intel.c 23.4 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5
#include <linux/kernel.h>

#include <linux/string.h>
#include <linux/bitops.h>
#include <linux/smp.h>
I
Ingo Molnar 已提交
6
#include <linux/sched.h>
L
Linus Torvalds 已提交
7
#include <linux/thread_info.h>
N
Nick Piggin 已提交
8
#include <linux/module.h>
A
Alan Cox 已提交
9
#include <linux/uaccess.h>
L
Linus Torvalds 已提交
10

11
#include <asm/cpufeature.h>
12
#include <asm/pgtable.h>
L
Linus Torvalds 已提交
13
#include <asm/msr.h>
14
#include <asm/bugs.h>
15
#include <asm/cpu.h>
L
Linus Torvalds 已提交
16

17
#ifdef CONFIG_X86_64
A
Alan Cox 已提交
18
#include <linux/topology.h>
19 20
#endif

L
Linus Torvalds 已提交
21 22 23 24 25 26 27
#include "cpu.h"

#ifdef CONFIG_X86_LOCAL_APIC
#include <asm/mpspec.h>
#include <asm/apic.h>
#endif

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
/*
 * Just in case our CPU detection goes bad, or you have a weird system,
 * allow a way to override the automatic disabling of MPX.
 */
static int forcempx;

static int __init forcempx_setup(char *__unused)
{
	forcempx = 1;

	return 1;
}
__setup("intel-skd-046-workaround=disable", forcempx_setup);

void check_mpx_erratum(struct cpuinfo_x86 *c)
{
	if (forcempx)
		return;
	/*
	 * Turn off the MPX feature on CPUs where SMEP is not
	 * available or disabled.
	 *
	 * Works around Intel Erratum SKD046: "Branch Instructions
	 * May Initialize MPX Bound Registers Incorrectly".
	 *
	 * This might falsely disable MPX on systems without
	 * SMEP, like Atom processors without SMEP.  But there
	 * is no such hardware known at the moment.
	 */
	if (cpu_has(c, X86_FEATURE_MPX) && !cpu_has(c, X86_FEATURE_SMEP)) {
		setup_clear_cpu_cap(X86_FEATURE_MPX);
		pr_warn("x86/mpx: Disabling MPX since SMEP not present\n");
	}
}

63
static void early_init_intel(struct cpuinfo_x86 *c)
L
Linus Torvalds 已提交
64
{
65 66
	u64 misc_enable;

67
	/* Unmask CPUID levels if masked: */
68
	if (c->x86 > 6 || (c->x86 == 6 && c->x86_model >= 0xd)) {
69 70
		if (msr_clear_bit(MSR_IA32_MISC_ENABLE,
				  MSR_IA32_MISC_ENABLE_LIMIT_CPUID_BIT) > 0) {
71
			c->cpuid_level = cpuid_eax(0);
72
			get_cpu_cap(c);
73
		}
74 75
	}

76 77 78
	if ((c->x86 == 0xf && c->x86_model >= 0x03) ||
		(c->x86 == 0x6 && c->x86_model >= 0x0e))
		set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC);
79

80 81 82 83 84 85 86 87 88
	if (c->x86 >= 6 && !cpu_has(c, X86_FEATURE_IA64)) {
		unsigned lower_word;

		wrmsr(MSR_IA32_UCODE_REV, 0, 0);
		/* Required by the SDM */
		sync_core();
		rdmsr(MSR_IA32_UCODE_REV, lower_word, c->microcode);
	}

89 90 91 92 93 94 95 96
	/*
	 * Atom erratum AAE44/AAF40/AAG38/AAH41:
	 *
	 * A race condition between speculative fetches and invalidating
	 * a large page.  This is worked around in microcode, but we
	 * need the microcode to have already been loaded... so if it is
	 * not, recommend a BIOS update and disable large pages.
	 */
97 98
	if (c->x86 == 6 && c->x86_model == 0x1c && c->x86_mask <= 2 &&
	    c->microcode < 0x20e) {
99
		pr_warn("Atom PSE erratum detected, BIOS microcode update recommended\n");
100
		clear_cpu_cap(c, X86_FEATURE_PSE);
101 102
	}

103 104 105 106 107 108 109
#ifdef CONFIG_X86_64
	set_cpu_cap(c, X86_FEATURE_SYSENTER32);
#else
	/* Netburst reports 64 bytes clflush size, but does IO in 128 bytes */
	if (c->x86 == 15 && c->x86_cache_alignment == 64)
		c->x86_cache_alignment = 128;
#endif
110

111 112 113 114 115
	/* CPUID workaround for 0F33/0F34 CPU */
	if (c->x86 == 0xF && c->x86_model == 0x3
	    && (c->x86_mask == 0x3 || c->x86_mask == 0x4))
		c->x86_phys_bits = 36;

