get_ramdisk_image();
	u64 ramdisk_size  = get_ramdisk_size();
316
	u64 area_size     = PAGE_ALIGN(ramdisk_size);
317
	u64 ramdisk_here;
318 319
	unsigned long slop, clen, mapaddr;
	char *p, *q;
320

321
	/* We need to move the initrd down into directly mapped mem */
322
	ramdisk_here = memblock_find_in_range(0, PFN_PHYS(max_pfn_mapped),
323
						 area_size, PAGE_SIZE);
324

T
Tejun Heo 已提交
325
	if (!ramdisk_here)
326 327 328
		panic("Cannot find place for new RAMDISK of size %lld\n",
			 ramdisk_size);

329
	/* Note: this includes all the mem currently occupied by
330
	   the initrd, we rely on that fact to keep the data intact. */
331
	memblock_reserve(ramdisk_here, area_size);
332 333
	initrd_start = ramdisk_here + PAGE_OFFSET;
	initrd_end   = initrd_start + ramdisk_size;
334 335
	printk(KERN_INFO "Allocated new RAMDISK: [mem %#010llx-%#010llx]\n",
			 ramdisk_here, ramdisk_here + ramdisk_size - 1);
336 337 338

	q = (char *)initrd_start;

339
	/* Copy the initrd */
340 341 342 343 344 345
	while (ramdisk_size) {
		slop = ramdisk_image & ~PAGE_MASK;
		clen = ramdisk_size;
		if (clen > MAX_MAP_CHUNK-slop)
			clen = MAX_MAP_CHUNK-slop;
		mapaddr = ramdisk_image & PAGE_MASK;
346
		p = early_memremap(mapaddr, clen+slop);
347
		memcpy(q, p+slop, clen);
348
		early_iounmap(p, clen+slop);
349 350 351 352
		q += clen;
		ramdisk_image += clen;
		ramdisk_size  -= clen;
	}
353

Y
Yinghai Lu 已提交
354 355
	ramdisk_image = get_ramdisk_image();
	ramdisk_size  = get_ramdisk_size();
356 357
	printk(KERN_INFO "Move RAMDISK from [mem %#010llx-%#010llx] to"
		" [mem %#010llx-%#010llx]\n",
358 359
		ramdisk_image, ramdisk_image + ramdisk_size - 1,
		ramdisk_here, ramdisk_here + ramdisk_size - 1);
360
}
Y
Yinghai Lu 已提交
361

362 363 364 365 366 367 368 369 370 371 372 373 374 375
static u64 __init get_mem_size(unsigned long limit_pfn)
{
	int i;
	u64 mapped_pages = 0;
	unsigned long start_pfn, end_pfn;

	for_each_mem_pfn_range(i, MAX_NUMNODES, &start_pfn, &end_pfn, NULL) {
		start_pfn = min_t(unsigned long, start_pfn, limit_pfn);
		end_pfn = min_t(unsigned long, end_pfn, limit_pfn);
		mapped_pages += end_pfn - start_pfn;
	}

	return mapped_pages << PAGE_SHIFT;
}
376 377 378
static void __init early_reserve_initrd(void)
{
	/* Assume only end is not page aligned */
Y
Yinghai Lu 已提交
379 380
	u64 ramdisk_image = get_ramdisk_image();
	u64 ramdisk_size  = get_ramdisk_size();
381 382 383 384 385 386 387 388
	u64 ramdisk_end   = PAGE_ALIGN(ramdisk_image + ramdisk_size);

	if (!boot_params.hdr.type_of_loader ||
	    !ramdisk_image || !ramdisk_size)
		return;		/* No initrd provided by bootloader */

	memblock_reserve(ramdisk_image, ramdisk_end - ramdisk_image);
}
389 390
static void __init reserve_initrd(void)
{
391
	/* Assume only end is not page aligned */
Y
Yinghai Lu 已提交
392 393
	u64 ramdisk_image = get_ramdisk_image();
	u64 ramdisk_size  = get_ramdisk_size();
394
	u64 ramdisk_end   = PAGE_ALIGN(ramdisk_image + ramdisk_size);
395
	u64 mapped_size;
396 397 398 399 400 401 402

	if (!boot_params.hdr.type_of_loader ||
	    !ramdisk_image || !ramdisk_size)
		return;		/* No initrd provided by bootloader */

	initrd_start = 0;

