aboutsummaryrefslogtreecommitdiff
path: root/arch/x86/kernel/cpu
diff options
context:
space:
mode:
Diffstat (limited to 'arch/x86/kernel/cpu')
-rw-r--r--arch/x86/kernel/cpu/Makefile20
-rw-r--r--arch/x86/kernel/cpu/addon_cpuid_features.c50
-rw-r--r--arch/x86/kernel/cpu/amd.c337
-rw-r--r--arch/x86/kernel/cpu/bugs.c192
-rw-r--r--arch/x86/kernel/cpu/centaur.c471
-rw-r--r--arch/x86/kernel/cpu/common.c733
-rw-r--r--arch/x86/kernel/cpu/cpu.h28
-rw-r--r--arch/x86/kernel/cpu/cyrix.c463
-rw-r--r--arch/x86/kernel/cpu/intel.c333
-rw-r--r--arch/x86/kernel/cpu/intel_cacheinfo.c806
-rw-r--r--arch/x86/kernel/cpu/nexgen.c60
-rw-r--r--arch/x86/kernel/cpu/perfctr-watchdog.c713
-rw-r--r--arch/x86/kernel/cpu/proc.c192
-rw-r--r--arch/x86/kernel/cpu/transmeta.c116
-rw-r--r--arch/x86/kernel/cpu/umc.c26
15 files changed, 4540 insertions, 0 deletions
diff --git a/arch/x86/kernel/cpu/Makefile b/arch/x86/kernel/cpu/Makefile
new file mode 100644
index 00000000000..6687f6d5ad2
--- /dev/null
+++ b/arch/x86/kernel/cpu/Makefile
@@ -0,0 +1,20 @@
+#
+# Makefile for x86-compatible CPU details and quirks
+#
+
+obj-y := common.o proc.o bugs.o
+
+obj-y += amd.o
+obj-y += cyrix.o
+obj-y += centaur.o
+obj-y += transmeta.o
+obj-y += intel.o intel_cacheinfo.o addon_cpuid_features.o
+obj-y += nexgen.o
+obj-y += umc.o
+
+obj-$(CONFIG_X86_MCE) += ../../../x86/kernel/cpu/mcheck/
+
+obj-$(CONFIG_MTRR) += ../../../x86/kernel/cpu/mtrr/
+obj-$(CONFIG_CPU_FREQ) += ../../../x86/kernel/cpu/cpufreq/
+
+obj-$(CONFIG_X86_LOCAL_APIC) += perfctr-watchdog.o
diff --git a/arch/x86/kernel/cpu/addon_cpuid_features.c b/arch/x86/kernel/cpu/addon_cpuid_features.c
new file mode 100644
index 00000000000..3e91d3ee26e
--- /dev/null
+++ b/arch/x86/kernel/cpu/addon_cpuid_features.c
@@ -0,0 +1,50 @@
+
+/*
+ * Routines to indentify additional cpu features that are scattered in
+ * cpuid space.
+ */
+
+#include <linux/cpu.h>
+
+#include <asm/processor.h>
+
+struct cpuid_bit {
+ u16 feature;
+ u8 reg;
+ u8 bit;
+ u32 level;
+};
+
+enum cpuid_regs {
+ CR_EAX = 0,
+ CR_ECX,
+ CR_EDX,
+ CR_EBX
+};
+
+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, &regs[CR_EAX], &regs[CR_EBX],
+ &regs[CR_ECX], &regs[CR_EDX]);
+
+ if (regs[cb->reg] & (1 << cb->bit))
+ set_bit(cb->feature, c->x86_capability);
+ }
+}
diff --git a/arch/x86/kernel/cpu/amd.c b/arch/x86/kernel/cpu/amd.c
new file mode 100644
index 00000000000..dcf6bbb1c7c
--- /dev/null
+++ b/arch/x86/kernel/cpu/amd.c
@@ -0,0 +1,337 @@
+#include <linux/init.h>
+#include <linux/bitops.h>
+#include <linux/mm.h>
+#include <asm/io.h>
+#include <asm/processor.h>
+#include <asm/apic.h>
+
+#include "cpu.h"
+
+/*
+ * B step AMD K6 before B 9730xxxx have hardware bugs that can cause
+ * misexecution of code under Linux. Owners of such processors should
+ * contact AMD for precise details and a CPU swap.
