summaryrefslogtreecommitdiffstats
path: root/arch/x86/cpu/cpu.c
blob: a4e639dcf3379f26613aa7ab0888905488f2e848 (plain)
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
26
27
28
29
30
31
32
33
34
35
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
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
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
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
/*
 * (C) Copyright 2008-2011
 * Graeme Russ, <graeme.russ@gmail.com>
 *
 * (C) Copyright 2002
 * Daniel Engström, Omicron Ceti AB, <daniel@omicron.se>
 *
 * (C) Copyright 2002
 * Sysgo Real-Time Solutions, GmbH <www.elinos.com>
 * Marius Groeger <mgroeger@sysgo.de>
 *
 * (C) Copyright 2002
 * Sysgo Real-Time Solutions, GmbH <www.elinos.com>
 * Alex Zuepke <azu@sysgo.de>
 *
 * Part of this file is adapted from coreboot
 * src/arch/x86/lib/cpu.c
 *
 * SPDX-License-Identifier:	GPL-2.0+
 */

#include <common.h>
#include <command.h>
#include <errno.h>
#include <malloc.h>
#include <asm/control_regs.h>
#include <asm/cpu.h>
#include <asm/processor.h>
#include <asm/processor-flags.h>
#include <asm/interrupt.h>
#include <linux/compiler.h>

DECLARE_GLOBAL_DATA_PTR;

/*
 * Constructor for a conventional segment GDT (or LDT) entry
 * This is a macro so it can be used in initialisers
 */
#define GDT_ENTRY(flags, base, limit)			\
	((((base)  & 0xff000000ULL) << (56-24)) |	\
	 (((flags) & 0x0000f0ffULL) << 40) |		\
	 (((limit) & 0x000f0000ULL) << (48-16)) |	\
	 (((base)  & 0x00ffffffULL) << 16) |		\
	 (((limit) & 0x0000ffffULL)))

struct gdt_ptr {
	u16 len;
	u32 ptr;
} __packed;

struct cpu_device_id {
	unsigned vendor;
	unsigned device;
};

struct cpuinfo_x86 {
	uint8_t x86;            /* CPU family */
	uint8_t x86_vendor;     /* CPU vendor */
	uint8_t x86_model;
	uint8_t x86_mask;
};

/*
 * List of cpu vendor strings along with their normalized
 * id values.
 */
static struct {
	int vendor;
	const char *name;
} x86_vendors[] = {
	{ X86_VENDOR_INTEL,     "GenuineIntel", },
	{ X86_VENDOR_CYRIX,     "CyrixInstead", },
	{ X86_VENDOR_AMD,       "AuthenticAMD", },
	{ X86_VENDOR_UMC,       "UMC UMC UMC ", },
	{ X86_VENDOR_NEXGEN,    "NexGenDriven", },
	{ X86_VENDOR_CENTAUR,   "CentaurHauls", },
	{ X86_VENDOR_RISE,      "RiseRiseRise", },
	{ X86_VENDOR_TRANSMETA, "GenuineTMx86", },
	{ X86_VENDOR_TRANSMETA, "TransmetaCPU", },
	{ X86_VENDOR_NSC,       "Geode by NSC", },
	{ X86_VENDOR_SIS,       "SiS SiS SiS ", },
};

static const char *const x86_vendor_name[] = {
	[X86_VENDOR_INTEL]     = "Intel",
	[X86_VENDOR_CYRIX]     = "Cyrix",
	[X86_VENDOR_AMD]       = "AMD",
	[X86_VENDOR_UMC]       = "UMC",
	[X86_VENDOR_NEXGEN]    = "NexGen",
	[X86_VENDOR_CENTAUR]   = "Centaur",
	[X86_VENDOR_RISE]      = "Rise",
	[X86_VENDOR_TRANSMETA] = "Transmeta",
	[X86_VENDOR_NSC]       = "NSC",
	[X86_VENDOR_SIS]       = "SiS",
};

static void load_ds(u32 segment)
{
	asm volatile("movl %0, %%ds" : : "r" (segment * X86_GDT_ENTRY_SIZE));
}

static void load_es(u32 segment)
{
	asm volatile("movl %0, %%es" : : "r" (segment * X86_GDT_ENTRY_SIZE));
}

static void load_fs(u32 segment)
{
	asm volatile("movl %0, %%fs" : : "r" (segment * X86_GDT_ENTRY_SIZE));
}

static void load_gs(u32 segment)
{
	asm volatile("movl %0, %%gs" : : "r" (segment * X86_GDT_ENTRY_SIZE));
}

static void load_ss(u32 segment)
{
	asm volatile("movl %0, %%ss" : : "r" (segment * X86_GDT_ENTRY_SIZE));
}

static void load_gdt(const u64 *boot_gdt, u16 num_entries)
{
	struct gdt_ptr gdt;

	gdt.len = (num_entries * 8) - 1;
	gdt.ptr = (u32)boot_gdt;

	asm volatile("lgdtl %0\n" : : "m" (gdt));
}

void setup_gdt(gd_t *id, u64 *gdt_addr)
{
	/* CS: code, read/execute, 4 GB, base 0 */
	gdt_addr[X86_GDT_ENTRY_32BIT_CS] = GDT_ENTRY(0xc09b, 0, 0xfffff);