403
	mapped_size = get_mem_size(max_pfn_mapped);
404
	if (ramdisk_size >= (mapped_size>>1))
405 406
		panic("initrd too large to handle, "
		       "disabling initrd (%lld needed, %lld available)\n",
407
		       ramdisk_size, mapped_size>>1);
408

409 410
	printk(KERN_INFO "RAMDISK: [mem %#010llx-%#010llx]\n", ramdisk_image,
			ramdisk_end - 1);
411

412
	if (pfn_range_is_mapped(PFN_DOWN(ramdisk_image),
413 414
				PFN_DOWN(ramdisk_end))) {
		/* All are mapped, easy case */
415 416 417 418 419 420
		initrd_start = ramdisk_image + PAGE_OFFSET;
		initrd_end = initrd_start + ramdisk_size;
		return;
	}

	relocate_initrd();
421

422
	memblock_free(ramdisk_image, ramdisk_end - ramdisk_image);
423
}
Y
Yinghai Lu 已提交
424
#else
425 426 427
static void __init early_reserve_initrd(void)
{
}
428
static void __init reserve_initrd(void)
Y
Yinghai Lu 已提交
429 430
{
}
431 432
#endif /* CONFIG_BLK_DEV_INITRD */

433
static void __init parse_setup_data(void)
434 435 436 437 438 439 440 441
{
	struct setup_data *data;
	u64 pa_data;

	if (boot_params.hdr.version < 0x0209)
		return;
	pa_data = boot_params.hdr.setup_data;
	while (pa_data) {
442 443 444 445 446 447 448 449 450 451 452 453
		u32 data_len, map_len;

		map_len = max(PAGE_SIZE - (pa_data & ~PAGE_MASK),
			      (u64)sizeof(struct setup_data));
		data = early_memremap(pa_data, map_len);
		data_len = data->len + sizeof(struct setup_data);
		if (data_len > map_len) {
			early_iounmap(data, map_len);
			data = early_memremap(pa_data, data_len);
			map_len = data_len;
		}

454 455
		switch (data->type) {
		case SETUP_E820_EXT:
456
			parse_e820_ext(data);
457
			break;
458 459
		case SETUP_DTB:
			add_dtb(pa_data);
460 461 462 463 464
			break;
		default:
			break;
		}
		pa_data = data->next;
465
		early_iounmap(data, map_len);
466 467 468
	}
}

469
static void __init e820_reserve_setup_data(void)
470 471 472
{
	struct setup_data *data;
	u64 pa_data;
473
	int found = 0;
474 475 476 477 478

	if (boot_params.hdr.version < 0x0209)
		return;
	pa_data = boot_params.hdr.setup_data;
	while (pa_data) {
479
		data = early_memremap(pa_data, sizeof(*data));
480 481
		e820_update_range(pa_data, sizeof(*data)+data->len,
			 E820_RAM, E820_RESERVED_KERN);
482
		found = 1;
483 484 485
		pa_data = data->next;
		early_iounmap(data, sizeof(*data));
	}
486 487 488
	if (!found)
		return;

489
	sanitize_e820_map(e820.map, ARRAY_SIZE(e820.map), &e820.nr_map);
490
	memcpy(&e820_saved, &e820, sizeof(struct e820map));
491 492 493 494
	printk(KERN_INFO "extended physical RAM map:\n");
	e820_print_map("reserve setup_data");
}

495
static void __init memblock_x86_reserve_range_setup_data(void)
496 497 498 499 500 501 502 503
{
	struct setup_data *data;
	u64 pa_data;

	if (boot_params.hdr.version < 0x0209)
		return;
	pa_data = boot_params.hdr.setup_data;
	while (pa_data) {
504
		data = early_memremap(pa_data, sizeof(*data));
505
		memblock_reserve(pa_data, sizeof(*data) + data->len);
506 507 508 509 510
		pa_data = data->next;
		early_iounmap(data, sizeof(*data));
	}
}

511 512 513 514 515
/*
 * --------- Crashkernel reservation ------------------------------
 */