116 117
	/*
	 * c->x86_power is 8000_0007 edx. Bit 8 is TSC runs at constant rate
I
Ingo Molnar 已提交
118 119 120 121
	 * with P/T states and does not stop in deep C-states.
	 *
	 * It is also reliable across cores and sockets. (but not across
	 * cabinets - we turn it off in that case explicitly.)
122 123 124 125
	 */
	if (c->x86_power & (1 << 8)) {
		set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC);
		set_cpu_cap(c, X86_FEATURE_NONSTOP_TSC);
126
		if (!check_tsc_unstable())
127
			set_sched_clock_stable();
128 129
	}

130 131 132 133 134
	/* Penwell and Cloverview have the TSC which doesn't sleep on S3 */
	if (c->x86 == 6) {
		switch (c->x86_model) {
		case 0x27:	/* Penwell */
		case 0x35:	/* Cloverview */
135
		case 0x4a:	/* Merrifield */
136 137 138 139 140 141 142
			set_cpu_cap(c, X86_FEATURE_NONSTOP_TSC_S3);
			break;
		default:
			break;
		}
	}

143 144 145 146 147 148 149 150 151 152 153 154
	/*
	 * There is a known erratum on Pentium III and Core Solo
	 * and Core Duo CPUs.
	 * " Page with PAT set to WC while associated MTRR is UC
	 *   may consolidate to UC "
	 * Because of this erratum, it is better to stick with
	 * setting WC in MTRR rather than using PAT on these CPUs.
	 *
	 * Enable PAT WC only on P4, Core 2 or later CPUs.
	 */
	if (c->x86 == 6 && c->x86_model < 15)
		clear_cpu_cap(c, X86_FEATURE_PAT);
V
Vegard Nossum 已提交
155 156 157 158 159 160 161 162 163 164

#ifdef CONFIG_KMEMCHECK
	/*
	 * P4s have a "fast strings" feature which causes single-
	 * stepping REP instructions to only generate a #DB on
	 * cache-line boundaries.
	 *
	 * Ingo Molnar reported a Pentium D (model 6) and a Xeon
	 * (model 2) with the same problem.
	 */
165
	if (c->x86 == 15)
166 167
		if (msr_clear_bit(MSR_IA32_MISC_ENABLE,
				  MSR_IA32_MISC_ENABLE_FAST_STRING_BIT) > 0)
168
			pr_info("kmemcheck: Disabling fast string operations\n");
V
Vegard Nossum 已提交
169
#endif
170 171 172 173 174 175 176 177

	/*
	 * If fast string is not enabled in IA32_MISC_ENABLE for any reason,
	 * clear the fast string and enhanced fast string CPU capabilities.
	 */
	if (c->x86 > 6 || (c->x86 == 6 && c->x86_model >= 0xd)) {
		rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable);
		if (!(misc_enable & MSR_IA32_MISC_ENABLE_FAST_STRING)) {
178
			pr_info("Disabled fast string operations\n");
179 180 181 182
			setup_clear_cpu_cap(X86_FEATURE_REP_GOOD);
			setup_clear_cpu_cap(X86_FEATURE_ERMS);
		}
	}
183 184 185 186 187 188 189 190 191 192 193 194 195 196 197

	/*
	 * Intel Quark Core DevMan_001.pdf section 6.4.11
	 * "The operating system also is required to invalidate (i.e., flush)
	 *  the TLB when any changes are made to any of the page table entries.
	 *  The operating system must reload CR3 to cause the TLB to be flushed"
	 *
	 * As a result cpu_has_pge() in arch/x86/include/asm/tlbflush.h should
	 * be false so that __flush_tlb_all() causes CR3 insted of CR4.PGE
	 * to be modified
	 */
	if (c->x86 == 5 && c->x86_model == 9) {
		pr_info("Disabling PGE capability bit\n");
		setup_clear_cpu_cap(X86_FEATURE_PGE);
	}
198 199 200 201 202 203 204 205 206 207 208 209 210

	if (c->cpuid_level >= 0x00000001) {
		u32 eax, ebx, ecx, edx;

		cpuid(0x00000001, &eax, &ebx, &ecx, &edx);
		/*
		 * If HTT (EDX[28]) is set EBX[16:23] contain the number of
		 * apicids which are reserved per package. Store the resulting
		 * shift value for the package management code.
		 */
		if (edx & (1U << 28))
			c->x86_coreid_bits = get_count_order((ebx >> 16) & 0xff);
	}
211 212

	check_mpx_erratum(c);
L
Linus Torvalds 已提交
213 214
}

215
#ifdef CONFIG_X86_32
L
Linus Torvalds 已提交
216 217 218 219 220
/*
 *	Early probe support logic for ppro memory erratum #50
 *
 *	This is called before we do cpu ident work
 */
221

222
int ppro_with_ram_bug(void)
L
Linus Torvalds 已提交
223 224 225 226 227 228
{
	/* Uses data from early_cpu_detect now */
	if (boot_cpu_data.x86_vendor == X86_VENDOR_INTEL &&
	    boot_cpu_data.x86 == 6 &&
	    boot_cpu_data.x86_model == 1 &&
	    boot_cpu_data.x86_mask < 8) {
229
		pr_info("Pentium Pro with Errata#50 detected. Taking evasive action.\n");
L
Linus Torvalds 已提交
230 231 232 233
		return 1;
	}
	return 0;
}
234

