static void __cpuinit early_init_centaur(struct cpuinfo_x86 *c)
{
if (c->x86 == 0x6 && c->x86_model >= 0xf)
set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC);
set_cpu_cap(c, X86_FEATURE_SYSENTER32);
}
static void __cpuinit early_init_intel(struct cpuinfo_x86 *c)
{
/* Unmask CPUID levels if masked: */
if (c->x86 > 6 || (c->x86 == 6 && c->x86_model >= 0xd)) {
u64 misc_enable;
rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable);
if (misc_enable & MSR_IA32_MISC_ENABLE_LIMIT_CPUID) {
misc_enable &= ~MSR_IA32_MISC_ENABLE_LIMIT_CPUID;
wrmsrl(MSR_IA32_MISC_ENABLE, misc_enable);
c->cpuid_level = cpuid_eax(0);
}
}
if ((c->x86 == 0xf && c->x86_model >= 0x03) ||
(c->x86 == 0x6 && c->x86_model >= 0x0e))
set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC);
#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
/* 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;
/*
* c->x86_power is 8000_0007 edx. Bit 8 is TSC runs at constant rate
* 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.)
*/
if (c->x86_power & (1 << 8)) {
set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC);
set_cpu_cap(c, X86_FEATURE_NONSTOP_TSC);
set_cpu_cap(c, X86_FEATURE_TSC_RELIABLE);
sched_clock_stable = 1;
}
/*
* 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);
}
static void __cpuinit init_centaur(struct cpuinfo_x86 *c)
{
if (c->x86 == 0x6 && c->x86_model >= 0xf) {
c->x86_cache_alignment = c->x86_clflush_size * 2;
set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC);
set_cpu_cap(c, X86_FEATURE_REP_GOOD);
}
set_cpu_cap(c, X86_FEATURE_LFENCE_RDTSC);
}
static void init_c3(struct cpuinfo_x86 *c)
{
u32 lo, hi;
/* Test for Centaur Extended Feature Flags presence */
if (cpuid_eax(0xC0000000) >= 0xC0000001) {
u32 tmp = cpuid_edx(0xC0000001);
/* enable ACE unit, if present and disabled */
if ((tmp & (ACE_PRESENT | ACE_ENABLED)) == ACE_PRESENT) {
rdmsr(MSR_VIA_FCR, lo, hi);
lo |= ACE_FCR; /* enable ACE unit */
wrmsr(MSR_VIA_FCR, lo, hi);
pr_info("CPU: Enabled ACE h/w crypto\n");
}
/* enable RNG unit, if present and disabled */
if ((tmp & (RNG_PRESENT | RNG_ENABLED)) == RNG_PRESENT) {
rdmsr(MSR_VIA_RNG, lo, hi);
lo |= RNG_ENABLE; /* enable RNG unit */
wrmsr(MSR_VIA_RNG, lo, hi);
pr_info("CPU: Enabled h/w RNG\n");
}
/* store Centaur Extended Feature Flags as
* word 5 of the CPU capability bit array
*/
c->x86_capability[CPUID_C000_0001_EDX] = cpuid_edx(0xC0000001);
}
#ifdef CONFIG_X86_32
/* Cyrix III family needs CX8 & PGE explicitly enabled. */
if (c->x86_model >= 6 && c->x86_model <= 13) {
rdmsr(MSR_VIA_FCR, lo, hi);
lo |= (1<<1 | 1<<7);
wrmsr(MSR_VIA_FCR, lo, hi);
set_cpu_cap(c, X86_FEATURE_CX8);
}
/* Before Nehemiah, the C3's had 3dNOW! */
if (c->x86_model >= 6 && c->x86_model < 9)
set_cpu_cap(c, X86_FEATURE_3DNOW);
#endif
if (c->x86 == 0x6 && c->x86_model >= 0xf) {
c->x86_cache_alignment = c->x86_clflush_size * 2;
set_cpu_cap(c, X86_FEATURE_REP_GOOD);
}
cpu_detect_cache_sizes(c);
}
static void probe_xeon_phi_r3mwait(struct cpuinfo_x86 *c)
{
/*
* Ring 3 MONITOR/MWAIT feature cannot be detected without
* cpu model and family comparison.