	/* DS: data, read/write, 4 GB, base 0 */
	gdt_addr[X86_GDT_ENTRY_32BIT_DS] = GDT_ENTRY(0xc093, 0, 0xfffff);

	/* FS: data, read/write, 4 GB, base (Global Data Pointer) */
	id->arch.gd_addr = id;
	gdt_addr[X86_GDT_ENTRY_32BIT_FS] = GDT_ENTRY(0xc093,
		     (ulong)&id->arch.gd_addr, 0xfffff);

	/* 16-bit CS: code, read/execute, 64 kB, base 0 */
	gdt_addr[X86_GDT_ENTRY_16BIT_CS] = GDT_ENTRY(0x109b, 0, 0x0ffff);

	/* 16-bit DS: data, read/write, 64 kB, base 0 */
	gdt_addr[X86_GDT_ENTRY_16BIT_DS] = GDT_ENTRY(0x1093, 0, 0x0ffff);

	load_gdt(gdt_addr, X86_GDT_NUM_ENTRIES);
	load_ds(X86_GDT_ENTRY_32BIT_DS);
	load_es(X86_GDT_ENTRY_32BIT_DS);
	load_gs(X86_GDT_ENTRY_32BIT_DS);
	load_ss(X86_GDT_ENTRY_32BIT_DS);
	load_fs(X86_GDT_ENTRY_32BIT_FS);
}

int __weak x86_cleanup_before_linux(void)
{
#ifdef CONFIG_BOOTSTAGE_STASH
	bootstage_stash((void *)CONFIG_BOOTSTAGE_STASH,
			CONFIG_BOOTSTAGE_STASH_SIZE);
#endif

	return 0;
}

/*
 * Cyrix CPUs without cpuid or with cpuid not yet enabled can be detected
 * by the fact that they preserve the flags across the division of 5/2.
 * PII and PPro exhibit this behavior too, but they have cpuid available.
 */

/*
 * Perform the Cyrix 5/2 test. A Cyrix won't change
 * the flags, while other 486 chips will.
 */
static inline int test_cyrix_52div(void)
{
	unsigned int test;

	__asm__ __volatile__(
	     "sahf\n\t"		/* clear flags (%eax = 0x0005) */
	     "div %b2\n\t"	/* divide 5 by 2 */
	     "lahf"		/* store flags into %ah */
	     : "=a" (test)
	     : "0" (5), "q" (2)
	     : "cc");

	/* AH is 0x02 on Cyrix after the divide.. */
	return (unsigned char) (test >> 8) == 0x02;
}

/*
 *	Detect a NexGen CPU running without BIOS hypercode new enough
 *	to have CPUID. (Thanks to Herbert Oppmann)
 */

static int deep_magic_nexgen_probe(void)
{
	int ret;

	__asm__ __volatile__ (
		"	movw	$0x5555, %%ax\n"
		"	xorw	%%dx,%%dx\n"
		"	movw	$2, %%cx\n"
		"	divw	%%cx\n"
		"	movl	$0, %%eax\n"
		"	jnz	1f\n"
		"	movl	$1, %%eax\n"
		"1:\n"
		: "=a" (ret) : : "cx", "dx");
	return  ret;
}

static bool has_cpuid(void)
{
	return flag_is_changeable_p(X86_EFLAGS_ID);
}

static int build_vendor_name(char *vendor_name)
{
	struct cpuid_result result;
	result = cpuid(0x00000000);
	unsigned int *name_as_ints = (unsigned int *)vendor_name;

	name_as_ints[0] = result.ebx;
	name_as_ints[1] = result.edx;
	name_as_ints[2] = result.ecx;

	return result.eax;
}

static void identify_cpu(struct cpu_device_id *cpu)
{
	char vendor_name[16];
	int i;

	vendor_name[0] = '\0'; /* Unset */

	/* Find the id and vendor_name */
	if (!has_cpuid()) {
		/* Its a 486 if we can modify the AC flag */
		if (flag_is_changeable_p(X86_EFLAGS_AC))
			cpu->device = 0x00000400; /* 486 */
		else
			cpu->device = 0x00000300; /* 386 */
		if ((cpu->device == 0x00000400) && test_cyrix_52div()) {
			memcpy(vendor_name, "CyrixInstead", 13);
			/* If we ever care we can enable cpuid here */
		}
		/* Detect NexGen with old hypercode */
		else if (deep_magic_nexgen_probe())
			memcpy(vendor_name, "NexGenDriven", 13);
	}
	if (has_cpuid()) {
		int  cpuid_level;

		cpuid_level = build_vendor_name(vendor_name);
		vendor_name[12] = '\0';