235
static void intel_smp_check(struct cpuinfo_x86 *c)
236 237
{
	/* calling is from identify_secondary_cpu() ? */
238
	if (!c->cpu_index)
239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254
		return;

	/*
	 * Mask B, Pentium, but not Pentium MMX
	 */
	if (c->x86 == 5 &&
	    c->x86_mask >= 1 && c->x86_mask <= 4 &&
	    c->x86_model <= 3) {
		/*
		 * Remember we have B step Pentia with bugs
		 */
		WARN_ONCE(1, "WARNING: SMP operation may be unreliable"
				    "with B stepping processors.\n");
	}
}

255 256 257 258 259 260 261 262
static int forcepae;
static int __init forcepae_setup(char *__unused)
{
	forcepae = 1;
	return 1;
}
__setup("forcepae", forcepae_setup);

263
static void intel_workarounds(struct cpuinfo_x86 *c)
L
Linus Torvalds 已提交
264
{
265 266
#ifdef CONFIG_X86_F00F_BUG
	/*
267
	 * All models of Pentium and Pentium with MMX technology CPUs
A
Alan Cox 已提交
268
	 * have the F0 0F bug, which lets nonprivileged users lock up the
269
	 * system. Announce that the fault handler will be checking for it.
270
	 * The Quark is also family 5, but does not have the same bug.
271
	 */
272
	clear_cpu_bug(c, X86_BUG_F00F);
273
	if (!paravirt_enabled() && c->x86 == 5 && c->x86_model < 9) {
274 275
		static int f00f_workaround_enabled;

276
		set_cpu_bug(c, X86_BUG_F00F);
277
		if (!f00f_workaround_enabled) {
278
			pr_notice("Intel Pentium with F0 0F bug - workaround enabled.\n");
279 280 281 282 283 284 285 286 287 288 289 290
			f00f_workaround_enabled = 1;
		}
	}
#endif

	/*
	 * SEP CPUID bug: Pentium Pro reports SEP but doesn't have it until
	 * model 3 mask 3
	 */
	if ((c->x86<<8 | c->x86_model<<4 | c->x86_mask) < 0x633)
		clear_cpu_cap(c, X86_FEATURE_SEP);

291 292 293 294 295 296
	/*
	 * PAE CPUID issue: many Pentium M report no PAE but may have a
	 * functionally usable PAE implementation.
	 * Forcefully enable PAE if kernel parameter "forcepae" is present.
	 */
	if (forcepae) {
297
		pr_warn("PAE forced!\n");
298 299 300 301
		set_cpu_cap(c, X86_FEATURE_PAE);
		add_taint(TAINT_CPU_OUT_OF_SPEC, LOCKDEP_NOW_UNRELIABLE);
	}

302 303 304 305
	/*
	 * P4 Xeon errata 037 workaround.
	 * Hardware prefetcher may cause stale data to be loaded into the cache.
	 */
L
Linus Torvalds 已提交
306
	if ((c->x86 == 15) && (c->x86_model == 1) && (c->x86_mask == 1)) {
307 308 309
		if (msr_set_bit(MSR_IA32_MISC_ENABLE,
				MSR_IA32_MISC_ENABLE_PREFETCH_DISABLE_BIT)
		    > 0) {
310 311
			pr_info("CPU: C0 stepping P4 Xeon detected.\n");
			pr_info("CPU: Disabling hardware prefetching (Errata 037)\n");
L
Linus Torvalds 已提交
312 313 314
		}
	}

315 316 317 318 319 320 321 322
	/*
	 * See if we have a good local APIC by checking for buggy Pentia,
	 * i.e. all B steppings and the C2 stepping of P54C when using their
	 * integrated APIC (see 11AP erratum in "Pentium Processor
	 * Specification Update").
	 */
	if (cpu_has_apic && (c->x86<<8 | c->x86_model<<4) == 0x520 &&
	    (c->x86_mask < 0x6 || c->x86_mask == 0xb))
323
		set_cpu_bug(c, X86_BUG_11AP);
324 325


326
#ifdef CONFIG_X86_INTEL_USERCOPY
327
	/*
328
	 * Set up the preferred alignment for movsl bulk memory moves
329
	 */
330 331 332 333 334 335 336 337 338 339 340 341
	switch (c->x86) {
	case 4:		/* 486: untested */
		break;
	case 5:		/* Old Pentia: untested */
		break;
	case 6:		/* PII/PIII only like movsl with 8-byte alignment */
		movsl_mask.mask = 7;
		break;
	case 15:	/* P4 is OK down to 8-byte alignment */
		movsl_mask.mask = 7;
		break;
	}
342
#endif
343

344
	intel_smp_check(c);
345 346
}
#else
347
static void intel_workarounds(struct cpuinfo_x86 *c)
348 349
{
}
350 351
#endif