*/
if (c->x86 != 6)
return;
switch (c->x86_model) {
case INTEL_FAM6_XEON_PHI_KNL:
case INTEL_FAM6_XEON_PHI_KNM:
break;
default:
return;
}
if (ring3mwait_disabled)
return;
set_cpu_cap(c, X86_FEATURE_RING3MWAIT);
this_cpu_or(msr_misc_features_shadow,
1UL << MSR_MISC_FEATURES_ENABLES_RING3MWAIT_BIT);
if (c == &boot_cpu_data)
ELF_HWCAP2 |= HWCAP2_RING3MWAIT;
}
/*
* The earliest FPU detection code.
*
* Set the X86_FEATURE_FPU CPU-capability bit based on
* trying to execute an actual sequence of FPU instructions:
*/
static void fpu__init_system_early_generic(struct cpuinfo_x86 *c)
{
unsigned long cr0;
u16 fsw, fcw;
fsw = fcw = 0xffff;
cr0 = read_cr0();
cr0 &= ~(X86_CR0_TS | X86_CR0_EM);
write_cr0(cr0);
asm volatile("fninit ; fnstsw %0 ; fnstcw %1"
: "+m" (fsw), "+m" (fcw));
if (fsw == 0 && (fcw & 0x103f) == 0x003f)
set_cpu_cap(c, X86_FEATURE_FPU);
else
clear_cpu_cap(c, X86_FEATURE_FPU);
#ifndef CONFIG_MATH_EMULATION
if (!cpu_has_fpu) {
pr_emerg("x86/fpu: Giving up, no FPU found and no math emulation present\n");
for (;;)
asm volatile("hlt");
}
int detect_extended_topology_early(struct cpuinfo_x86 *c)
{
#ifdef CONFIG_SMP
unsigned int eax, ebx, ecx, edx;
if (c->cpuid_level < 0xb)
return -1;
cpuid_count(0xb, SMT_LEVEL, &eax, &ebx, &ecx, &edx);
/*
* check if the cpuid leaf 0xb is actually implemented.
*/
if (ebx == 0 || (LEAFB_SUBTYPE(ecx) != SMT_TYPE))
return -1;
set_cpu_cap(c, X86_FEATURE_XTOPOLOGY);
/*
* initial apic id, which also represents 32-bit extended x2apic id.
*/
c->initial_apicid = edx;
smp_num_siblings = LEVEL_MAX_SIBLINGS(ebx);
#endif
return 0;
}
void __cpuinit init_scattered_cpuid_features(struct cpuinfo_x86 *c)
{
u32 max_level;
u32 regs[4];
const struct cpuid_bit *cb;
static const struct cpuid_bit cpuid_bits[] = {
{ X86_FEATURE_IDA, CR_EAX, 1, 0x00000006 },
{ 0, 0, 0, 0 }
};
for (cb = cpuid_bits; cb->feature; cb++) {
/* Verify that the level is valid */
max_level = cpuid_eax(cb->level & 0xffff0000);
if (max_level < cb->level ||
max_level > (cb->level | 0xffff))
continue;
cpuid(cb->level, ®s[CR_EAX], ®s[CR_EBX],
®s[CR_ECX], ®s[CR_EDX]);
if (regs[cb->reg] & (1 << cb->bit))
set_cpu_cap(c, cb->feature);
}
}
static void early_init_centaur(struct cpuinfo_x86 *c)
{
switch (c->x86) {
#ifdef CONFIG_X86_32
case 5:
/* Emulate MTRRs using Centaur's MCR. */
set_cpu_cap(c, X86_FEATURE_CENTAUR_MCR);
break;
#endif
case 6:
if (c->x86_model >= 0xf)
set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC);
break;
}
#ifdef CONFIG_X86_64
set_cpu_cap(c, X86_FEATURE_SYSENTER32);
#endif
}
static void __cpuinit early_init_intel(struct cpuinfo_x86 *c)
{
/* 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;
if ((c->x86 == 0xf && c->x86_model >= 0x03) ||
(c->x86 == 0x6 && c->x86_model >= 0x0e))
set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC);
}