+ *
+ * See http://www.multimania.com/poulot/k6bug.html
+ * http://www.amd.com/K6/k6docs/revgd.html
+ *
+ * The following test is erm.. interesting. AMD neglected to up
+ * the chip setting when fixing the bug but they also tweaked some
+ * performance at the same time..
+ */
+
+extern void vide(void);
+__asm__(".align 4\nvide: ret");
+
+#ifdef CONFIG_X86_LOCAL_APIC
+#define ENABLE_C1E_MASK 0x18000000
+#define CPUID_PROCESSOR_SIGNATURE 1
+#define CPUID_XFAM 0x0ff00000
+#define CPUID_XFAM_K8 0x00000000
+#define CPUID_XFAM_10H 0x00100000
+#define CPUID_XFAM_11H 0x00200000
+#define CPUID_XMOD 0x000f0000
+#define CPUID_XMOD_REV_F 0x00040000
+
+/* AMD systems with C1E don't have a working lAPIC timer. Check for that. */
+static __cpuinit int amd_apic_timer_broken(void)
+{
+ u32 lo, hi;
+ u32 eax = cpuid_eax(CPUID_PROCESSOR_SIGNATURE);
+ switch (eax & CPUID_XFAM) {
+ case CPUID_XFAM_K8:
+ if ((eax & CPUID_XMOD) < CPUID_XMOD_REV_F)
+ break;
+ case CPUID_XFAM_10H:
+ case CPUID_XFAM_11H:
+ rdmsr(MSR_K8_ENABLE_C1E, lo, hi);
+ if (lo & ENABLE_C1E_MASK)
+ return 1;
+ break;
+ default:
+ /* err on the side of caution */
+ return 1;
+ }
+ return 0;
+}
+#endif
+
+int force_mwait __cpuinitdata;
+
+static void __cpuinit init_amd(struct cpuinfo_x86 *c)
+{
+ u32 l, h;
+ int mbytes = num_physpages >> (20-PAGE_SHIFT);
+ int r;
+
+#ifdef CONFIG_SMP
+ unsigned long long value;
+
+ /* Disable TLB flush filter by setting HWCR.FFDIS on K8
+ * bit 6 of msr C001_0015
+ *
+ * Errata 63 for SH-B3 steppings
+ * Errata 122 for all steppings (F+ have it disabled by default)
+ */
+ if (c->x86 == 15) {
+ rdmsrl(MSR_K7_HWCR, value);
+ value |= 1 << 6;
+ wrmsrl(MSR_K7_HWCR, value);
+ }
+#endif
+
+ /*
+ * FIXME: We should handle the K5 here. Set up the write
+ * range and also turn on MSR 83 bits 4 and 31 (write alloc,
+ * no bus pipeline)
+ */
+
+ /* Bit 31 in normal CPUID used for nonstandard 3DNow ID;
+ 3DNow is IDd by bit 31 in extended CPUID (1*32+31) anyway */
+ clear_bit(0*32+31, c->x86_capability);
+
+ r = get_model_name(c);
+
+ switch(c->x86)
+ {
+ case 4:
+ /*
+ * General Systems BIOSen alias the cpu frequency registers
+ * of the Elan at 0x000df000. Unfortuantly, one of the Linux
+ * drivers subsequently pokes it, and changes the CPU speed.
+ * Workaround : Remove the unneeded alias.
+ */
+#define CBAR (0xfffc) /* Configuration Base Address (32-bit) */
+#define CBAR_ENB (0x80000000)
+#define CBAR_KEY (0X000000CB)
+ if (c->x86_model==9 || c->x86_model == 10) {
+ if (inl (CBAR) & CBAR_ENB)
+ outl (0 | CBAR_KEY, CBAR);
+ }
+ break;
+ case 5:
+ if( c->x86_model < 6 )
+ {
+ /* Based on AMD doc 20734R - June 2000 */
+ if ( c->x86_model == 0 ) {
+ clear_bit(X86_FEATURE_APIC, c->x86_capability);
+ set_bit(X86_FEATURE_PGE, c->x86_capability);
+ }
+ break;
+ }
+
+ if ( c->x86_model == 6 && c->x86_mask == 1 ) {
+ const int K6_BUG_LOOP = 1000000;
+ int n;
+ void (*f_vide)(void);
+ unsigned long d, d2;
+
+ printk(KERN_INFO "AMD K6 stepping B detected - ");
+
+ /*
+ * It looks like AMD fixed the 2.6.2 bug and improved indirect
+ * calls at the same time.