#ifdef CONFIG_KEXEC
516

517 518 519 520 521 522 523 524 525 526 527 528
/*
 * Keep the crash kernel below this limit.  On 32 bits earlier kernels
 * would limit the kernel to the low 512 MiB due to mapping restrictions.
 * On 64 bits, kexec-tools currently limits us to 896 MiB; increase this
 * limit once kexec-tools are fixed.
 */
#ifdef CONFIG_X86_32
# define CRASH_KERNEL_ADDR_MAX	(512 << 20)
#else
# define CRASH_KERNEL_ADDR_MAX	(896 << 20)
#endif

529
static void __init reserve_crashkernel(void)
530 531 532 533 534
{
	unsigned long long total_mem;
	unsigned long long crash_size, crash_base;
	int ret;

535
	total_mem = memblock_phys_mem_size();
536 537 538

	ret = parse_crashkernel(boot_command_line, total_mem,
			&crash_size, &crash_base);
539 540 541 542 543
	if (ret != 0 || crash_size <= 0)
		return;

	/* 0 means: find the address automatically */
	if (crash_base <= 0) {
544 545
		const unsigned long long alignment = 16<<20;	/* 16M */

546
		/*
547
		 *  kexec want bzImage is below CRASH_KERNEL_ADDR_MAX
548 549
		 */
		crash_base = memblock_find_in_range(alignment,
550
			       CRASH_KERNEL_ADDR_MAX, crash_size, alignment);
551

T
Tejun Heo 已提交
552
		if (!crash_base) {
553
			pr_info("crashkernel reservation failed - No suitable area found.\n");
554 555
			return;
		}
556
	} else {
557 558
		unsigned long long start;

559 560
		start = memblock_find_in_range(crash_base,
				 crash_base + crash_size, crash_size, 1<<20);
561 562
		if (start != crash_base) {
			pr_info("crashkernel reservation failed - memory is in use.\n");
563 564
			return;
		}
565
	}
566
	memblock_reserve(crash_base, crash_size);
567

568 569 570 571 572
	printk(KERN_INFO "Reserving %ldMB of memory at %ldMB "
			"for crashkernel (System RAM: %ldMB)\n",
			(unsigned long)(crash_size >> 20),
			(unsigned long)(crash_base >> 20),
			(unsigned long)(total_mem >> 20));
573

574 575 576
	crashk_res.start = crash_base;
	crashk_res.end   = crash_base + crash_size - 1;
	insert_resource(&iomem_resource, &crashk_res);
577 578
}
#else
579
static void __init reserve_crashkernel(void)
580 581 582 583
{
}
#endif

584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606
static struct resource standard_io_resources[] = {
	{ .name = "dma1", .start = 0x00, .end = 0x1f,
		.flags = IORESOURCE_BUSY | IORESOURCE_IO },
	{ .name = "pic1", .start = 0x20, .end = 0x21,
		.flags = IORESOURCE_BUSY | IORESOURCE_IO },
	{ .name = "timer0", .start = 0x40, .end = 0x43,
		.flags = IORESOURCE_BUSY | IORESOURCE_IO },
	{ .name = "timer1", .start = 0x50, .end = 0x53,
		.flags = IORESOURCE_BUSY | IORESOURCE_IO },
	{ .name = "keyboard", .start = 0x60, .end = 0x60,
		.flags = IORESOURCE_BUSY | IORESOURCE_IO },
	{ .name = "keyboard", .start = 0x64, .end = 0x64,
		.flags = IORESOURCE_BUSY | IORESOURCE_IO },
	{ .name = "dma page reg", .start = 0x80, .end = 0x8f,
		.flags = IORESOURCE_BUSY | IORESOURCE_IO },
	{ .name = "pic2", .start = 0xa0, .end = 0xa1,
		.flags = IORESOURCE_BUSY | IORESOURCE_IO },
	{ .name = "dma2", .start = 0xc0, .end = 0xdf,
		.flags = IORESOURCE_BUSY | IORESOURCE_IO },
	{ .name = "fpu", .start = 0xf0, .end = 0xff,
		.flags = IORESOURCE_BUSY | IORESOURCE_IO }
};

607
void __init reserve_standard_io_resources(void)
608 609 610 611 612 613 614 615 616
{
	int i;

	/* request I/O space for devices used on all i[345]86 PCs */
	for (i = 0; i < ARRAY_SIZE(standard_io_resources); i++)
		request_resource(&ioport_resource, &standard_io_resources[i]);