void __exit lguest_arch_host_fini(void)
{
/* If we had PGE before we started, turn it back on now. */
get_online_cpus();
if (cpu_had_pge) {
set_cpu_cap(&boot_cpu_data, X86_FEATURE_PGE);
/* adjust_pge's argument "1" means set PGE. */
on_each_cpu(adjust_pge, (void *)1, 1);
}
put_online_cpus();
}
void __exit lguest_arch_host_fini(void)
{
get_online_cpus();
if (cpu_had_pge) {
set_cpu_cap(&boot_cpu_data, X86_FEATURE_PGE);
on_each_cpu(adjust_pge, (void *)1, 1);
}
put_online_cpus();
}
static void __cpuinit early_init_intel(struct cpuinfo_x86 *c)
{
/* Unmask CPUID levels if masked: */
if (c->x86 == 6 && c->x86_model >= 15) {
u64 misc_enable;
rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable);
if (misc_enable & MSR_IA32_MISC_ENABLE_LIMIT_CPUID) {
misc_enable &= ~MSR_IA32_MISC_ENABLE_LIMIT_CPUID;
wrmsrl(MSR_IA32_MISC_ENABLE, misc_enable);
c->cpuid_level = cpuid_eax(0);
}
}
if ((c->x86 == 0xf && c->x86_model >= 0x03) ||
(c->x86 == 0x6 && c->x86_model >= 0x0e))
set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC);
#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
/*
* c->x86_power is 8000_0007 edx. Bit 8 is TSC runs at constant rate
* with P/T states and does not stop in deep C-states
*/
if (c->x86_power & (1 << 8)) {
set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC);
set_cpu_cap(c, X86_FEATURE_NONSTOP_TSC);
}
}
void init_scattered_cpuid_features(struct cpuinfo_x86 *c)
{
u32 max_level;
u32 regs[4];
const struct cpuid_bit *cb;
static const struct cpuid_bit cpuid_bits[] = {
{ X86_FEATURE_DTS, CR_EAX, 0, 0x00000006, 0 },
{ X86_FEATURE_DTHERM, CR_EAX, 0, 0x00000006, 0 },
{ X86_FEATURE_IDA, CR_EAX, 1, 0x00000006, 0 },
{ X86_FEATURE_ARAT, CR_EAX, 2, 0x00000006, 0 },
{ X86_FEATURE_PLN, CR_EAX, 4, 0x00000006, 0 },
{ X86_FEATURE_PTS, CR_EAX, 6, 0x00000006, 0 },
{ X86_FEATURE_APERFMPERF, CR_ECX, 0, 0x00000006, 0 },
{ X86_FEATURE_EPB, CR_ECX, 3, 0x00000006, 0 },
{ X86_FEATURE_XSAVEOPT, CR_EAX, 0, 0x0000000d, 1 },
{ X86_FEATURE_CPB, CR_EDX, 9, 0x80000007, 0 },
{ X86_FEATURE_HW_PSTATE, CR_EDX, 7, 0x80000007, 0 },
{ X86_FEATURE_NPT, CR_EDX, 0, 0x8000000a, 0 },
{ X86_FEATURE_LBRV, CR_EDX, 1, 0x8000000a, 0 },
{ X86_FEATURE_SVML, CR_EDX, 2, 0x8000000a, 0 },
{ X86_FEATURE_NRIPS, CR_EDX, 3, 0x8000000a, 0 },
{ X86_FEATURE_TSCRATEMSR, CR_EDX, 4, 0x8000000a, 0 },
{ X86_FEATURE_VMCBCLEAN, CR_EDX, 5, 0x8000000a, 0 },
{ X86_FEATURE_FLUSHBYASID, CR_EDX, 6, 0x8000000a, 0 },
{ X86_FEATURE_DECODEASSISTS, CR_EDX, 7, 0x8000000a, 0 },
{ X86_FEATURE_PAUSEFILTER, CR_EDX,10, 0x8000000a, 0 },
{ X86_FEATURE_PFTHRESHOLD, CR_EDX,12, 0x8000000a, 0 },
{ 0, 0, 0, 0, 0 }
};
for (cb = cpuid_bits; cb->feature; cb++) {
/* Verify that the level is valid */
max_level = cpuid_eax(cb->level & 0xffff0000);
if (max_level < cb->level ||
max_level > (cb->level | 0xffff))
continue;