		/* Intel-defined flags: level 0x00000001 */
		if (cpuid_level >= 0x00000001) {
			cpu->device = cpuid_eax(0x00000001);
		} else {
			/* Have CPUID level 0 only unheard of */
			cpu->device = 0x00000400;
		}
	}
	cpu->vendor = X86_VENDOR_UNKNOWN;
	for (i = 0; i < ARRAY_SIZE(x86_vendors); i++) {
		if (memcmp(vendor_name, x86_vendors[i].name, 12) == 0) {
			cpu->vendor = x86_vendors[i].vendor;
			break;
		}
	}
}

static inline void get_fms(struct cpuinfo_x86 *c, uint32_t tfms)
{
	c->x86 = (tfms >> 8) & 0xf;
	c->x86_model = (tfms >> 4) & 0xf;
	c->x86_mask = tfms & 0xf;
	if (c->x86 == 0xf)
		c->x86 += (tfms >> 20) & 0xff;
	if (c->x86 >= 0x6)
		c->x86_model += ((tfms >> 16) & 0xF) << 4;
}

int x86_cpu_init_f(void)
{
	const u32 em_rst = ~X86_CR0_EM;
	const u32 mp_ne_set = X86_CR0_MP | X86_CR0_NE;

	/* initialize FPU, reset EM, set MP and NE */
	asm ("fninit\n" \
	     "movl %%cr0, %%eax\n" \
	     "andl %0, %%eax\n" \
	     "orl  %1, %%eax\n" \
	     "movl %%eax, %%cr0\n" \
	     : : "i" (em_rst), "i" (mp_ne_set) : "eax");

	/* identify CPU via cpuid and store the decoded info into gd->arch */
	if (has_cpuid()) {
		struct cpu_device_id cpu;
		struct cpuinfo_x86 c;

		identify_cpu(&cpu);
		get_fms(&c, cpu.device);
		gd->arch.x86 = c.x86;
		gd->arch.x86_vendor = cpu.vendor;
		gd->arch.x86_model = c.x86_model;
		gd->arch.x86_mask = c.x86_mask;
		gd->arch.x86_device = cpu.device;
	}

	return 0;
}

int x86_cpu_init_r(void)
{
	/* Initialize core interrupt and exception functionality of CPU */
	cpu_init_interrupts();
	return 0;
}
int cpu_init_r(void) __attribute__((weak, alias("x86_cpu_init_r")));

void x86_enable_caches(void)
{
	unsigned long cr0;

	cr0 = read_cr0();
	cr0 &= ~(X86_CR0_NW | X86_CR0_CD);
	write_cr0(cr0);
	wbinvd();
}
void enable_caches(void) __attribute__((weak, alias("x86_enable_caches")));

void x86_disable_caches(void)
{
	unsigned long cr0;

	cr0 = read_cr0();
	cr0 |= X86_CR0_NW | X86_CR0_CD;
	wbinvd();
	write_cr0(cr0);
	wbinvd();
}
void disable_caches(void) __attribute__((weak, alias("x86_disable_caches")));

int x86_init_cache(void)
{
	enable_caches();

	return 0;
}
int init_cache(void) __attribute__((weak, alias("x86_init_cache")));

int do_reset(cmd_tbl_t *cmdtp, int flag, int argc, char * const argv[])
{
	printf("resetting ...\n");

	/* wait 50 ms */
	udelay(50000);
	disable_interrupts();
	reset_cpu(0);

	/*NOTREACHED*/
	return 0;
}

void  flush_cache(unsigned long dummy1, unsigned long dummy2)
{
	asm("wbinvd\n");
}

void __attribute__ ((regparm(0))) generate_gpf(void);

/* segment 0x70 is an arbitrary segment which does not exist */
asm(".globl generate_gpf\n"
	".hidden generate_gpf\n"
	".type generate_gpf, @function\n"
	"generate_gpf:\n"
	"ljmp   $0x70, $0x47114711\n");

