head_32.S 19.3 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9 10
/*
 *
 *  Copyright (C) 1991, 1992  Linus Torvalds
 *
 *  Enhanced CPU detection and feature setting code by Mike Jagdis
 *  and Martin Mares, November 1997.
 */

.text
#include <linux/threads.h>
11
#include <linux/init.h>
L
Linus Torvalds 已提交
12 13
#include <linux/linkage.h>
#include <asm/segment.h>
14 15
#include <asm/page_types.h>
#include <asm/pgtable_types.h>
L
Linus Torvalds 已提交
16 17
#include <asm/cache.h>
#include <asm/thread_info.h>
18
#include <asm/asm-offsets.h>
L
Linus Torvalds 已提交
19
#include <asm/setup.h>
20
#include <asm/processor-flags.h>
21
#include <asm/msr-index.h>
22
#include <asm/cpufeatures.h>
23
#include <asm/percpu.h>
24
#include <asm/nops.h>
25
#include <asm/bootparam.h>
26 27 28

/* Physical address */
#define pa(X) ((X) - __PAGE_OFFSET)
L
Linus Torvalds 已提交
29 30 31 32 33 34 35 36 37 38 39 40 41 42 43

/*
 * References to members of the new_cpu_data structure.
 */

#define X86		new_cpu_data+CPUINFO_x86
#define X86_VENDOR	new_cpu_data+CPUINFO_x86_vendor
#define X86_MODEL	new_cpu_data+CPUINFO_x86_model
#define X86_MASK	new_cpu_data+CPUINFO_x86_mask
#define X86_HARD_MATH	new_cpu_data+CPUINFO_hard_math
#define X86_CPUID	new_cpu_data+CPUINFO_cpuid_level
#define X86_CAPABILITY	new_cpu_data+CPUINFO_x86_capability
#define X86_VENDOR_ID	new_cpu_data+CPUINFO_x86_vendor_id

/*
44 45
 * This is how much memory in addition to the memory covered up to
 * and including _end we need mapped initially.
46
 * We need:
47 48
 *     (KERNEL_IMAGE_SIZE/4096) / 1024 pages (worst case, non PAE)
 *     (KERNEL_IMAGE_SIZE/4096) / 512 + 4 pages (worst case for PAE)
L
Linus Torvalds 已提交
49 50 51 52 53
 *
 * Modulo rounding, each megabyte assigned here requires a kilobyte of
 * memory, which is currently unreclaimed.
 *
 * This should be a multiple of a page.
54 55 56
 *
 * KERNEL_IMAGE_SIZE should be greater than pa(_end)
 * and small than max_low_pfn, otherwise will waste some page table entries
L
Linus Torvalds 已提交
57 58
 */

59
#if PTRS_PER_PMD > 1
60
#define PAGE_TABLE_SIZE(pages) (((pages) / PTRS_PER_PMD) + PTRS_PER_PGD)
61
#else
62
#define PAGE_TABLE_SIZE(pages) ((pages) / PTRS_PER_PGD)
63 64
#endif

65 66 67 68 69 70 71 72 73 74
/*
 * Number of possible pages in the lowmem region.
 *
 * We shift 2 by 31 instead of 1 by 32 to the left in order to avoid a
 * gas warning about overflowing shift count when gas has been compiled
 * with only a host target support using a 32-bit type for internal
 * representation.
 */
LOWMEM_PAGES = (((2<<31) - __PAGE_OFFSET) >> PAGE_SHIFT)

75
/* Enough space to fit pagetables for the low memory linear map */
76
MAPPING_BEYOND_END = PAGE_TABLE_SIZE(LOWMEM_PAGES) << PAGE_SHIFT
77 78 79

/*
 * Worst-case size of the kernel mapping we need to make:
80 81
 * a relocatable kernel can live anywhere in lowmem, so we need to be able
 * to map all of lowmem.
82
 */
83
KERNEL_PAGES = LOWMEM_PAGES
84

85
INIT_MAP_SIZE = PAGE_TABLE_SIZE(KERNEL_PAGES) * PAGE_SIZE
86
RESERVE_BRK(pagetables, INIT_MAP_SIZE)
87

L
Linus Torvalds 已提交
88 89 90 91 92 93 94
/*
 * 32-bit kernel entrypoint; only used by the boot CPU.  On entry,
 * %esi points to the real-mode code as a 32-bit pointer.
 * CS and DS must be 4 GB flat segments, but we don't depend on
 * any particular GDT layout, because we load our own as soon as we
 * can.
 */
95
__HEAD
L
Linus Torvalds 已提交
96
ENTRY(startup_32)
97 98
	movl pa(stack_start),%ecx
	