cpuid_count(cb->level, cb->sub_leaf, ®s[CR_EAX],
®s[CR_EBX], ®s[CR_ECX], ®s[CR_EDX]);
if (regs[cb->reg] & (1 << cb->bit))
set_cpu_cap(c, cb->feature);
}
}
static void __cpuinit early_init_intel(struct cpuinfo_x86 *c)
{
u64 misc_enable;
/* Unmask CPUID levels if masked: */
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_LIMIT_CPUID) {
misc_enable &= ~MSR_IA32_MISC_ENABLE_LIMIT_CPUID;
wrmsrl(MSR_IA32_MISC_ENABLE, misc_enable);
c->cpuid_level = cpuid_eax(0);
get_cpu_cap(c);
}
}
if ((c->x86 == 0xf && c->x86_model >= 0x03) ||
(c->x86 == 0x6 && c->x86_model >= 0x0e))
set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC);
<<<<<<< HEAD
static void __cpuinit early_init_intel(struct cpuinfo_x86 *c)
{
/* Unmask CPUID levels if masked: */
if (c->x86 == 6 && c->x86_model >= 15) {
u64 misc_enable;
rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable);
if (misc_enable & MSR_IA32_MISC_ENABLE_LIMIT_CPUID) {
misc_enable &= ~MSR_IA32_MISC_ENABLE_LIMIT_CPUID;
wrmsrl(MSR_IA32_MISC_ENABLE, misc_enable);
c->cpuid_level = cpuid_eax(0);
}
}
/* 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;
if ((c->x86 == 0xf && c->x86_model >= 0x03) ||
(c->x86 == 0x6 && c->x86_model >= 0x0e))
set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC);
}
void init_scattered_cpuid_features(struct cpuinfo_x86 *c)
{
u32 max_level;
u32 regs[4];
const struct cpuid_bit *cb;
for (cb = cpuid_bits; cb->feature; cb++) {
/* Verify that the level is valid */
max_level = cpuid_eax(cb->level & 0xffff0000);
if (max_level < cb->level ||
max_level > (cb->level | 0xffff))
continue;
cpuid_count(cb->level, cb->sub_leaf, ®s[CPUID_EAX],
®s[CPUID_EBX], ®s[CPUID_ECX],
®s[CPUID_EDX]);
if (regs[cb->reg] & (1 << cb->bit))
set_cpu_cap(c, cb->feature);
}
}
static void __cpuinit early_init_intel(struct cpuinfo_x86 *c)
{
u64 misc_enable;
/* Unmask CPUID levels if masked: */
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_LIMIT_CPUID) {
misc_enable &= ~MSR_IA32_MISC_ENABLE_LIMIT_CPUID;
wrmsrl(MSR_IA32_MISC_ENABLE, misc_enable);
c->cpuid_level = cpuid_eax(0);
get_cpu_cap(c);
}
}
if ((c->x86 == 0xf && c->x86_model >= 0x03) ||
(c->x86 == 0x6 && c->x86_model >= 0x0e))
set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC);
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);
}
/*
* 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.
*/
if (c->x86 == 6 && c->x86_model == 0x1c && c->x86_mask <= 2 &&
c->microcode < 0x20e) {
printk(KERN_WARNING "Atom PSE erratum detected, BIOS microcode update recommended\n");
clear_cpu_cap(c, X86_FEATURE_PSE);
}
#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
/* 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;
/*
* c->x86_power is 8000_0007 edx. Bit 8 is TSC runs at constant rate
* 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.)