__weak void reset_cpu(ulong addr)
{
	printf("Resetting using x86 Triple Fault\n");
	set_vector(13, generate_gpf);	/* general protection fault handler */
	set_vector(8, generate_gpf);	/* double fault handler */
	generate_gpf();			/* start the show */
}

int dcache_status(void)
{
	return !(read_cr0() & 0x40000000);
}

/* Define these functions to allow ehch-hcd to function */
void flush_dcache_range(unsigned long start, unsigned long stop)
{
}

void invalidate_dcache_range(unsigned long start, unsigned long stop)
{
}

void dcache_enable(void)
{
	enable_caches();
}

void dcache_disable(void)
{
	disable_caches();
}

void icache_enable(void)
{
}

void icache_disable(void)
{
}

int icache_status(void)
{
	return 1;
}

void cpu_enable_paging_pae(ulong cr3)
{
	__asm__ __volatile__(
		/* Load the page table address */
		"movl	%0, %%cr3\n"
		/* Enable pae */
		"movl	%%cr4, %%eax\n"
		"orl	$0x00000020, %%eax\n"
		"movl	%%eax, %%cr4\n"
		/* Enable paging */
		"movl	%%cr0, %%eax\n"
		"orl	$0x80000000, %%eax\n"
		"movl	%%eax, %%cr0\n"
		:
		: "r" (cr3)
		: "eax");
}

void cpu_disable_paging_pae(void)
{
	/* Turn off paging */
	__asm__ __volatile__ (
		/* Disable paging */
		"movl	%%cr0, %%eax\n"
		"andl	$0x7fffffff, %%eax\n"
		"movl	%%eax, %%cr0\n"
		/* Disable pae */
		"movl	%%cr4, %%eax\n"
		"andl	$0xffffffdf, %%eax\n"
		"movl	%%eax, %%cr4\n"
		:
		:
		: "eax");
}

static bool can_detect_long_mode(void)
{
	return cpuid_eax(0x80000000) > 0x80000000UL;
}

static bool has_long_mode(void)
{
	return cpuid_edx(0x80000001) & (1 << 29) ? true : false;
}

int cpu_has_64bit(void)
{
	return has_cpuid() && can_detect_long_mode() &&
		has_long_mode();
}

const char *cpu_vendor_name(int vendor)
{
	const char *name;
	name = "<invalid cpu vendor>";
	if ((vendor < (ARRAY_SIZE(x86_vendor_name))) &&
	    (x86_vendor_name[vendor] != 0))
		name = x86_vendor_name[vendor];

	return name;
}

void fill_processor_name(char *processor_name)
{
	struct cpuid_result regs;
	char temp_processor_name[49];
	char *processor_name_start;
	unsigned int *name_as_ints = (unsigned int *)temp_processor_name;
	int i;

	for (i = 0; i < 3; i++) {
		regs = cpuid(0x80000002 + i);
		name_as_ints[i * 4 + 0] = regs.eax;
		name_as_ints[i * 4 + 1] = regs.ebx;
		name_as_ints[i * 4 + 2] = regs.ecx;
		name_as_ints[i * 4 + 3] = regs.edx;
	}

	temp_processor_name[48] = 0;

	/* Skip leading spaces. */
	processor_name_start = temp_processor_name;
	while (*processor_name_start == ' ')
		processor_name_start++;

	memset(processor_name, 0, 49);
	strcpy(processor_name, processor_name_start);
}

int print_cpuinfo(void)
{
	printf("CPU: %s, vendor %s, device %xh\n",
	       cpu_has_64bit() ? "x86_64" : "x86",
	       cpu_vendor_name(gd->arch.x86_vendor), gd->arch.x86_device);

	return 0;
}

#define PAGETABLE_SIZE		(6 * 4096)

/**
 * build_pagetable() - build a flat 4GiB page table structure for 64-bti mode
 *
 * @pgtable: Pointer to a 24iKB block of memory
 */
static void build_pagetable(uint32_t *pgtable)
{
	uint i;

	memset(pgtable, '\0', PAGETABLE_SIZE);

	/* Level 4 needs a single entry */
	pgtable[0] = (uint32_t)&pgtable[1024] + 7;

	/* Level 3 has one 64-bit entry for each GiB of memory */
	for (i = 0; i < 4; i++) {
		pgtable[1024 + i * 2] = (uint32_t)&pgtable[2048] +
							0x1000 * i + 7;
	}

	/* Level 2 has 2048 64-bit entries, each repesenting 2MiB */
	for (i = 0; i < 2048; i++)
		pgtable[2048 + i * 2] = 0x183 + (i << 21UL);
}

int cpu_jump_to_64bit(ulong setup_base, ulong target)
{
	uint32_t *pgtable;

	pgtable = memalign(4096, PAGETABLE_SIZE);
	if (!pgtable)
		return -ENOMEM;

	build_pagetable(pgtable);
	cpu_call64((ulong)pgtable, setup_base, target);
	free(pgtable);

	return -EFAULT;
}
OpenPOWER on IntegriCloud