R
Rusty Russell 已提交
99 100
	/* test KEEP_SEGMENTS flag to see if the bootloader is asking
		us to not reload segments */
101
	testb $KEEP_SEGMENTS, BP_loadflags(%esi)
R
Rusty Russell 已提交
102
	jnz 2f
L
Linus Torvalds 已提交
103 104 105 106

/*
 * Set segments to known values.
 */
107
	lgdt pa(boot_gdt_descr)
L
Linus Torvalds 已提交
108 109 110 111 112
	movl $(__BOOT_DS),%eax
	movl %eax,%ds
	movl %eax,%es
	movl %eax,%fs
	movl %eax,%gs
113
	movl %eax,%ss
R
Rusty Russell 已提交
114
2:
115
	leal -__PAGE_OFFSET(%ecx),%esp
L
Linus Torvalds 已提交
116 117 118 119

/*
 * Clear BSS first so that there are no surprises...
 */
R
Rusty Russell 已提交
120
	cld
L
Linus Torvalds 已提交
121
	xorl %eax,%eax
122 123
	movl $pa(__bss_start),%edi
	movl $pa(__bss_stop),%ecx
L
Linus Torvalds 已提交
124 125 126
	subl %edi,%ecx
	shrl $2,%ecx
	rep ; stosl
127 128 129 130 131 132 133 134
/*
 * Copy bootup parameters out of the way.
 * Note: %esi still has the pointer to the real-mode data.
 * With the kexec as boot loader, parameter segment might be loaded beyond
 * kernel image and might not even be addressable by early boot page tables.
 * (kexec on panic case). Hence copy out the parameters before initializing
 * page tables.
 */
135
	movl $pa(boot_params),%edi
136 137 138 139
	movl $(PARAM_SIZE/4),%ecx
	cld
	rep
	movsl
140
	movl pa(boot_params) + NEW_CL_POINTER,%esi
141
	andl %esi,%esi
142
	jz 1f			# No command line
143
	movl $pa(boot_command_line),%edi
144 145 146 147
	movl $(COMMAND_LINE_SIZE/4),%ecx
	rep
	movsl
1:
L
Linus Torvalds 已提交
148

149
#ifdef CONFIG_OLPC
150 151 152 153 154
	/* save OFW's pgdir table for later use when calling into OFW */
	movl %cr3, %eax
	movl %eax, pa(olpc_ofw_pgd)
#endif

155
#ifdef CONFIG_MICROCODE
156 157 158 159
	/* Early load ucode on BSP. */
	call load_ucode_bsp
#endif

L
Linus Torvalds 已提交
160 161
/*
 * Initialize page tables.  This creates a PDE and a set of page
162
 * tables, which are located immediately beyond __brk_base.  The variable
163
 * _brk_end is set up to point to the first "safe" location.
L
Linus Torvalds 已提交
164
 * Mappings are created both at virtual address 0 (identity mapping)
165
 * and PAGE_OFFSET for up to _end.
L
Linus Torvalds 已提交
166
 */
167 168 169
#ifdef CONFIG_X86_PAE

	/*
170 171 172 173
	 * In PAE mode initial_page_table is statically defined to contain
	 * enough entries to cover the VMSPLIT option (that is the top 1, 2 or 3
	 * entries). The identity mapping is handled by pointing two PGD entries
	 * to the first kernel PMD.
174
	 *
175
	 * Note the upper half of each PMD or PTE are always zero at this stage.
176 177
	 */

J
Joe Korty 已提交
178
#define KPMDS (((-__PAGE_OFFSET) >> 30) & 3) /* Number of kernel PMDs */
179 180 181

	xorl %ebx,%ebx				/* %ebx is kept at zero */

182
	movl $pa(__brk_base), %edi
183
	movl $pa(initial_pg_pmd), %edx
184
	movl $PTE_IDENT_ATTR, %eax
185
10:
186
	leal PDE_IDENT_ATTR(%edi),%ecx		/* Create PMD entry */
187 188 189 190 191 192 193 194 195 196 197 198 199
	movl %ecx,(%edx)			/* Store PMD entry */
						/* Upper half already zero */
	addl $8,%edx
	movl $512,%ecx
11:
	stosl
	xchgl %eax,%ebx
	stosl
	xchgl %eax,%ebx
	addl $0x1000,%eax
	loop 11b