+ */
+
+ n = K6_BUG_LOOP;
+ f_vide = vide;
+ rdtscl(d);
+ while (n--)
+ f_vide();
+ rdtscl(d2);
+ d = d2-d;
+
+ if (d > 20*K6_BUG_LOOP)
+ printk("system stability may be impaired when more than 32 MB are used.\n");
+ else
+ printk("probably OK (after B9730xxxx).\n");
+ printk(KERN_INFO "Please see http://membres.lycos.fr/poulot/k6bug.html\n");
+ }
+
+ /* K6 with old style WHCR */
+ if (c->x86_model < 8 ||
+ (c->x86_model== 8 && c->x86_mask < 8)) {
+ /* We can only write allocate on the low 508Mb */
+ if(mbytes>508)
+ mbytes=508;
+
+ rdmsr(MSR_K6_WHCR, l, h);
+ if ((l&0x0000FFFF)==0) {
+ unsigned long flags;
+ l=(1<<0)|((mbytes/4)<<1);
+ local_irq_save(flags);
+ wbinvd();
+ wrmsr(MSR_K6_WHCR, l, h);
+ local_irq_restore(flags);
+ printk(KERN_INFO "Enabling old style K6 write allocation for %d Mb\n",
+ mbytes);
+ }
+ break;
+ }
+
+ if ((c->x86_model == 8 && c->x86_mask >7) ||
+ c->x86_model == 9 || c->x86_model == 13) {
+ /* The more serious chips .. */
+
+ if(mbytes>4092)
+ mbytes=4092;
+
+ rdmsr(MSR_K6_WHCR, l, h);
+ if ((l&0xFFFF0000)==0) {
+ unsigned long flags;
+ l=((mbytes>>2)<<22)|(1<<16);
+ local_irq_save(flags);
+ wbinvd();
+ wrmsr(MSR_K6_WHCR, l, h);
+ local_irq_restore(flags);
+ printk(KERN_INFO "Enabling new style K6 write allocation for %d Mb\n",
+ mbytes);
+ }
+
+ /* Set MTRR capability flag if appropriate */
+ if (c->x86_model == 13 || c->x86_model == 9 ||
+ (c->x86_model == 8 && c->x86_mask >= 8))
+ set_bit(X86_FEATURE_K6_MTRR, c->x86_capability);
+ break;
+ }
+
+ if (c->x86_model == 10) {
+ /* AMD Geode LX is model 10 */
+ /* placeholder for any needed mods */
+ break;
+ }
+ break;
+ case 6: /* An Athlon/Duron */
+
+ /* Bit 15 of Athlon specific MSR 15, needs to be 0
+ * to enable SSE on Palomino/Morgan/Barton CPU's.
+ * If the BIOS didn't enable it already, enable it here.
+ */
+ if (c->x86_model >= 6 && c->x86_model <= 10) {
+ if (!cpu_has(c, X86_FEATURE_XMM)) {
+ printk(KERN_INFO "Enabling disabled K7/SSE Support.\n");
+ rdmsr(MSR_K7_HWCR, l, h);
+ l &= ~0x00008000;
+ wrmsr(MSR_K7_HWCR, l, h);
+ set_bit(X86_FEATURE_XMM, c->x86_capability);
+ }
+ }
+
+ /* It's been determined by AMD that Athlons since model 8 stepping 1
+ * are more robust with CLK_CTL set to 200xxxxx instead of 600xxxxx
+ * As per AMD technical note 27212 0.2
+ */
+ if ((c->x86_model == 8 && c->x86_mask>=1) || (c->x86_model > 8)) {
+ rdmsr(MSR_K7_CLK_CTL, l, h);
+ if ((l & 0xfff00000) != 0x20000000) {
+ printk ("CPU: CLK_CTL MSR was %x. Reprogramming to %x\n", l,
+ ((l & 0x000fffff)|0x20000000));
+ wrmsr(MSR_K7_CLK_CTL, (l & 0x000fffff)|0x20000000, h);
+ }
+ }
+ break;
+ }
+
+ switch (c->x86) {
+ case 15:
+ /* Use K8 tuning for Fam10h and Fam11h */
+ case 0x10:
+ case 0x11:
+ set_bit(X86_FEATURE_K8, c->x86_capability);
+ break;
+ case 6:
+ set_bit(X86_FEATURE_K7, c->x86_capability);
+ break;
+ }
+ if (c->x86 >= 6)
+ set_bit(X86_FEATURE_FXSAVE_LEAK, c->x86_capability);
+
+ display_cacheinfo(c);
+
+ if (cpuid_eax(0x80000000) >= 0x80000008) {
+ c->x86_max_cores = (cpuid_ecx(0x80000008) & 0xff) + 1;
+ }
+
+ if (cpuid_eax(0x80000000) >= 0x80000007) {
+ c->x86_power = cpuid_edx(0x80000007);
+ if (c->x86_power & (1<<8))
+ set_bit(X86_FEATURE_CONSTANT_TSC, c->x86_capability);
+ }
+
+#ifdef CONFIG_X86_HT
+ /*
+ * On a AMD multi core setup the lower bits of the APIC id
+ * distingush the cores.