}

617 618 619 620 621 622 623
static __init void reserve_ibft_region(void)
{
	unsigned long addr, size = 0;

	addr = find_ibft_region(&size);

	if (size)
624
		memblock_reserve(addr, size);
625 626
}

627
static unsigned reserve_low = CONFIG_X86_RESERVE_LOW << 10;
628

629 630
static bool __init snb_gfx_workaround_needed(void)
{
631
#ifdef CONFIG_PCI
632 633
	int i;
	u16 vendor, devid;
634
	static const __initconst u16 snb_ids[] = {
635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655
		0x0102,
		0x0112,
		0x0122,
		0x0106,
		0x0116,
		0x0126,
		0x010a,
	};

	/* Assume no if something weird is going on with PCI */
	if (!early_pci_allowed())
		return false;

	vendor = read_pci_config_16(0, 2, 0, PCI_VENDOR_ID);
	if (vendor != 0x8086)
		return false;

	devid = read_pci_config_16(0, 2, 0, PCI_DEVICE_ID);
	for (i = 0; i < ARRAY_SIZE(snb_ids); i++)
		if (devid == snb_ids[i])
			return true;
656
#endif
657 658 659 660 661 662 663 664 665 666

	return false;
}

/*
 * Sandy Bridge graphics has trouble with certain ranges, exclude
 * them from allocation.
 */
static void __init trim_snb_memory(void)
{
667
	static const __initconst unsigned long bad_pages[] = {
668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705
		0x20050000,
		0x20110000,
		0x20130000,
		0x20138000,
		0x40004000,
	};
	int i;

	if (!snb_gfx_workaround_needed())
		return;

	printk(KERN_DEBUG "reserving inaccessible SNB gfx pages\n");

	/*
	 * Reserve all memory below the 1 MB mark that has not
	 * already been reserved.
	 */
	memblock_reserve(0, 1<<20);
	
	for (i = 0; i < ARRAY_SIZE(bad_pages); i++) {
		if (memblock_reserve(bad_pages[i], PAGE_SIZE))
			printk(KERN_WARNING "failed to reserve 0x%08lx\n",
			       bad_pages[i]);
	}
}

/*
 * Here we put platform-specific memory range workarounds, i.e.
 * memory known to be corrupt or otherwise in need to be reserved on
 * specific platforms.
 *
 * If this gets used more widely it could use a real dispatch mechanism.
 */
static void __init trim_platform_memory_ranges(void)
{
	trim_snb_memory();
}

706 707 708 709 710 711
static void __init trim_bios_range(void)
{
	/*
	 * A special case is the first 4Kb of memory;
	 * This is a BIOS owned area, not kernel ram, but generally
	 * not listed as such in the E820 table.
712 713 714
	 *
	 * This typically reserves additional memory (64KiB by default)
	 * since some BIOSes are known to corrupt low memory.  See the
715
	 * Kconfig help text for X86_RESERVE_LOW.
716
	 */
717
	e820_update_range(0, ALIGN(reserve_low, PAGE_SIZE),
718 719
			  E820_RAM, E820_RESERVED);

720 721 722 723 724 725
	/*
	 * special case: Some BIOSen report the PC BIOS
	 * area (640->1Mb) as ram even though it is not.
	 * take them out.
	 */
	e820_remove_range(BIOS_BEGIN, BIOS_END - BIOS_BEGIN, E820_RAM, 1);
726

727 728 729
	sanitize_e820_map(e820.map, ARRAY_SIZE(e820.map), &e820.nr_map);
}

730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750
/* called before trim_bios_range() to spare extra sanitize */
static void __init e820_add_kernel_range(void)
{
	u64 start = __pa_symbol(_text);
	u64 size = __pa_symbol(_end) - start;

	/*
	 * Complain if .text .data and .bss are not marked as E820_RAM and
	 * attempt to fix it by adding the range. We may have a confused BIOS,
	 * or the user may have used memmap=exactmap or memmap=xxM$yyM to
	 * exclude kernel range. If we really are running on top non-RAM,
	 * we will crash later anyways.
	 */
	if (e820_all_mapped(start, start + size, E820_RAM))
		return;

	pr_warn(".text .data .bss are not marked as E820_RAM!\n");
	e820_remove_range(start, size, E820_RAM, 0);
	e820_add_region(start, size, E820_RAM);
}