	/*
200
	 * End condition: we must map up to the end + MAPPING_BEYOND_END.
201
	 */
202
	movl $pa(_end) + MAPPING_BEYOND_END + PTE_IDENT_ATTR, %ebp
203 204 205
	cmpl %ebp,%eax
	jb 10b
1:
206 207
	addl $__PAGE_OFFSET, %edi
	movl %edi, pa(_brk_end)
208 209
	shrl $12, %eax
	movl %eax, pa(max_pfn_mapped)
210 211

	/* Do early initialization of the fixmap area */
212 213
	movl $pa(initial_pg_fixmap)+PDE_IDENT_ATTR,%eax
	movl %eax,pa(initial_pg_pmd+0x1000*KPMDS-8)
214 215 216 217
#else	/* Not PAE */

page_pde_offset = (__PAGE_OFFSET >> 20);

218
	movl $pa(__brk_base), %edi
219
	movl $pa(initial_page_table), %edx
220
	movl $PTE_IDENT_ATTR, %eax
L
Linus Torvalds 已提交
221
10:
222
	leal PDE_IDENT_ATTR(%edi),%ecx		/* Create PDE entry */
L
Linus Torvalds 已提交
223 224 225 226 227 228 229 230
	movl %ecx,(%edx)			/* Store identity PDE entry */
	movl %ecx,page_pde_offset(%edx)		/* Store kernel PDE entry */
	addl $4,%edx
	movl $1024, %ecx
11:
	stosl
	addl $0x1000,%eax
	loop 11b
231
	/*
232
	 * End condition: we must map up to the end + MAPPING_BEYOND_END.
233
	 */
234
	movl $pa(_end) + MAPPING_BEYOND_END + PTE_IDENT_ATTR, %ebp
L
Linus Torvalds 已提交
235 236
	cmpl %ebp,%eax
	jb 10b
237 238
	addl $__PAGE_OFFSET, %edi
	movl %edi, pa(_brk_end)
239 240
	shrl $12, %eax
	movl %eax, pa(max_pfn_mapped)
241

242
	/* Do early initialization of the fixmap area */
243 244
	movl $pa(initial_pg_fixmap)+PDE_IDENT_ATTR,%eax
	movl %eax,pa(initial_page_table+0xffc)
245
#endif
246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281

#ifdef CONFIG_PARAVIRT
	/* This is can only trip for a broken bootloader... */
	cmpw $0x207, pa(boot_params + BP_version)
	jb default_entry

	/* Paravirt-compatible boot parameters.  Look to see what architecture
		we're booting under. */
	movl pa(boot_params + BP_hardware_subarch), %eax
	cmpl $num_subarch_entries, %eax
	jae bad_subarch

	movl pa(subarch_entries)(,%eax,4), %eax
	subl $__PAGE_OFFSET, %eax
	jmp *%eax

bad_subarch:
WEAK(lguest_entry)
WEAK(xen_entry)
	/* Unknown implementation; there's really
	   nothing we can do at this point. */
	ud2a

	__INITDATA

subarch_entries:
	.long default_entry		/* normal x86/PC */
	.long lguest_entry		/* lguest hypervisor */
	.long xen_entry			/* Xen hypervisor */
	.long default_entry		/* Moorestown MID */
num_subarch_entries = (. - subarch_entries) / 4
.previous
#else
	jmp default_entry
#endif /* CONFIG_PARAVIRT */

282 283 284 285 286 287 288 289 290 291 292 293 294
#ifdef CONFIG_HOTPLUG_CPU
/*
 * Boot CPU0 entry point. It's called from play_dead(). Everything has been set
 * up already except stack. We just set up stack here. Then call
 * start_secondary().
 */
ENTRY(start_cpu0)
	movl stack_start, %ecx
	movl %ecx, %esp
	jmp  *(initial_code)
ENDPROC(start_cpu0)
#endif