352
static void srat_detect_node(struct cpuinfo_x86 *c)
353
{
354
#ifdef CONFIG_NUMA
355 356 357 358 359
	unsigned node;
	int cpu = smp_processor_id();

	/* Don't do the funky fallback heuristics the AMD version employs
	   for now. */
360
	node = numa_cpu_node(cpu);
361
	if (node == NUMA_NO_NODE || !node_online(node)) {
362 363 364
		/* reuse the value from init_cpu_to_node() */
		node = cpu_to_node(cpu);
	}
365 366 367 368
	numa_set_node(cpu, node);
#endif
}

369 370 371
/*
 * find out the number of processor cores on the die
 */
372
static int intel_num_cpu_cores(struct cpuinfo_x86 *c)
373
{
Z
Zachary Amsden 已提交
374
	unsigned int eax, ebx, ecx, edx;
375

376
	if (!IS_ENABLED(CONFIG_SMP) || c->cpuid_level < 4)
377 378
		return 1;

Z
Zachary Amsden 已提交
379 380
	/* Intel has a non-standard dependency on %ecx for this CPUID level. */
	cpuid_count(4, 0, &eax, &ebx, &ecx, &edx);
381
	if (eax & 0x1f)
A
Alan Cox 已提交
382
		return (eax >> 26) + 1;
383 384 385 386
	else
		return 1;
}

387
static void detect_vmx_virtcap(struct cpuinfo_x86 *c)
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
{
	/* Intel VMX MSR indicated features */
#define X86_VMX_FEATURE_PROC_CTLS_TPR_SHADOW	0x00200000
#define X86_VMX_FEATURE_PROC_CTLS_VNMI		0x00400000
#define X86_VMX_FEATURE_PROC_CTLS_2ND_CTLS	0x80000000
#define X86_VMX_FEATURE_PROC_CTLS2_VIRT_APIC	0x00000001
#define X86_VMX_FEATURE_PROC_CTLS2_EPT		0x00000002
#define X86_VMX_FEATURE_PROC_CTLS2_VPID		0x00000020

	u32 vmx_msr_low, vmx_msr_high, msr_ctl, msr_ctl2;

	clear_cpu_cap(c, X86_FEATURE_TPR_SHADOW);
	clear_cpu_cap(c, X86_FEATURE_VNMI);
	clear_cpu_cap(c, X86_FEATURE_FLEXPRIORITY);
	clear_cpu_cap(c, X86_FEATURE_EPT);
	clear_cpu_cap(c, X86_FEATURE_VPID);

	rdmsr(MSR_IA32_VMX_PROCBASED_CTLS, vmx_msr_low, vmx_msr_high);
	msr_ctl = vmx_msr_high | vmx_msr_low;
	if (msr_ctl & X86_VMX_FEATURE_PROC_CTLS_TPR_SHADOW)
		set_cpu_cap(c, X86_FEATURE_TPR_SHADOW);
	if (msr_ctl & X86_VMX_FEATURE_PROC_CTLS_VNMI)
		set_cpu_cap(c, X86_FEATURE_VNMI);
	if (msr_ctl & X86_VMX_FEATURE_PROC_CTLS_2ND_CTLS) {
		rdmsr(MSR_IA32_VMX_PROCBASED_CTLS2,
		      vmx_msr_low, vmx_msr_high);
		msr_ctl2 = vmx_msr_high | vmx_msr_low;
		if ((msr_ctl2 & X86_VMX_FEATURE_PROC_CTLS2_VIRT_APIC) &&
		    (msr_ctl & X86_VMX_FEATURE_PROC_CTLS_TPR_SHADOW))
			set_cpu_cap(c, X86_FEATURE_FLEXPRIORITY);
		if (msr_ctl2 & X86_VMX_FEATURE_PROC_CTLS2_EPT)
			set_cpu_cap(c, X86_FEATURE_EPT);
		if (msr_ctl2 & X86_VMX_FEATURE_PROC_CTLS2_VPID)
			set_cpu_cap(c, X86_FEATURE_VPID);
	}
}

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
static void init_intel_energy_perf(struct cpuinfo_x86 *c)
{
	u64 epb;

	/*
	 * Initialize MSR_IA32_ENERGY_PERF_BIAS if not already initialized.
	 * (x86_energy_perf_policy(8) is available to change it at run-time.)
	 */
	if (!cpu_has(c, X86_FEATURE_EPB))
		return;

	rdmsrl(MSR_IA32_ENERGY_PERF_BIAS, epb);
	if ((epb & 0xF) != ENERGY_PERF_BIAS_PERFORMANCE)
		return;

	pr_warn_once("ENERGY_PERF_BIAS: Set to 'normal', was 'performance'\n");
	pr_warn_once("ENERGY_PERF_BIAS: View and update with x86_energy_perf_policy(8)\n");
	epb = (epb & ~0xF) | ENERGY_PERF_BIAS_NORMAL;
	wrmsrl(MSR_IA32_ENERGY_PERF_BIAS, epb);
}

static void intel_bsp_resume(struct cpuinfo_x86 *c)
{
	/*
	 * MSR_IA32_ENERGY_PERF_BIAS is lost across suspend/resume,
	 * so reinitialize it properly like during bootup:
	 */
	init_intel_energy_perf(c);
}