751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772
static int __init parse_reservelow(char *p)
{
	unsigned long long size;

	if (!p)
		return -EINVAL;

	size = memparse(p, &p);

	if (size < 4096)
		size = 4096;

	if (size > 640*1024)
		size = 640*1024;

	reserve_low = size;

	return 0;
}

early_param("reservelow", parse_reservelow);

L
Linus Torvalds 已提交
773 774 775 776 777 778 779
/*
 * Determine if we were loaded by an EFI loader.  If so, then we have also been
 * passed the efi memmap, systab, etc., so we should use these data structures
 * for initialization.  Note, the efi init code path is determined by the
 * global efi_enabled. This allows the same kernel image to be used on existing
 * systems (with a traditional BIOS) as well as on EFI systems.
 */
780 781 782 783 784 785
/*
 * setup_arch - architecture-specific boot-time initializations
 *
 * Note: On x86_64, fixmaps are ready for use even before this is called.
 */

L
Linus Torvalds 已提交
786 787
void __init setup_arch(char **cmdline_p)
{
788 789
	early_reserve_initrd();

790
#ifdef CONFIG_X86_32
L
Linus Torvalds 已提交
791
	memcpy(&boot_cpu_data, &new_cpu_data, sizeof(new_cpu_data));
I
Ingo Molnar 已提交
792
	visws_early_detect();
793 794 795 796 797 798 799 800 801 802 803

	/*
	 * copy kernel address range established so far and switch
	 * to the proper swapper page table
	 */
	clone_pgd_range(swapper_pg_dir     + KERNEL_PGD_BOUNDARY,
			initial_page_table + KERNEL_PGD_BOUNDARY,
			KERNEL_PGD_PTRS);

	load_cr3(swapper_pg_dir);
	__flush_tlb_all();
804 805 806
#else
	printk(KERN_INFO "Command line: %s\n", boot_command_line);
#endif
L
Linus Torvalds 已提交
807

808 809 810 811
	/*
	 * If we have OLPC OFW, we might end up relocating the fixmap due to
	 * reserve_top(), so do this before touching the ioremap area.
	 */
812 813
	olpc_ofw_detect();

814
	early_trap_init();
815
	early_cpu_init();
816 817
	early_ioremap_init();

818 819
	setup_olpc_ofw_pgd();

820 821 822
	ROOT_DEV = old_decode_dev(boot_params.hdr.root_dev);
	screen_info = boot_params.screen_info;
	edid_info = boot_params.edid_info;
823
#ifdef CONFIG_X86_32
824 825
	apm_info.bios = boot_params.apm_bios_info;
	ist_info = boot_params.ist_info;
826
	if (boot_params.sys_desc_table.length != 0) {
827 828 829
		machine_id = boot_params.sys_desc_table.table[0];
		machine_submodel_id = boot_params.sys_desc_table.table[1];
		BIOS_revision = boot_params.sys_desc_table.table[2];
L
Linus Torvalds 已提交
830
	}
831 832
#endif
	saved_video_mode = boot_params.hdr.vid_mode;
833
	bootloader_type = boot_params.hdr.type_of_loader;
834 835 836 837 838 839
	if ((bootloader_type >> 4) == 0xe) {
		bootloader_type &= 0xf;
		bootloader_type |= (boot_params.hdr.ext_loader_type+0x10) << 4;
	}
	bootloader_version  = bootloader_type & 0xf;
	bootloader_version |= boot_params.hdr.ext_loader_ver << 4;
L
Linus Torvalds 已提交
840 841

#ifdef CONFIG_BLK_DEV_RAM
842 843 844
	rd_image_start = boot_params.hdr.ram_size & RAMDISK_IMAGE_START_MASK;
	rd_prompt = ((boot_params.hdr.ram_size & RAMDISK_PROMPT_FLAG) != 0);
	rd_doload = ((boot_params.hdr.ram_size & RAMDISK_LOAD_FLAG) != 0);
L
Linus Torvalds 已提交
845
#endif
846 847
#ifdef CONFIG_EFI
	if (!strncmp((char *)&boot_params.efi_info.efi_loader_signature,
848
		     "EL32", 4)) {
849
		efi_enabled = 1;
850 851 852 853 854
		efi_64bit = false;
	} else if (!strncmp((char *)&boot_params.efi_info.efi_loader_signature,
		     "EL64", 4)) {
		efi_enabled = 1;
		efi_64bit = true;
855
	}
856 857
	if (efi_enabled && efi_memblock_x86_reserve_range())
		efi_enabled = 0;
858 859
#endif