L
Linus Torvalds 已提交
295 296 297
/*
 * Non-boot CPU entry point; entered from trampoline.S
 * We can't lgdt here, because lgdt itself uses a data segment, but
298
 * we know the trampoline has already loaded the boot_gdt for us.
299 300 301
 *
 * If cpu hotplug is not supported then this code can go in init section
 * which will be freed later
L
Linus Torvalds 已提交
302 303 304 305 306 307 308 309
 */
ENTRY(startup_32_smp)
	cld
	movl $(__BOOT_DS),%eax
	movl %eax,%ds
	movl %eax,%es
	movl %eax,%fs
	movl %eax,%gs
310 311 312
	movl pa(stack_start),%ecx
	movl %eax,%ss
	leal -__PAGE_OFFSET(%ecx),%esp
313

314
#ifdef CONFIG_MICROCODE
315 316 317 318
	/* Early load ucode on AP. */
	call load_ucode_ap
#endif

319
default_entry:
320 321 322 323 324 325
#define CR0_STATE	(X86_CR0_PE | X86_CR0_MP | X86_CR0_ET | \
			 X86_CR0_NE | X86_CR0_WP | X86_CR0_AM | \
			 X86_CR0_PG)
	movl $(CR0_STATE & ~X86_CR0_PG),%eax
	movl %eax,%cr0

L
Linus Torvalds 已提交
326
/*
327 328 329 330 331 332 333 334 335 336 337
 * We want to start out with EFLAGS unambiguously cleared. Some BIOSes leave
 * bits like NT set. This would confuse the debugger if this code is traced. So
 * initialize them properly now before switching to protected mode. That means
 * DF in particular (even though we have cleared it earlier after copying the
 * command line) because GCC expects it.
 */
	pushl $0
	popfl

/*
 * New page tables may be in 4Mbyte page mode and may be using the global pages.
L
Linus Torvalds 已提交
338
 *
339 340
 * NOTE! If we are on a 486 we may have no cr4 at all! Specifically, cr4 exists
 * if and only if CPUID exists and has flags other than the FPU flag set.
L
Linus Torvalds 已提交
341
 */
342
	movl $-1,pa(X86_CPUID)		# preset CPUID level
343 344
	movl $X86_EFLAGS_ID,%ecx
	pushl %ecx
345
	popfl				# set EFLAGS=ID
346
	pushfl
347 348
	popl %eax			# get EFLAGS
	testl $X86_EFLAGS_ID,%eax	# did EFLAGS.ID remained set?
349
	jz enable_paging		# hw disallowed setting of ID bit
350 351 352 353 354
					# which means no CPUID and no CR4

	xorl %eax,%eax
	cpuid
	movl %eax,pa(X86_CPUID)		# save largest std CPUID function
355

356 357
	movl $1,%eax
	cpuid
358
	andl $~1,%edx			# Ignore CPUID.FPU
359
	jz enable_paging		# No flags or only CPUID.FPU = no CR4
360

361
	movl pa(mmu_cr4_features),%eax
L
Linus Torvalds 已提交
362 363
	movl %eax,%cr4

364
	testb $X86_CR4_PAE, %al		# check if PAE is enabled
365
	jz enable_paging
L
Linus Torvalds 已提交
366 367 368 369

	/* Check if extended functions are implemented */
	movl $0x80000000, %eax
	cpuid
370 371 372
	/* Value must be in the range 0x80000001 to 0x8000ffff */
	subl $0x80000001, %eax
	cmpl $(0x8000ffff-0x80000001), %eax
373
	ja enable_paging
374 375 376 377

	/* Clear bogus XD_DISABLE bits */
	call verify_cpu

L
Linus Torvalds 已提交
378 379 380
	mov $0x80000001, %eax
	cpuid
	/* Execute Disable bit supported? */
381
	btl $(X86_FEATURE_NX & 31), %edx
382
	jnc enable_paging
L
Linus Torvalds 已提交
383 384

	/* Setup EFER (Extended Feature Enable Register) */
385
	movl $MSR_EFER, %ecx
L
Linus Torvalds 已提交
386 387
	rdmsr

388
	btsl $_EFER_NX, %eax
L
Linus Torvalds 已提交
389 390 391
	/* Make changes effective */
	wrmsr

392 393 394 395 396 397
	/*
	 * And make sure that all the mappings we set up have NX set from
	 * the beginning.
	 */
	orl $(1 << (_PAGE_BIT_NX - 32)), pa(__supported_pte_mask + 4)