+ */
+ if (c->x86_max_cores > 1) {
+ int cpu = smp_processor_id();
+ unsigned bits = (cpuid_ecx(0x80000008) >> 12) & 0xf;
+
+ if (bits == 0) {
+ while ((1 << bits) < c->x86_max_cores)
+ bits++;
+ }
+ c->cpu_core_id = c->phys_proc_id & ((1<<bits)-1);
+ c->phys_proc_id >>= bits;
+ printk(KERN_INFO "CPU %d(%d) -> Core %d\n",
+ cpu, c->x86_max_cores, c->cpu_core_id);
+ }
+#endif
+
+ if (cpuid_eax(0x80000000) >= 0x80000006) {
+ if ((c->x86 == 0x10) && (cpuid_edx(0x80000006) & 0xf000))
+ num_cache_leaves = 4;
+ else
+ num_cache_leaves = 3;
+ }
+
+#ifdef CONFIG_X86_LOCAL_APIC
+ if (amd_apic_timer_broken())
+ local_apic_timer_disabled = 1;
+#endif
+
+ if (c->x86 == 0x10 && !force_mwait)
+ clear_bit(X86_FEATURE_MWAIT, c->x86_capability);
+
+ /* K6s reports MCEs but don't actually have all the MSRs */
+ if (c->x86 < 6)
+ clear_bit(X86_FEATURE_MCE, c->x86_capability);
+}
+
+static unsigned int __cpuinit amd_size_cache(struct cpuinfo_x86 * c, unsigned int size)
+{
+ /* AMD errata T13 (order #21922) */
+ if ((c->x86 == 6)) {
+ if (c->x86_model == 3 && c->x86_mask == 0) /* Duron Rev A0 */
+ size = 64;
+ if (c->x86_model == 4 &&
+ (c->x86_mask==0 || c->x86_mask==1)) /* Tbird rev A1/A2 */
+ size = 256;
+ }
+ return size;
+}
+
+static struct cpu_dev amd_cpu_dev __cpuinitdata = {
+ .c_vendor = "AMD",
+ .c_ident = { "AuthenticAMD" },
+ .c_models = {
+ { .vendor = X86_VENDOR_AMD, .family = 4, .model_names =
+ {
+ [3] = "486 DX/2",
+ [7] = "486 DX/2-WB",
+ [8] = "486 DX/4",
+ [9] = "486 DX/4-WB",
+ [14] = "Am5x86-WT",
+ [15] = "Am5x86-WB"
+ }
+ },
+ },
+ .c_init = init_amd,
+ .c_size_cache = amd_size_cache,
+};
+
+int __init amd_init_cpu(void)
+{
+ cpu_devs[X86_VENDOR_AMD] = &amd_cpu_dev;
+ return 0;
+}
diff --git a/arch/x86/kernel/cpu/bugs.c b/arch/x86/kernel/cpu/bugs.c
new file mode 100644
index 00000000000..59266f03d1c
--- /dev/null
+++ b/arch/x86/kernel/cpu/bugs.c
@@ -0,0 +1,192 @@
+/*
+ * arch/i386/cpu/bugs.c
+ *
+ * Copyright (C) 1994 Linus Torvalds
+ *
+ * Cyrix stuff, June 1998 by:
+ * - Rafael R. Reilova (moved everything from head.S),
+ * <rreilova@ececs.uc.edu>
+ * - Channing Corn (tests & fixes),
+ * - Andrew D. Balsa (code cleanup).