860
	x86_init.oem.arch_setup();
861

862
	iomem_resource.end = (1ULL << boot_cpu_data.x86_phys_bits) - 1;
863
	setup_memory_map();
864
	parse_setup_data();
865 866
	/* update the e820_saved too */
	e820_reserve_setup_data();
867

L
Linus Torvalds 已提交
868 869
	copy_edd();

870
	if (!boot_params.hdr.root_flags)
L
Linus Torvalds 已提交
871 872 873 874
		root_mountflags &= ~MS_RDONLY;
	init_mm.start_code = (unsigned long) _text;
	init_mm.end_code = (unsigned long) _etext;
	init_mm.end_data = (unsigned long) _edata;
875
	init_mm.brk = _brk_end;
L
Linus Torvalds 已提交
876 877 878 879 880

	code_resource.start = virt_to_phys(_text);
	code_resource.end = virt_to_phys(_etext)-1;
	data_resource.start = virt_to_phys(_etext);
	data_resource.end = virt_to_phys(_edata)-1;
B
Bernhard Walle 已提交
881 882
	bss_resource.start = virt_to_phys(&__bss_start);
	bss_resource.end = virt_to_phys(&__bss_stop)-1;
L
Linus Torvalds 已提交
883

884 885 886 887 888 889 890 891 892 893 894 895 896
#ifdef CONFIG_CMDLINE_BOOL
#ifdef CONFIG_CMDLINE_OVERRIDE
	strlcpy(boot_command_line, builtin_cmdline, COMMAND_LINE_SIZE);
#else
	if (builtin_cmdline[0]) {
		/* append boot loader cmdline to builtin */
		strlcat(builtin_cmdline, " ", COMMAND_LINE_SIZE);
		strlcat(builtin_cmdline, boot_command_line, COMMAND_LINE_SIZE);
		strlcpy(boot_command_line, builtin_cmdline, COMMAND_LINE_SIZE);
	}
#endif
#endif

897 898 899 900
	strlcpy(command_line, boot_command_line, COMMAND_LINE_SIZE);
	*cmdline_p = command_line;

	/*
901 902 903 904 905
	 * x86_configure_nx() is called before parse_early_param() to detect
	 * whether hardware doesn't support NX (so that the early EHCI debug
	 * console setup can safely call set_fixmap()). It may then be called
	 * again from within noexec_setup() during parsing early parameters
	 * to honor the respective command line option.
906
	 */
907
	x86_configure_nx();
908 909 910

	parse_early_param();

911
	x86_report_nx();
912

913
	/* after early param, so could get panic from serial */
914
	memblock_x86_reserve_range_setup_data();
915

916
	if (acpi_mps_check()) {
I
Ingo Molnar 已提交
917
#ifdef CONFIG_X86_LOCAL_APIC
918
		disable_apic = 1;
I
Ingo Molnar 已提交
919
#endif
920
		setup_clear_cpu_cap(X86_FEATURE_APIC);
921 922
	}

923 924 925 926 927
#ifdef CONFIG_PCI
	if (pci_early_dump_regs)
		early_dump_pci_devices();
#endif

928
	finish_e820_parsing();
929

B
Brian Maly 已提交
930 931 932
	if (efi_enabled)
		efi_init();

933 934
	dmi_scan_machine();

935 936 937 938
	/*
	 * VMware detection requires dmi to be available, so this
	 * needs to be done after dmi_scan_machine, for the BP.
	 */
939
	init_hypervisor_platform();
940

941
	x86_init.resources.probe_roms();
942 943 944 945 946 947

	/* after parse_early_param, so could debug it */
	insert_resource(&iomem_resource, &code_resource);
	insert_resource(&iomem_resource, &data_resource);
	insert_resource(&iomem_resource, &bss_resource);