398
enable_paging:
L
Linus Torvalds 已提交
399 400 401 402

/*
 * Enable paging
 */
403
	movl $pa(initial_page_table), %eax
L
Linus Torvalds 已提交
404
	movl %eax,%cr3		/* set the page table pointer.. */
405
	movl $CR0_STATE,%eax
L
Linus Torvalds 已提交
406 407 408
	movl %eax,%cr0		/* ..and set paging (PG) bit */
	ljmp $__BOOT_CS,$1f	/* Clear prefetch and normalize %eip */
1:
409 410
	/* Shift the stack pointer to a virtual address */
	addl $__PAGE_OFFSET, %esp
L
Linus Torvalds 已提交
411 412 413 414 415

/*
 * start system 32-bit setup. We need to re-do some of the things done
 * in 16-bit mode for the "real" operations.
 */
416 417 418 419 420
	movl setup_once_ref,%eax
	andl %eax,%eax
	jz 1f				# Did we do this already?
	call *%eax
1:
421

L
Linus Torvalds 已提交
422
/*
423
 * Check if it is 486
L
Linus Torvalds 已提交
424
 */
425
	movb $4,X86			# at least 486
426
	cmpl $-1,X86_CPUID
L
Linus Torvalds 已提交
427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451
	je is486

	/* get vendor info */
	xorl %eax,%eax			# call CPUID with 0 -> return vendor ID
	cpuid
	movl %eax,X86_CPUID		# save CPUID level
	movl %ebx,X86_VENDOR_ID		# lo 4 chars
	movl %edx,X86_VENDOR_ID+4	# next 4 chars
	movl %ecx,X86_VENDOR_ID+8	# last 4 chars

	orl %eax,%eax			# do we have processor info as well?
	je is486

	movl $1,%eax		# Use the CPUID instruction to get CPU type
	cpuid
	movb %al,%cl		# save reg for future use
	andb $0x0f,%ah		# mask processor family
	movb %ah,X86
	andb $0xf0,%al		# mask model
	shrb $4,%al
	movb %al,X86_MODEL
	andb $0x0f,%cl		# mask mask revision
	movb %cl,X86_MASK
	movl %edx,X86_CAPABILITY

452 453
is486:
	movl $0x50022,%ecx	# set AM, WP, NE and MP
454
	movl %cr0,%eax
L
Linus Torvalds 已提交
455 456 457 458
	andl $0x80000011,%eax	# Save PG,PE,ET
	orl %ecx,%eax
	movl %eax,%cr0

459
	lgdt early_gdt_descr
L
Linus Torvalds 已提交
460 461 462 463 464 465 466 467 468
	lidt idt_descr
	ljmp $(__KERNEL_CS),$1f
1:	movl $(__KERNEL_DS),%eax	# reload all the segment registers
	movl %eax,%ss			# after changing gdt.

	movl $(__USER_DS),%eax		# DS/ES contains default USER segment
	movl %eax,%ds
	movl %eax,%es

469 470 471
	movl $(__KERNEL_PERCPU), %eax
	movl %eax,%fs			# set this cpu's percpu

472
	movl $(__KERNEL_STACK_CANARY),%eax
473
	movl %eax,%gs
474 475

	xorl %eax,%eax			# Clear LDT
L
Linus Torvalds 已提交
476
	lldt %ax
477

478
	pushl $0		# fake return address for unwinder
G
Glauber Costa 已提交
479
	jmp *(initial_code)
L
Linus Torvalds 已提交
480

481 482
#include "verify_cpu.S"

L
Linus Torvalds 已提交
483
/*
484
 *  setup_once
L
Linus Torvalds 已提交
485
 *
486
 *  The setup work we only want to run on the BSP.
L
Linus Torvalds 已提交
487 488 489
 *
 *  Warning: %esi is live across this function.
 */
490 491 492
__INIT
setup_once:
	/*
493 494 495 496 497
	 * Set up a idt with 256 interrupt gates that push zero if there
	 * is no error code and then jump to early_idt_handler_common.
	 * It doesn't actually load the idt - that needs to be done on
	 * each CPU. Interrupts are enabled elsewhere, when we can be
	 * relatively sure everything is ok.
498
	 */
L
Linus Torvalds 已提交
499