*/
if (c->x86_power & (1 << 8)) {
set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC);
set_cpu_cap(c, X86_FEATURE_NONSTOP_TSC);
if (!check_tsc_unstable())
sched_clock_stable = 1;
}
/*
* 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);
#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.
*/
if (c->x86 == 15) {
rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable);
if (misc_enable & MSR_IA32_MISC_ENABLE_FAST_STRING) {
printk(KERN_INFO "kmemcheck: Disabling fast string operations\n");
misc_enable &= ~MSR_IA32_MISC_ENABLE_FAST_STRING;
wrmsrl(MSR_IA32_MISC_ENABLE, misc_enable);
}
}
#endif
/*
* 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)) {
printk(KERN_INFO "Disabled fast string operations\n");
setup_clear_cpu_cap(X86_FEATURE_REP_GOOD);
setup_clear_cpu_cap(X86_FEATURE_ERMS);
}
}
}
static void __cpuinit intel_workarounds(struct cpuinfo_x86 *c)
{
unsigned long lo, hi;
#ifdef CONFIG_X86_F00F_BUG
/*
* All current models of Pentium and Pentium with MMX technology CPUs
* have the F0 0F bug, which lets nonprivileged users lock up the
* system.
* Note that the workaround only should be initialized once...
*/
c->f00f_bug = 0;
if (!paravirt_enabled() && c->x86 == 5) {
static int f00f_workaround_enabled;
c->f00f_bug = 1;
if (!f00f_workaround_enabled) {
trap_init_f00f_bug();
printk(KERN_NOTICE "Intel Pentium with F0 0F bug - workaround enabled.\n");
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);
/*
* P4 Xeon errata 037 workaround.
* Hardware prefetcher may cause stale data to be loaded into the cache.
*/
if ((c->x86 == 15) && (c->x86_model == 1) && (c->x86_mask == 1)) {
rdmsr(MSR_IA32_MISC_ENABLE, lo, hi);
if ((lo & MSR_IA32_MISC_ENABLE_PREFETCH_DISABLE) == 0) {
printk (KERN_INFO "CPU: C0 stepping P4 Xeon detected.\n");
printk (KERN_INFO "CPU: Disabling hardware prefetching (Errata 037)\n");
lo |= MSR_IA32_MISC_ENABLE_PREFETCH_DISABLE;
wrmsr(MSR_IA32_MISC_ENABLE, lo, hi);
}
}
/*
* 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))
set_cpu_cap(c, X86_FEATURE_11AP);
#ifdef CONFIG_X86_INTEL_USERCOPY
/*
* Set up the preferred alignment for movsl bulk memory moves
*/
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;
}
#endif
#ifdef CONFIG_X86_NUMAQ
numaq_tsc_disable();
#endif
intel_smp_check(c);
}
/*
* VMware hypervisor takes care of exporting a reliable TSC to the guest.
* Still, due to timing difference when running on virtual cpus, the TSC can
* be marked as unstable in some cases. For example, the TSC sync check at
* bootup can fail due to a marginal offset between vcpus' TSCs (though the
* TSCs do not drift from each other). Also, the ACPI PM timer clocksource
* is not suitable as a watchdog when running on a hypervisor because the
* kernel may miss a wrap of the counter if the vcpu is descheduled for a
* long time. To skip these checks at runtime we set these capability bits,
* so that the kernel could just trust the hypervisor with providing a
* reliable virtual TSC that is suitable for timekeeping.
*/
static void vmware_set_cpu_features(struct cpuinfo_x86 *c)
{
set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC);
set_cpu_cap(c, X86_FEATURE_TSC_RELIABLE);
}
static void early_init_intel(struct cpuinfo_x86 *c)
{
u64 misc_enable;
/* Unmask CPUID levels if masked: */
if (c->x86 > 6 || (c->x86 == 6 && c->x86_model >= 0xd)) {
if (msr_clear_bit(MSR_IA32_MISC_ENABLE,
MSR_IA32_MISC_ENABLE_LIMIT_CPUID_BIT) > 0) {
c->cpuid_level = cpuid_eax(0);
get_cpu_cap(c);
}
}
if ((c->x86 == 0xf && c->x86_model >= 0x03) ||
(c->x86 == 0x6 && c->x86_model >= 0x0e))
set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC);
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);
}
/*
* 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.