948
	e820_add_kernel_range();
949
	trim_bios_range();
950
#ifdef CONFIG_X86_32
Y
Yinghai Lu 已提交
951 952 953 954 955 956 957
	if (ppro_with_ram_bug()) {
		e820_update_range(0x70000000ULL, 0x40000ULL, E820_RAM,
				  E820_RESERVED);
		sanitize_e820_map(e820.map, ARRAY_SIZE(e820.map), &e820.nr_map);
		printk(KERN_INFO "fixed physical RAM map:\n");
		e820_print_map("bad_ppro");
	}
958 959 960
#else
	early_gart_iommu_check();
#endif
Y
Yinghai Lu 已提交
961

962 963 964 965
	/*
	 * partially used pages are not usable - thus
	 * we are rounding upwards:
	 */
966
	max_pfn = e820_end_of_ram_pfn();
967

968 969
	/* update e820 for memory not covered by WB MTRRs */
	mtrr_bp_init();
970
	if (mtrr_trim_uncached_memory(max_pfn))
971
		max_pfn = e820_end_of_ram_pfn();
972

973
#ifdef CONFIG_X86_32
974
	/* max_low_pfn get updated here */
975
	find_low_pfn_range();
976 977 978
#else
	num_physpages = max_pfn;

Y
Yinghai Lu 已提交
979
	check_x2apic();
980 981 982

	/* How many end-of-memory variables you have, grandma! */
	/* need this before calling reserve_initrd */
983 984 985 986 987
	if (max_pfn > (1UL<<(32 - PAGE_SHIFT)))
		max_low_pfn = e820_end_of_low_ram_pfn();
	else
		max_low_pfn = max_pfn;

988
	high_memory = (void *)__va(max_pfn * PAGE_SIZE - 1) + 1;
989 990
#endif

991 992 993 994 995
	/*
	 * Find and reserve possible boot-time SMP configuration:
	 */
	find_smp_config();

996 997
	reserve_ibft_region();

Y
Yinghai Lu 已提交
998 999
	early_alloc_pgt_buf();

1000 1001 1002 1003 1004 1005 1006
	/*
	 * Need to conclude brk, before memblock_x86_fill()
	 *  it could use memblock_find_in_range, could overlap with
	 *  brk area.
	 */
	reserve_brk();

1007 1008
	cleanup_highmap();

1009
	memblock.current_limit = ISA_END_ADDRESS;
1010 1011
	memblock_x86_fill();

1012 1013 1014 1015 1016 1017 1018
	/*
	 * The EFI specification says that boot service code won't be called
	 * after ExitBootServices(). This is, in fact, a lie.
	 */
	if (efi_enabled)
		efi_reserve_boot_services();

1019 1020 1021 1022 1023 1024 1025
	/* preallocate 4k for mptable mpc */
	early_reserve_e820_mpc_new();

#ifdef CONFIG_X86_CHECK_BIOS_CORRUPTION
	setup_bios_corruption_check();
#endif

1026
#ifdef CONFIG_X86_32
1027 1028
	printk(KERN_DEBUG "initial memory mapped: [mem 0x00000000-%#010lx]\n",
			(max_pfn_mapped<<PAGE_SHIFT) - 1);
1029
#endif
1030

1031
	reserve_real_mode();
1032

1033 1034
	trim_platform_memory_ranges();

1035
	init_mem_mapping();
1036

1037 1038
	early_trap_pf_init();

1039 1040
	setup_real_mode();

1041
	memblock.current_limit = get_max_mapped();
1042
	dma_contiguous_reserve(0);
1043

1044 1045 1046 1047 1048 1049 1050 1051
	/*
	 * NOTE: On x86-32, only from this point on, fixmaps are ready for use.
	 */

#ifdef CONFIG_PROVIDE_OHCI1394_DMA_INIT
	if (init_ohci1394_dma_early)
		init_ohci1394_dma_on_all_controllers();
#endif
1052 1053
	/* Allocate bigger log buffer */
	setup_log_buf(1);
1054

1055 1056
	reserve_initrd();

D
David Rientjes 已提交
1057
#if defined(CONFIG_ACPI) && defined(CONFIG_BLK_DEV_INITRD)
1058
	acpi_initrd_override((void *)initrd_start, initrd_end - initrd_start);
D
David Rientjes 已提交
1059
#endif
1060

1061 1062
	reserve_crashkernel();

1063 1064
	vsmp_init();

1065 1066 1067 1068 1069 1070 1071
	io_delay_init();