500
	movl $idt_table,%edi
501
	movl $early_idt_handler_array,%eax
502 503
	movl $NUM_EXCEPTION_VECTORS,%ecx
1:
L
Linus Torvalds 已提交
504
	movl %eax,(%edi)
505 506 507
	movl %eax,4(%edi)
	/* interrupt gate, dpl=0, present */
	movl $(0x8E000000 + __KERNEL_CS),2(%edi)
508
	addl $EARLY_IDT_HANDLER_SIZE,%eax
L
Linus Torvalds 已提交
509
	addl $8,%edi
510
	loop 1b
511

512 513
	movl $256 - NUM_EXCEPTION_VECTORS,%ecx
	movl $ignore_int,%edx
514
	movl $(__KERNEL_CS << 16),%eax
515
	movw %dx,%ax		/* selector = 0x0010 = cs */
516
	movw $0x8E00,%dx	/* interrupt gate - dpl=0, present */
517 518 519 520 521
2:
	movl %eax,(%edi)
	movl %edx,4(%edi)
	addl $8,%edi
	loop 2b
522

523 524 525 526 527 528 529 530 531 532 533 534 535
#ifdef CONFIG_CC_STACKPROTECTOR
	/*
	 * Configure the stack canary. The linker can't handle this by
	 * relocation.  Manually set base address in stack canary
	 * segment descriptor.
	 */
	movl $gdt_page,%eax
	movl $stack_canary,%ecx
	movw %cx, 8 * GDT_ENTRY_STACK_CANARY + 2(%eax)
	shrl $16, %ecx
	movb %cl, 8 * GDT_ENTRY_STACK_CANARY + 4(%eax)
	movb %ch, 8 * GDT_ENTRY_STACK_CANARY + 7(%eax)
#endif
536

537
	andl $0,setup_once_ref	/* Once is enough, thanks */
L
Linus Torvalds 已提交
538 539
	ret

540
ENTRY(early_idt_handler_array)
541 542 543 544 545 546
	# 36(%esp) %eflags
	# 32(%esp) %cs
	# 28(%esp) %eip
	# 24(%rsp) error code
	i = 0
	.rept NUM_EXCEPTION_VECTORS
547
	.ifeq (EXCEPTION_ERRCODE_MASK >> i) & 1
548 549 550
	pushl $0		# Dummy error code, to make stack frame uniform
	.endif
	pushl $i		# 20(%esp) Vector number
551
	jmp early_idt_handler_common
552
	i = i + 1
553
	.fill early_idt_handler_array + i*EARLY_IDT_HANDLER_SIZE - ., 1, 0xcc
554
	.endr
555
ENDPROC(early_idt_handler_array)
556
	
557 558 559 560 561
early_idt_handler_common:
	/*
	 * The stack is the hardware frame, an error code or zero, and the
	 * vector number.
	 */
562
	cld
563

564 565 566
	cmpl $2,%ss:early_recursion_flag
	je hlt_loop
	incl %ss:early_recursion_flag
567

568
	/* The vector number is in pt_regs->gs */
569

570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597
	cld
	pushl	%fs		/* pt_regs->fs */
	movw	$0, 2(%esp)	/* clear high bits (some CPUs leave garbage) */
	pushl	%es		/* pt_regs->es */
	movw	$0, 2(%esp)	/* clear high bits (some CPUs leave garbage) */
	pushl	%ds		/* pt_regs->ds */
	movw	$0, 2(%esp)	/* clear high bits (some CPUs leave garbage) */
	pushl	%eax		/* pt_regs->ax */
	pushl	%ebp		/* pt_regs->bp */
	pushl	%edi		/* pt_regs->di */
	pushl	%esi		/* pt_regs->si */
	pushl	%edx		/* pt_regs->dx */
	pushl	%ecx		/* pt_regs->cx */
	pushl	%ebx		/* pt_regs->bx */

	/* Fix up DS and ES */
	movl	$(__KERNEL_DS), %ecx
	movl	%ecx, %ds
	movl	%ecx, %es