*/
if (c->x86 == 6 && c->x86_model == 0x1c && c->x86_mask <= 2 &&
c->microcode < 0x20e) {
pr_warn("Atom PSE erratum detected, BIOS microcode update recommended\n");
clear_cpu_cap(c, X86_FEATURE_PSE);
}
#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
/* 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;
/*
* c->x86_power is 8000_0007 edx. Bit 8 is TSC runs at constant rate
* 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.)
*/
if (c->x86_power & (1 << 8)) {
set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC);
set_cpu_cap(c, X86_FEATURE_NONSTOP_TSC);
if (!check_tsc_unstable())
set_sched_clock_stable();
}
/* 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 */
case 0x4a: /* Merrifield */
set_cpu_cap(c, X86_FEATURE_NONSTOP_TSC_S3);
break;
default:
break;
}
}
/*
* 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);
#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.
*/
if (c->x86 == 15)
if (msr_clear_bit(MSR_IA32_MISC_ENABLE,
MSR_IA32_MISC_ENABLE_FAST_STRING_BIT) > 0)
pr_info("kmemcheck: Disabling fast string operations\n");
#endif
/*
* 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)) {
pr_info("Disabled fast string operations\n");
setup_clear_cpu_cap(X86_FEATURE_REP_GOOD);
setup_clear_cpu_cap(X86_FEATURE_ERMS);
}
}
/*
* 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, boot_cpu_has(X86_FEATURE_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);
}
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);
}
/*
* VMware hypervisor takes care of exporting a reliable TSC to the guest.
* Still, due to timing difference when running on virtual cpus, the TSC can
* be marked as unstable in some cases. For example, the TSC sync check at
* bootup can fail due to a marginal offset between vcpus' TSCs (though the
* TSCs do not drift from each other). Also, the ACPI PM timer clocksource
* is not suitable as a watchdog when running on a hypervisor because the
* kernel may miss a wrap of the counter if the vcpu is descheduled for a
* long time. To skip these checks at runtime we set these capability bits,
* so that the kernel could just trust the hypervisor with providing a
* reliable virtual TSC that is suitable for timekeeping.
*/
void __cpuinit vmware_set_feature_bits(struct cpuinfo_x86 *c)
{
set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC);
set_cpu_cap(c, X86_FEATURE_TSC_RELIABLE);
}
static void init_centaur(struct cpuinfo_x86 *c)
{
#ifdef CONFIG_X86_32
char *name;
u32 fcr_set = 0;
u32 fcr_clr = 0;
u32 lo, hi, newlo;
u32 aa, bb, cc, dd;
/*
* Bit 31 in normal CPUID used for nonstandard 3DNow ID;
* 3DNow is IDd by bit 31 in extended CPUID (1*32+31) anyway
*/
clear_cpu_cap(c, 0*32+31);
#endif
early_init_centaur(c);
switch (c->x86) {
#ifdef CONFIG_X86_32
case 5:
switch (c->x86_model) {
case 4:
name = "C6";
fcr_set = ECX8|DSMC|EDCTLB|EMMX|ERETSTK;
fcr_clr = DPDC;
pr_notice("Disabling bugged TSC.\n");
clear_cpu_cap(c, X86_FEATURE_TSC);
break;
case 8:
switch (c->x86_mask) {
default:
name = "2";
break;
case 7 ... 9:
name = "2A";
break;
case 10 ... 15:
name = "2B";
break;
}
fcr_set = ECX8|DSMC|DTLOCK|EMMX|EBRPRED|ERETSTK|