+ */
+#include <linux/init.h>
+#include <linux/utsname.h>
+#include <asm/bugs.h>
+#include <asm/processor.h>
+#include <asm/i387.h>
+#include <asm/msr.h>
+#include <asm/paravirt.h>
+#include <asm/alternative.h>
+
+static int __init no_halt(char *s)
+{
+ boot_cpu_data.hlt_works_ok = 0;
+ return 1;
+}
+
+__setup("no-hlt", no_halt);
+
+static int __init mca_pentium(char *s)
+{
+ mca_pentium_flag = 1;
+ return 1;
+}
+
+__setup("mca-pentium", mca_pentium);
+
+static int __init no_387(char *s)
+{
+ boot_cpu_data.hard_math = 0;
+ write_cr0(0xE | read_cr0());
+ return 1;
+}
+
+__setup("no387", no_387);
+
+static double __initdata x = 4195835.0;
+static double __initdata y = 3145727.0;
+
+/*
+ * This used to check for exceptions..
+ * However, it turns out that to support that,
+ * the XMM trap handlers basically had to
+ * be buggy. So let's have a correct XMM trap
+ * handler, and forget about printing out
+ * some status at boot.
+ *
+ * We should really only care about bugs here
+ * anyway. Not features.
+ */
+static void __init check_fpu(void)
+{
+ if (!boot_cpu_data.hard_math) {
+#ifndef CONFIG_MATH_EMULATION
+ printk(KERN_EMERG "No coprocessor found and no math emulation present.\n");
+ printk(KERN_EMERG "Giving up.\n");
+ for (;;) ;
+#endif
+ return;
+ }
+
+/* trap_init() enabled FXSR and company _before_ testing for FP problems here. */
+ /* Test for the divl bug.. */
+ __asm__("fninit\n\t"
+ "fldl %1\n\t"
+ "fdivl %2\n\t"
+ "fmull %2\n\t"
+ "fldl %1\n\t"
+ "fsubp %%st,%%st(1)\n\t"
+ "fistpl %0\n\t"
+ "fwait\n\t"
+ "fninit"
+ : "=m" (*&boot_cpu_data.fdiv_bug)
+ : "m" (*&x), "m" (*&y));
+ if (boot_cpu_data.fdiv_bug)
+ printk("Hmm, FPU with FDIV bug.\n");
+}
+
+static void __init check_hlt(void)
+{
+ if (paravirt_enabled())
+ return;
+
+ printk(KERN_INFO "Checking 'hlt' instruction... ");
+ if (!boot_cpu_data.hlt_works_ok) {
+ printk("disabled\n");
+ return;
+ }
+ halt();
+ halt();
+ halt();
+ halt();
+ printk("OK.\n");
+}
+
+/*
+ * Most 386 processors have a bug where a POPAD can lock the
+ * machine even from user space.
+ */
+
+static void __init check_popad(void)
+{
+#ifndef CONFIG_X86_POPAD_OK
+ int res, inp = (int) &res;
+
+ printk(KERN_INFO "Checking for popad bug... ");
+ __asm__ __volatile__(
+ "movl $12345678,%%eax; movl $0,%%edi; pusha; popa; movl (%%edx,%%edi),%%ecx "
+ : "=&a" (res)
+ : "d" (inp)
+ : "ecx", "edi" );
+ /* If this fails, it means that any user program may lock the CPU hard. Too bad. */
+ if (res != 12345678) printk( "Buggy.\n" );
+ else printk( "OK.\n" );
+#endif
+}
+
+/*
+ * Check whether we are able to run this kernel safely on SMP.
+ *
+ * - In order to run on a i386, we need to be compiled for i386
+ * (for due to lack of "invlpg" and working WP on a i386)
+ * - In order to run on anything without a TSC, we need to be
+ * compiled for a i486.
+ * - In order to support the local APIC on a buggy Pentium machine,
+ * we need to be compiled with CONFIG_X86_GOOD_APIC disabled,
+ * which happens implicitly if compiled for a Pentium or lower
+ * (unless an advanced selection of CPU features is used) as an
+ * otherwise config implies a properly working local APIC without
+ * the need to do extra reads from the APIC.
+*/
+
+static void __init check_config(void)
+{
+/*
+ * We'd better not be a i386 if we're configured to use some
+ * i486+ only features! (WP works in supervisor mode and the
+ * new "invlpg" and "bswap" instructions)
+ */
+#if defined(CONFIG_X86_WP_WORKS_OK) || defined(CONFIG_X86_INVLPG) || defined(CONFIG_X86_BSWAP)
+ if (boot_cpu_data.x86 == 3)
+ panic("Kernel requires i486+ for 'invlpg' and other features");
+#endif
+
+/*
+ * If we configured ourselves for a TSC, we'd better have one!