455
static void init_intel(struct cpuinfo_x86 *c)
L
Linus Torvalds 已提交
456 457 458
{
	unsigned int l2 = 0;

459 460
	early_init_intel(c);

461
	intel_workarounds(c);
L
Linus Torvalds 已提交
462

463 464 465 466 467 468 469
	/*
	 * Detect the extended topology information if available. This
	 * will reinitialise the initial_apicid which will be used
	 * in init_intel_cacheinfo()
	 */
	detect_extended_topology(c);

470 471 472 473 474 475 476 477 478 479 480
	if (!cpu_has(c, X86_FEATURE_XTOPOLOGY)) {
		/*
		 * let's use the legacy cpuid vector 0x1 and 0x4 for topology
		 * detection.
		 */
		c->x86_max_cores = intel_num_cpu_cores(c);
#ifdef CONFIG_X86_32
		detect_ht(c);
#endif
	}

L
Linus Torvalds 已提交
481
	l2 = init_intel_cacheinfo(c);
482 483 484 485 486 487 488

	/* Detect legacy cache sizes if init_intel_cacheinfo did not */
	if (l2 == 0) {
		cpu_detect_cache_sizes(c);
		l2 = c->x86_cache_size;
	}

489
	if (c->cpuid_level > 9) {
490 491 492
		unsigned eax = cpuid_eax(10);
		/* Check for version and the number of counters */
		if ((eax & 0xff) && (((eax>>8) & 0xff) > 1))
493
			set_cpu_cap(c, X86_FEATURE_ARCH_PERFMON);
494
	}
L
Linus Torvalds 已提交
495

496 497
	if (cpu_has_xmm2)
		set_cpu_cap(c, X86_FEATURE_LFENCE_RDTSC);
498 499

	if (boot_cpu_has(X86_FEATURE_DS)) {
500 501 502 503 504 505 506
		unsigned int l1;
		rdmsr(MSR_IA32_MISC_ENABLE, l1, l2);
		if (!(l1 & (1<<11)))
			set_cpu_cap(c, X86_FEATURE_BTS);
		if (!(l1 & (1<<12)))
			set_cpu_cap(c, X86_FEATURE_PEBS);
	}
L
Linus Torvalds 已提交
507

508 509
	if (c->x86 == 6 && cpu_has_clflush &&
	    (c->x86_model == 29 || c->x86_model == 46 || c->x86_model == 47))
510
		set_cpu_bug(c, X86_BUG_CLFLUSH_MONITOR);
511

512 513 514 515 516 517
#ifdef CONFIG_X86_64
	if (c->x86 == 15)
		c->x86_cache_alignment = c->x86_clflush_size * 2;
	if (c->x86 == 6)
		set_cpu_cap(c, X86_FEATURE_REP_GOOD);
#else
518 519 520 521 522
	/*
	 * Names for the Pentium II/Celeron processors
	 * detectable only by also checking the cache size.
	 * Dixon is NOT a Celeron.
	 */
L
Linus Torvalds 已提交
523
	if (c->x86 == 6) {
524 525
		char *p = NULL;

L
Linus Torvalds 已提交
526 527
		switch (c->x86_model) {
		case 5:
528 529 530 531
			if (l2 == 0)
				p = "Celeron (Covington)";
			else if (l2 == 256)
				p = "Mobile Pentium II (Dixon)";
L
Linus Torvalds 已提交
532
			break;
533

L
Linus Torvalds 已提交
534 535 536 537 538 539
		case 6:
			if (l2 == 128)
				p = "Celeron (Mendocino)";
			else if (c->x86_mask == 0 || c->x86_mask == 5)
				p = "Celeron-A";
			break;
540

L
Linus Torvalds 已提交
541 542 543 544 545 546
		case 8:
			if (l2 == 128)
				p = "Celeron (Coppermine)";
			break;
		}

547 548
		if (p)
			strcpy(c->x86_model_id, p);
L
Linus Torvalds 已提交
549 550
	}

551 552 553 554
	if (c->x86 == 15)
		set_cpu_cap(c, X86_FEATURE_P4);
	if (c->x86 == 6)
		set_cpu_cap(c, X86_FEATURE_P3);
555
#endif
556 557

	/* Work around errata */
558
	srat_detect_node(c);
559 560 561

	if (cpu_has(c, X86_FEATURE_VMX))
		detect_vmx_virtcap(c);
562

563
	init_intel_energy_perf(c);
564
}
L
Linus Torvalds 已提交
565

566
#ifdef CONFIG_X86_32
567
static unsigned int intel_size_cache(struct cpuinfo_x86 *c, unsigned int size)
L
Linus Torvalds 已提交
568
{
569 570
	/*
	 * Intel PIII Tualatin. This comes in two flavours.
L
Linus Torvalds 已提交
571 572 573 574 575 576
	 * One has 256kb of cache, the other 512. We have no way
	 * to determine which, so we use a boottime override
	 * for the 512kb model, and assume 256 otherwise.
	 */
	if ((c->x86 == 6) && (c->x86_model == 11) && (size == 0))
		size = 256;
577 578 579 580 581 582 583