E2MMX|EAMD3D;
fcr_clr = DPDC;
break;
case 9:
name = "3";
fcr_set = ECX8|DSMC|DTLOCK|EMMX|EBRPRED|ERETSTK|
E2MMX|EAMD3D;
fcr_clr = DPDC;
break;
default:
name = "??";
}
rdmsr(MSR_IDT_FCR1, lo, hi);
newlo = (lo|fcr_set) & (~fcr_clr);
if (newlo != lo) {
pr_info("Centaur FCR was 0x%X now 0x%X\n",
lo, newlo);
wrmsr(MSR_IDT_FCR1, newlo, hi);
} else {
pr_info("Centaur FCR is 0x%X\n", lo);
}
/* Emulate MTRRs using Centaur's MCR. */
set_cpu_cap(c, X86_FEATURE_CENTAUR_MCR);
/* Report CX8 */
set_cpu_cap(c, X86_FEATURE_CX8);
/* Set 3DNow! on Winchip 2 and above. */
if (c->x86_model >= 8)
set_cpu_cap(c, X86_FEATURE_3DNOW);
/* See if we can find out some more. */
if (cpuid_eax(0x80000000) >= 0x80000005) {
/* Yes, we can. */
cpuid(0x80000005, &aa, &bb, &cc, &dd);
/* Add L1 data and code cache sizes. */
c->x86_cache_size = (cc>>24)+(dd>>24);
}
sprintf(c->x86_model_id, "WinChip %s", name);
break;
#endif
case 6:
init_c3(c);
break;
}
/*
* Check for extended topology enumeration cpuid leaf 0xb and if it
* exists, use it for populating initial_apicid and cpu topology
* detection.
*/
void __cpuinit detect_extended_topology(struct cpuinfo_x86 *c)
{
#ifdef CONFIG_SMP
unsigned int eax, ebx, ecx, edx, sub_index;
unsigned int ht_mask_width, core_plus_mask_width;
unsigned int core_select_mask, core_level_siblings;
if (c->cpuid_level < 0xb)
return;
cpuid_count(0xb, SMT_LEVEL, &eax, &ebx, &ecx, &edx);
/*
* check if the cpuid leaf 0xb is actually implemented.
*/
if (ebx == 0 || (LEAFB_SUBTYPE(ecx) != SMT_TYPE))
return;
set_cpu_cap(c, X86_FEATURE_XTOPOLOGY);
/*
* initial apic id, which also represents 32-bit extended x2apic id.
*/
c->initial_apicid = edx;
/*
* Populate HT related information from sub-leaf level 0.
*/
core_level_siblings = smp_num_siblings = LEVEL_MAX_SIBLINGS(ebx);
core_plus_mask_width = ht_mask_width = BITS_SHIFT_NEXT_LEVEL(eax);
sub_index = 1;
do {
cpuid_count(0xb, sub_index, &eax, &ebx, &ecx, &edx);
/*
* Check for the Core type in the implemented sub leaves.
*/
if (LEAFB_SUBTYPE(ecx) == CORE_TYPE) {
core_level_siblings = LEVEL_MAX_SIBLINGS(ebx);
core_plus_mask_width = BITS_SHIFT_NEXT_LEVEL(eax);
break;
}
sub_index++;
} while (LEAFB_SUBTYPE(ecx) != INVALID_TYPE);
core_select_mask = (~(-1 << core_plus_mask_width)) >> ht_mask_width;
#ifdef CONFIG_X86_32
c->cpu_core_id = phys_pkg_id(c->initial_apicid, ht_mask_width)
& core_select_mask;
c->phys_proc_id = phys_pkg_id(c->initial_apicid, core_plus_mask_width);
#else
c->cpu_core_id = phys_pkg_id(ht_mask_width) & core_select_mask;
c->phys_proc_id = phys_pkg_id(core_plus_mask_width);
#endif
c->x86_max_cores = (core_level_siblings / smp_num_siblings);
printk(KERN_INFO "CPU: Physical Processor ID: %d\n",
c->phys_proc_id);
if (c->x86_max_cores > 1)
printk(KERN_INFO "CPU: Processor Core ID: %d\n",
c->cpu_core_id);
return;
#endif
}
static void intel_workarounds(struct cpuinfo_x86 *c)
{
#ifdef CONFIG_X86_F00F_BUG
/*
* All models of Pentium and Pentium with MMX technology CPUs
* have the F0 0F bug, which lets nonprivileged users lock up the
* system. Announce that the fault handler will be checking for it.