	/* Load the vector number into EDX */
	movl	PT_GS(%esp), %edx

	/* Load GS into pt_regs->gs and clear high bits */
	movw	%gs, PT_GS(%esp)
	movw	$0, PT_GS+2(%esp)

	cmpl $(__KERNEL_CS),PT_CS(%esp)
598
	jne 10f
599

600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618
	movl	%esp, %eax	/* args are pt_regs (EAX), trapnr (EDX) */
	call	early_fixup_exception
	andl	%eax,%eax
	jz	10f		/* Exception wasn't fixed up */

	popl	%ebx		/* pt_regs->bx */
	popl	%ecx		/* pt_regs->cx */
	popl	%edx		/* pt_regs->dx */
	popl	%esi		/* pt_regs->si */
	popl	%edi		/* pt_regs->di */
	popl	%ebp		/* pt_regs->bp */
	popl	%eax		/* pt_regs->ax */
	popl	%ds		/* pt_regs->ds */
	popl	%es		/* pt_regs->es */
	popl	%fs		/* pt_regs->fs */
	popl	%gs		/* pt_regs->gs */
	decl	%ss:early_recursion_flag
	addl	$4, %esp	/* pop pt_regs->orig_ax */
	iret
619

620
10:
621
#ifdef CONFIG_PRINTK
622
	xorl %eax,%eax
623 624 625
	movw %ax,PT_FS+2(%esp)	/* clean up the segment values on some cpus */
	movw %ax,PT_DS+2(%esp)
	movw %ax,PT_ES+2(%esp)
626 627 628 629 630 631
	leal  40(%esp),%eax
	pushl %eax		/* %esp before the exception */
	pushl %ebx
	pushl %ebp
	pushl %esi
	pushl %edi
632 633
	movl %cr2,%eax
	pushl %eax
634
	pushl (20+6*4)(%esp)	/* trapno */
635 636 637
	pushl $fault_msg
	call printk
#endif
638
	call dump_stack
639 640 641
hlt_loop:
	hlt
	jmp hlt_loop
642
ENDPROC(early_idt_handler_common)
643

L
Linus Torvalds 已提交
644 645 646 647
/* This is the default interrupt "handler" :-) */
	ALIGN
ignore_int:
	cld
M
Matt Mackall 已提交
648
#ifdef CONFIG_PRINTK
L
Linus Torvalds 已提交
649 650 651 652 653 654 655 656
	pushl %eax
	pushl %ecx
	pushl %edx
	pushl %es
	pushl %ds
	movl $(__KERNEL_DS),%eax
	movl %eax,%ds
	movl %eax,%es
657 658 659
	cmpl $2,early_recursion_flag
	je hlt_loop
	incl early_recursion_flag
L
Linus Torvalds 已提交
660 661 662 663 664 665
	pushl 16(%esp)
	pushl 24(%esp)
	pushl 32(%esp)
	pushl 40(%esp)
	pushl $int_msg
	call printk
666 667 668

	call dump_stack

L
Linus Torvalds 已提交
669 670 671 672 673 674
	addl $(5*4),%esp
	popl %ds
	popl %es
	popl %edx
	popl %ecx
	popl %eax
M
Matt Mackall 已提交
675
#endif
L
Linus Torvalds 已提交
676
	iret
677 678 679 680 681
ENDPROC(ignore_int)
__INITDATA
	.align 4
early_recursion_flag:
	.long 0
L
Linus Torvalds 已提交
682

683 684
__REFDATA
	.align 4
T
Thomas Gleixner 已提交
685 686
ENTRY(initial_code)
	.long i386_start_kernel
687 688
ENTRY(setup_once_ref)
	.long setup_once
T
Thomas Gleixner 已提交
689

L
Linus Torvalds 已提交
690 691 692
/*
 * BSS section
 */
693
__PAGE_ALIGNED_BSS
694
	.align PAGE_SIZE
695
#ifdef CONFIG_X86_PAE
696
initial_pg_pmd:
697 698
	.fill 1024*KPMDS,4,0
#else
699
ENTRY(initial_page_table)
L
Linus Torvalds 已提交
700
	.fill 1024,4,0
701
#endif
702
initial_pg_fixmap:
E
Eric W. Biderman 已提交
703
	.fill 1024,4,0
L
Linus Torvalds 已提交
704 705
ENTRY(empty_zero_page)
	.fill 4096,1,0
706 707
ENTRY(swapper_pg_dir)
	.fill 1024,4,0
708