	/*
	 * Intel Quark SoC X1000 contains a 4-way set associative
	 * 16K cache with a 16 byte cache line and 256 lines per tag
	 */
	if ((c->x86 == 5) && (c->x86_model == 9))
		size = 16;
L
Linus Torvalds 已提交
584 585
	return size;
}
586
#endif
L
Linus Torvalds 已提交
587

588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606
#define TLB_INST_4K	0x01
#define TLB_INST_4M	0x02
#define TLB_INST_2M_4M	0x03

#define TLB_INST_ALL	0x05
#define TLB_INST_1G	0x06

#define TLB_DATA_4K	0x11
#define TLB_DATA_4M	0x12
#define TLB_DATA_2M_4M	0x13
#define TLB_DATA_4K_4M	0x14

#define TLB_DATA_1G	0x16

#define TLB_DATA0_4K	0x21
#define TLB_DATA0_4M	0x22
#define TLB_DATA0_2M_4M	0x23

#define STLB_4K		0x41
607
#define STLB_4K_2M	0x42
608

609
static const struct _tlb_table intel_tlb_table[] = {
610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627
	{ 0x01, TLB_INST_4K,		32,	" TLB_INST 4 KByte pages, 4-way set associative" },
	{ 0x02, TLB_INST_4M,		2,	" TLB_INST 4 MByte pages, full associative" },
	{ 0x03, TLB_DATA_4K,		64,	" TLB_DATA 4 KByte pages, 4-way set associative" },
	{ 0x04, TLB_DATA_4M,		8,	" TLB_DATA 4 MByte pages, 4-way set associative" },
	{ 0x05, TLB_DATA_4M,		32,	" TLB_DATA 4 MByte pages, 4-way set associative" },
	{ 0x0b, TLB_INST_4M,		4,	" TLB_INST 4 MByte pages, 4-way set associative" },
	{ 0x4f, TLB_INST_4K,		32,	" TLB_INST 4 KByte pages */" },
	{ 0x50, TLB_INST_ALL,		64,	" TLB_INST 4 KByte and 2-MByte or 4-MByte pages" },
	{ 0x51, TLB_INST_ALL,		128,	" TLB_INST 4 KByte and 2-MByte or 4-MByte pages" },
	{ 0x52, TLB_INST_ALL,		256,	" TLB_INST 4 KByte and 2-MByte or 4-MByte pages" },
	{ 0x55, TLB_INST_2M_4M,		7,	" TLB_INST 2-MByte or 4-MByte pages, fully associative" },
	{ 0x56, TLB_DATA0_4M,		16,	" TLB_DATA0 4 MByte pages, 4-way set associative" },
	{ 0x57, TLB_DATA0_4K,		16,	" TLB_DATA0 4 KByte pages, 4-way associative" },
	{ 0x59, TLB_DATA0_4K,		16,	" TLB_DATA0 4 KByte pages, fully associative" },
	{ 0x5a, TLB_DATA0_2M_4M,	32,	" TLB_DATA0 2-MByte or 4 MByte pages, 4-way set associative" },
	{ 0x5b, TLB_DATA_4K_4M,		64,	" TLB_DATA 4 KByte and 4 MByte pages" },
	{ 0x5c, TLB_DATA_4K_4M,		128,	" TLB_DATA 4 KByte and 4 MByte pages" },
	{ 0x5d, TLB_DATA_4K_4M,		256,	" TLB_DATA 4 KByte and 4 MByte pages" },
628 629 630
	{ 0x61, TLB_INST_4K,		48,	" TLB_INST 4 KByte pages, full associative" },
	{ 0x63, TLB_DATA_1G,		4,	" TLB_DATA 1 GByte pages, 4-way set associative" },
	{ 0x76, TLB_INST_2M_4M,		8,	" TLB_INST 2-MByte or 4-MByte pages, fully associative" },
631 632 633 634 635
	{ 0xb0, TLB_INST_4K,		128,	" TLB_INST 4 KByte pages, 4-way set associative" },
	{ 0xb1, TLB_INST_2M_4M,		4,	" TLB_INST 2M pages, 4-way, 8 entries or 4M pages, 4-way entries" },
	{ 0xb2, TLB_INST_4K,		64,	" TLB_INST 4KByte pages, 4-way set associative" },
	{ 0xb3, TLB_DATA_4K,		128,	" TLB_DATA 4 KByte pages, 4-way set associative" },
	{ 0xb4, TLB_DATA_4K,		256,	" TLB_DATA 4 KByte pages, 4-way associative" },
636 637
	{ 0xb5, TLB_INST_4K,		64,	" TLB_INST 4 KByte pages, 8-way set associative" },
	{ 0xb6, TLB_INST_4K,		128,	" TLB_INST 4 KByte pages, 8-way set associative" },
638 639
	{ 0xba, TLB_DATA_4K,		64,	" TLB_DATA 4 KByte pages, 4-way associative" },
	{ 0xc0, TLB_DATA_4K_4M,		8,	" TLB_DATA 4 KByte and 4 MByte pages, 4-way associative" },
640 641
	{ 0xc1, STLB_4K_2M,		1024,	" STLB 4 KByte and 2 MByte pages, 8-way associative" },
	{ 0xc2, TLB_DATA_2M_4M,		16,	" DTLB 2 MByte/4MByte pages, 4-way associative" },
642 643 644 645
	{ 0xca, STLB_4K,		512,	" STLB 4 KByte pages, 4-way associative" },
	{ 0x00, 0, 0 }
};