* The Quark is also family 5, but does not have the same bug.
*/
clear_cpu_bug(c, X86_BUG_F00F);
if (!paravirt_enabled() && c->x86 == 5 && c->x86_model < 9) {
static int f00f_workaround_enabled;
set_cpu_bug(c, X86_BUG_F00F);
if (!f00f_workaround_enabled) {
printk(KERN_NOTICE "Intel Pentium with F0 0F bug - workaround enabled.\n");
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);
/*
* 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) {
printk(KERN_WARNING "PAE forced!\n");
set_cpu_cap(c, X86_FEATURE_PAE);
add_taint(TAINT_CPU_OUT_OF_SPEC, LOCKDEP_NOW_UNRELIABLE);
}
/*
* P4 Xeon errata 037 workaround.
* Hardware prefetcher may cause stale data to be loaded into the cache.
*/
if ((c->x86 == 15) && (c->x86_model == 1) && (c->x86_mask == 1)) {
if (msr_set_bit(MSR_IA32_MISC_ENABLE,
MSR_IA32_MISC_ENABLE_PREFETCH_DISABLE_BIT)
> 0) {
pr_info("CPU: C0 stepping P4 Xeon detected.\n");
pr_info("CPU: Disabling hardware prefetching (Errata 037)\n");
}
}
/*
* 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))
set_cpu_bug(c, X86_BUG_11AP);
#ifdef CONFIG_X86_INTEL_USERCOPY
/*
* Set up the preferred alignment for movsl bulk memory moves
*/
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;
}
#endif
intel_smp_check(c);
}
static int __init detect_init_APIC(void)
{
u32 h, l, features;
if (disable_apic)
return -1;
switch (boot_cpu_data.x86_vendor) {
case X86_VENDOR_AMD:
if ((boot_cpu_data.x86 == 6 && boot_cpu_data.x86_model > 1) ||
(boot_cpu_data.x86 >= 15))
break;
goto no_apic;
case X86_VENDOR_INTEL:
if (boot_cpu_data.x86 == 6 || boot_cpu_data.x86 == 15 ||
(boot_cpu_data.x86 == 5 && cpu_has_apic))
break;
goto no_apic;
default:
goto no_apic;
}
if (!cpu_has_apic) {
if (!force_enable_local_apic) {
pr_info("Local APIC disabled by BIOS -- "
"you can enable it with \"lapic\"\n");
return -1;
}
rdmsr(MSR_IA32_APICBASE, l, h);
if (!(l & MSR_IA32_APICBASE_ENABLE)) {
pr_info("Local APIC disabled by BIOS -- reenabling.\n");
l &= ~MSR_IA32_APICBASE_BASE;
l |= MSR_IA32_APICBASE_ENABLE | APIC_DEFAULT_PHYS_BASE;
wrmsr(MSR_IA32_APICBASE, l, h);
enabled_via_apicbase = 1;
}
}
features = cpuid_edx(1);
if (!(features & (1 << X86_FEATURE_APIC))) {
pr_warning("Could not enable APIC!\n");
return -1;
}
set_cpu_cap(&boot_cpu_data, X86_FEATURE_APIC);
mp_lapic_addr = APIC_DEFAULT_PHYS_BASE;
rdmsr(MSR_IA32_APICBASE, l, h);
if (l & MSR_IA32_APICBASE_ENABLE)
mp_lapic_addr = l & MSR_IA32_APICBASE_BASE;
pr_info("Found and enabled local APIC!\n");
apic_pm_activate();
return 0;
no_apic:
pr_info("No local APIC present or hardware disabled\n");
return -1;
}