L
Linus Torvalds 已提交
709 710 711
/*
 * This starts the data section.
 */
712
#ifdef CONFIG_X86_PAE
713
__PAGE_ALIGNED_DATA
714
	/* Page-aligned for the benefit of paravirt? */
715
	.align PAGE_SIZE
716 717
ENTRY(initial_page_table)
	.long	pa(initial_pg_pmd+PGD_IDENT_ATTR),0	/* low identity map */
718
# if KPMDS == 3
719 720 721
	.long	pa(initial_pg_pmd+PGD_IDENT_ATTR),0
	.long	pa(initial_pg_pmd+PGD_IDENT_ATTR+0x1000),0
	.long	pa(initial_pg_pmd+PGD_IDENT_ATTR+0x2000),0
722 723
# elif KPMDS == 2
	.long	0,0
724 725
	.long	pa(initial_pg_pmd+PGD_IDENT_ATTR),0
	.long	pa(initial_pg_pmd+PGD_IDENT_ATTR+0x1000),0
726 727 728
# elif KPMDS == 1
	.long	0,0
	.long	0,0
729
	.long	pa(initial_pg_pmd+PGD_IDENT_ATTR),0
730 731 732
# else
#  error "Kernel PMDs should be 1, 2 or 3"
# endif
733
	.align PAGE_SIZE		/* needs to be page-sized too */
734 735
#endif

L
Linus Torvalds 已提交
736
.data
737
.balign 4
L
Linus Torvalds 已提交
738 739 740
ENTRY(stack_start)
	.long init_thread_union+THREAD_SIZE

741
__INITRODATA
L
Linus Torvalds 已提交
742
int_msg:
743
	.asciz "Unknown interrupt or fault at: %p %p %p\n"
L
Linus Torvalds 已提交
744

745
fault_msg:
V
Vegard Nossum 已提交
746 747
/* fault info: */
	.ascii "BUG: Int %d: CR2 %p\n"
748 749 750 751
/* regs pushed in early_idt_handler: */
	.ascii "     EDI %p  ESI %p  EBP %p  EBX %p\n"
	.ascii "     ESP %p   ES %p   DS %p\n"
	.ascii "     EDX %p  ECX %p  EAX %p\n"
V
Vegard Nossum 已提交
752
/* fault frame: */
753
	.ascii "     vec %p  err %p  EIP %p   CS %p  flg %p\n"
V
Vegard Nossum 已提交
754 755 756
	.ascii "Stack: %p %p %p %p %p %p %p %p\n"
	.ascii "       %p %p %p %p %p %p %p %p\n"
	.asciz "       %p %p %p %p %p %p %p %p\n"
757

T
Thomas Gleixner 已提交
758
#include "../../x86/xen/xen-head.S"
759

L
Linus Torvalds 已提交
760 761 762 763 764 765 766
/*
 * The IDT and GDT 'descriptors' are a strange 48-bit object
 * only used by the lidt and lgdt instructions. They are not
 * like usual segment descriptors - they consist of a 16-bit
 * segment size, and 32-bit linear address value:
 */

767
	.data
L
Linus Torvalds 已提交
768 769 770 771 772 773 774 775
.globl boot_gdt_descr
.globl idt_descr

	ALIGN
# early boot GDT descriptor (must use 1:1 address mapping)
	.word 0				# 32 bit align gdt_desc.address
boot_gdt_descr:
	.word __BOOT_DS+7
776
	.long boot_gdt - __PAGE_OFFSET
L
Linus Torvalds 已提交
777 778 779 780 781 782 783 784

	.word 0				# 32-bit align idt_desc.address
idt_descr:
	.word IDT_ENTRIES*8-1		# idt contains 256 entries
	.long idt_table

# boot GDT descriptor (later on used by CPU#0):
	.word 0				# 32 bit align gdt_desc.address
785
ENTRY(early_gdt_descr)
L
Linus Torvalds 已提交
786
	.word GDT_ENTRIES*8-1
R
Rusty Russell 已提交
787
	.long gdt_page			/* Overwritten for secondary CPUs */
L
Linus Torvalds 已提交
788 789

/*
790
 * The boot_gdt must mirror the equivalent in setup.S and is
L
Linus Torvalds 已提交
791 792 793
 * used only for booting.
 */
	.align L1_CACHE_BYTES
794
ENTRY(boot_gdt)
L
Linus Torvalds 已提交
795 796 797
	.fill GDT_ENTRY_BOOT_CS,8,0
	.quad 0x00cf9a000000ffff	/* kernel 4GB code at 0x00000000 */
	.quad 0x00cf92000000ffff	/* kernel 4GB data at 0x00000000 */