646
static void intel_tlb_lookup(const unsigned char desc)
647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666
{
	unsigned char k;
	if (desc == 0)
		return;

	/* look up this descriptor in the table */
	for (k = 0; intel_tlb_table[k].descriptor != desc && \
			intel_tlb_table[k].descriptor != 0; k++)
		;

	if (intel_tlb_table[k].tlb_type == 0)
		return;

	switch (intel_tlb_table[k].tlb_type) {
	case STLB_4K:
		if (tlb_lli_4k[ENTRIES] < intel_tlb_table[k].entries)
			tlb_lli_4k[ENTRIES] = intel_tlb_table[k].entries;
		if (tlb_lld_4k[ENTRIES] < intel_tlb_table[k].entries)
			tlb_lld_4k[ENTRIES] = intel_tlb_table[k].entries;
		break;
667 668 669 670 671 672 673 674 675 676 677 678 679 680
	case STLB_4K_2M:
		if (tlb_lli_4k[ENTRIES] < intel_tlb_table[k].entries)
			tlb_lli_4k[ENTRIES] = intel_tlb_table[k].entries;
		if (tlb_lld_4k[ENTRIES] < intel_tlb_table[k].entries)
			tlb_lld_4k[ENTRIES] = intel_tlb_table[k].entries;
		if (tlb_lli_2m[ENTRIES] < intel_tlb_table[k].entries)
			tlb_lli_2m[ENTRIES] = intel_tlb_table[k].entries;
		if (tlb_lld_2m[ENTRIES] < intel_tlb_table[k].entries)
			tlb_lld_2m[ENTRIES] = intel_tlb_table[k].entries;
		if (tlb_lli_4m[ENTRIES] < intel_tlb_table[k].entries)
			tlb_lli_4m[ENTRIES] = intel_tlb_table[k].entries;
		if (tlb_lld_4m[ENTRIES] < intel_tlb_table[k].entries)
			tlb_lld_4m[ENTRIES] = intel_tlb_table[k].entries;
		break;
681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725
	case TLB_INST_ALL:
		if (tlb_lli_4k[ENTRIES] < intel_tlb_table[k].entries)
			tlb_lli_4k[ENTRIES] = intel_tlb_table[k].entries;
		if (tlb_lli_2m[ENTRIES] < intel_tlb_table[k].entries)
			tlb_lli_2m[ENTRIES] = intel_tlb_table[k].entries;
		if (tlb_lli_4m[ENTRIES] < intel_tlb_table[k].entries)
			tlb_lli_4m[ENTRIES] = intel_tlb_table[k].entries;
		break;
	case TLB_INST_4K:
		if (tlb_lli_4k[ENTRIES] < intel_tlb_table[k].entries)
			tlb_lli_4k[ENTRIES] = intel_tlb_table[k].entries;
		break;
	case TLB_INST_4M:
		if (tlb_lli_4m[ENTRIES] < intel_tlb_table[k].entries)
			tlb_lli_4m[ENTRIES] = intel_tlb_table[k].entries;
		break;
	case TLB_INST_2M_4M:
		if (tlb_lli_2m[ENTRIES] < intel_tlb_table[k].entries)
			tlb_lli_2m[ENTRIES] = intel_tlb_table[k].entries;
		if (tlb_lli_4m[ENTRIES] < intel_tlb_table[k].entries)
			tlb_lli_4m[ENTRIES] = intel_tlb_table[k].entries;
		break;
	case TLB_DATA_4K:
	case TLB_DATA0_4K:
		if (tlb_lld_4k[ENTRIES] < intel_tlb_table[k].entries)
			tlb_lld_4k[ENTRIES] = intel_tlb_table[k].entries;
		break;
	case TLB_DATA_4M:
	case TLB_DATA0_4M:
		if (tlb_lld_4m[ENTRIES] < intel_tlb_table[k].entries)
			tlb_lld_4m[ENTRIES] = intel_tlb_table[k].entries;
		break;
	case TLB_DATA_2M_4M:
	case TLB_DATA0_2M_4M:
		if (tlb_lld_2m[ENTRIES] < intel_tlb_table[k].entries)
			tlb_lld_2m[ENTRIES] = intel_tlb_table[k].entries;
		if (tlb_lld_4m[ENTRIES] < intel_tlb_table[k].entries)
			tlb_lld_4m[ENTRIES] = intel_tlb_table[k].entries;
		break;
	case TLB_DATA_4K_4M:
		if (tlb_lld_4k[ENTRIES] < intel_tlb_table[k].entries)
			tlb_lld_4k[ENTRIES] = intel_tlb_table[k].entries;
		if (tlb_lld_4m[ENTRIES] < intel_tlb_table[k].entries)
			tlb_lld_4m[ENTRIES] = intel_tlb_table[k].entries;
		break;
726 727 728
	case TLB_DATA_1G:
		if (tlb_lld_1g[ENTRIES] < intel_tlb_table[k].entries)
			tlb_lld_1g[ENTRIES] = intel_tlb_table[k].entries;
729 730 731 732
		break;
	}
}