	/*
	 * Parse the ACPI tables for possible boot-time SMP configuration.
	 */
	acpi_boot_table_init();

1072 1073
	early_acpi_boot_init();

1074
	initmem_init();
1075
	memblock_find_dma_reserve();
1076

1077
#ifdef CONFIG_KVM_GUEST
1078 1079 1080
	kvmclock_init();
#endif

1081
	x86_init.paging.pagetable_init();
1082

1083 1084 1085
	if (boot_cpu_data.cpuid_level >= 0) {
		/* A CPU has %cr4 if and only if it has CPUID */
		mmu_cr4_features = read_cr4();
1086 1087
		if (trampoline_cr4_features)
			*trampoline_cr4_features = mmu_cr4_features;
1088 1089
	}

1090 1091 1092 1093 1094 1095
#ifdef CONFIG_X86_32
	/* sync back kernel address range */
	clone_pgd_range(initial_page_table + KERNEL_PGD_BOUNDARY,
			swapper_pg_dir     + KERNEL_PGD_BOUNDARY,
			KERNEL_PGD_PTRS);
#endif
1096

1097 1098
	tboot_probe();

1099 1100 1101 1102
#ifdef CONFIG_X86_64
	map_vsyscall();
#endif

1103
	generic_apic_probe();
L
Linus Torvalds 已提交
1104

1105
	early_quirks();
1106

1107 1108 1109
	/*
	 * Read APIC and some other early information from ACPI tables.
	 */
L
Linus Torvalds 已提交
1110
	acpi_boot_init();
F
Feng Tang 已提交
1111
	sfi_init();
1112
	x86_dtb_init();
1113

1114 1115 1116
	/*
	 * get boot-time SMP configuration:
	 */
1117 1118
	if (smp_found_config)
		get_smp_config();
1119

1120
	prefill_possible_map();
Y
Yinghai Lu 已提交
1121

1122 1123
	init_cpu_to_node();

1124
	init_apic_mappings();
I
Ingo Molnar 已提交
1125 1126
	if (x86_io_apic_ops.init)
		x86_io_apic_ops.init();
1127

1128
	kvm_guest_init();
L
Linus Torvalds 已提交
1129

1130
	e820_reserve_resources();
1131
	e820_mark_nosave_regions(max_low_pfn);
L
Linus Torvalds 已提交
1132

1133
	x86_init.resources.reserve_resources();
1134 1135 1136

	e820_setup_gap();

L
Linus Torvalds 已提交
1137 1138 1139 1140 1141 1142 1143 1144
#ifdef CONFIG_VT
#if defined(CONFIG_VGA_CONSOLE)
	if (!efi_enabled || (efi_mem_type(0xa0000) != EFI_CONVENTIONAL_MEMORY))
		conswitchp = &vga_con;
#elif defined(CONFIG_DUMMY_CONSOLE)
	conswitchp = &dummy_con;
#endif
#endif
1145
	x86_init.oem.banner();
1146

1147 1148
	x86_init.timers.wallclock_init();

1149
	mcheck_init();
1150

1151
	arch_init_ideal_nops();
1152 1153

	register_refined_jiffies(CLOCK_TICK_RATE);
1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165

#ifdef CONFIG_EFI
	/* Once setup is done above, disable efi_enabled on mismatched
	 * firmware/kernel archtectures since there is no support for
	 * runtime services.
	 */
	if (efi_enabled && IS_ENABLED(CONFIG_X86_64) != efi_64bit) {
		pr_info("efi: Setup done, disabling due to 32/64-bit mismatch\n");
		efi_unmap_memmap();
		efi_enabled = 0;
	}
#endif
L
Linus Torvalds 已提交
1166
}
1167

1168 1169
#ifdef CONFIG_X86_32

1170 1171 1172 1173 1174
static struct resource video_ram_resource = {
	.name	= "Video RAM area",
	.start	= 0xa0000,
	.end	= 0xbffff,
	.flags	= IORESOURCE_BUSY | IORESOURCE_MEM
1175 1176
};

1177
void __init i386_reserve_resources(void)
1178
{
1179 1180
	request_resource(&iomem_resource, &video_ram_resource);
	reserve_standard_io_resources();
1181 1182 1183
}

#endif /* CONFIG_X86_32 */