733
static void intel_detect_tlb(struct cpuinfo_x86 *c)
734 735 736 737
{
	int i, j, n;
	unsigned int regs[4];
	unsigned char *desc = (unsigned char *)regs;
738 739 740 741

	if (c->cpuid_level < 2)
		return;

742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758
	/* Number of times to iterate */
	n = cpuid_eax(2) & 0xFF;

	for (i = 0 ; i < n ; i++) {
		cpuid(2, &regs[0], &regs[1], &regs[2], &regs[3]);

		/* If bit 31 is set, this is an unknown format */
		for (j = 0 ; j < 3 ; j++)
			if (regs[j] & (1 << 31))
				regs[j] = 0;

		/* Byte 0 is level count, not a descriptor */
		for (j = 1 ; j < 16 ; j++)
			intel_tlb_lookup(desc[j]);
	}
}

759
static const struct cpu_dev intel_cpu_dev = {
L
Linus Torvalds 已提交
760
	.c_vendor	= "Intel",
761
	.c_ident	= { "GenuineIntel" },
762
#ifdef CONFIG_X86_32
763 764
	.legacy_models = {
		{ .family = 4, .model_names =
765 766 767 768 769 770 771 772 773
		  {
			  [0] = "486 DX-25/33",
			  [1] = "486 DX-50",
			  [2] = "486 SX",
			  [3] = "486 DX/2",
			  [4] = "486 SL",
			  [5] = "486 SX/2",
			  [7] = "486 DX/2-WB",
			  [8] = "486 DX/4",
L
Linus Torvalds 已提交
774 775 776
			  [9] = "486 DX/4-WB"
		  }
		},
777
		{ .family = 5, .model_names =
778 779 780
		  {
			  [0] = "Pentium 60/66 A-step",
			  [1] = "Pentium 60/66",
L
Linus Torvalds 已提交
781
			  [2] = "Pentium 75 - 200",
782
			  [3] = "OverDrive PODP5V83",
L
Linus Torvalds 已提交
783
			  [4] = "Pentium MMX",
784
			  [7] = "Mobile Pentium 75 - 200",
785 786
			  [8] = "Mobile Pentium MMX",
			  [9] = "Quark SoC X1000",
L
Linus Torvalds 已提交
787 788
		  }
		},
789
		{ .family = 6, .model_names =
790
		  {
L
Linus Torvalds 已提交
791
			  [0] = "Pentium Pro A-step",
792 793 794 795
			  [1] = "Pentium Pro",
			  [3] = "Pentium II (Klamath)",
			  [4] = "Pentium II (Deschutes)",
			  [5] = "Pentium II (Deschutes)",
L
Linus Torvalds 已提交
796
			  [6] = "Mobile Pentium II",
797 798
			  [7] = "Pentium III (Katmai)",
			  [8] = "Pentium III (Coppermine)",
L
Linus Torvalds 已提交
799 800 801 802
			  [10] = "Pentium III (Cascades)",
			  [11] = "Pentium III (Tualatin)",
		  }
		},
803
		{ .family = 15, .model_names =
L
Linus Torvalds 已提交
804 805 806 807 808 809 810 811 812
		  {
			  [0] = "Pentium 4 (Unknown)",
			  [1] = "Pentium 4 (Willamette)",
			  [2] = "Pentium 4 (Northwood)",
			  [4] = "Pentium 4 (Foster)",
			  [5] = "Pentium 4 (Foster)",
		  }
		},
	},
813
	.legacy_cache_size = intel_size_cache,
814
#endif
815
	.c_detect_tlb	= intel_detect_tlb,
816
	.c_early_init   = early_init_intel,
L
Linus Torvalds 已提交
817
	.c_init		= init_intel,
818
	.c_bsp_resume	= intel_bsp_resume,
Y
Yinghai Lu 已提交
819
	.c_x86_vendor	= X86_VENDOR_INTEL,
L
Linus Torvalds 已提交
820 821
};

Y
Yinghai Lu 已提交
822
cpu_dev_register(intel_cpu_dev);
L
Linus Torvalds 已提交
823