+ */
+#ifdef CONFIG_X86_TSC
+ if (!cpu_has_tsc && !tsc_disable)
+ panic("Kernel compiled for Pentium+, requires TSC feature!");
+#endif
+
+/*
+ * If we were told we had a good local APIC, check 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 defined(CONFIG_X86_LOCAL_APIC) && defined(CONFIG_X86_GOOD_APIC)
+ if (boot_cpu_data.x86_vendor == X86_VENDOR_INTEL
+ && cpu_has_apic
+ && boot_cpu_data.x86 == 5
+ && boot_cpu_data.x86_model == 2
+ && (boot_cpu_data.x86_mask < 6 || boot_cpu_data.x86_mask == 11))
+ panic("Kernel compiled for PMMX+, assumes a local APIC without the read-before-write bug!");
+#endif
+}
+
+
+void __init check_bugs(void)
+{
+ identify_boot_cpu();
+#ifndef CONFIG_SMP
+ printk("CPU: ");
+ print_cpu_info(&boot_cpu_data);
+#endif
+ check_config();
+ check_fpu();
+ check_hlt();
+ check_popad();
+ init_utsname()->machine[1] = '0' + (boot_cpu_data.x86 > 6 ? 6 : boot_cpu_data.x86);
+ alternative_instructions();
+}
diff --git a/arch/x86/kernel/cpu/centaur.c b/arch/x86/kernel/cpu/centaur.c
new file mode 100644
index 00000000000..473eac883c7
--- /dev/null
+++ b/arch/x86/kernel/cpu/centaur.c
@@ -0,0 +1,471 @@
+#include <linux/kernel.h>
+#include <linux/init.h>
+#include <linux/bitops.h>
+#include <asm/processor.h>
+#include <asm/msr.h>
+#include <asm/e820.h>
+#include <asm/mtrr.h>
+#include "cpu.h"
+
+#ifdef CONFIG_X86_OOSTORE
+
+static u32 __cpuinit power2(u32 x)
+{
+ u32 s=1;
+ while(s<=x)
+ s<<=1;
+ return s>>=1;
+}
+
+
+/*
+ * Set up an actual MCR
+ */
+
+static void __cpuinit centaur_mcr_insert(int reg, u32 base, u32 size, int key)
+{
+ u32 lo, hi;
+
+ hi = base & ~0xFFF;
+ lo = ~(size-1); /* Size is a power of 2 so this makes a mask */
+ lo &= ~0xFFF; /* Remove the ctrl value bits */
+ lo |= key; /* Attribute we wish to set */
+ wrmsr(reg+MSR_IDT_MCR0, lo, hi);
+ mtrr_centaur_report_mcr(reg, lo, hi); /* Tell the mtrr driver */
+}
+
+/*
+ * Figure what we can cover with MCR's
+ *
+ * Shortcut: We know you can't put 4Gig of RAM on a winchip
+ */
+
+static u32 __cpuinit ramtop(void) /* 16388 */
+{
+ int i;
+ u32 top = 0;
+ u32 clip = 0xFFFFFFFFUL;
+
+ for (i = 0; i < e820.nr_map; i++) {
+ unsigned long start, end;
+
+ if (e820.map[i].addr > 0xFFFFFFFFUL)
+ continue;
+ /*
+ * Don't MCR over reserved space. Ignore the ISA hole
+ * we frob around that catastrophy already
+ */
+
+ if (e820.map[i].type == E820_RESERVED)
+ {
+ if(e820.map[i].addr >= 0x100000UL && e820.map[i].addr < clip)
+ clip = e820.map[i].addr;
+ continue;
+ }
+ start = e820.map[i].addr;
+ end = e820.map[i].addr + e820.map[i].size;
+ if (start >= end)
+ continue;
+ if (end > top)
+ top = end;
+ }
+ /* Everything below 'top' should be RAM except for the ISA hole.
+ Because of the limited MCR's we want to map NV/ACPI into our
+ MCR range for gunk in RAM
+
+ Clip might cause us to MCR insufficient RAM but that is an
+ acceptable failure mode and should only bite obscure boxes with
+ a VESA hole at 15Mb
+
+ The second case Clip sometimes kicks in is when the EBDA is marked
+ as reserved. Again we fail safe with reasonable results
+ */
+
+ if(top>clip)
+ top=clip;
+
+ return top;
+}
+
+/*
+ * Compute a set of MCR's to give maximum coverage
+ */
+
+static int __cpuinit centaur_mcr_compute(int nr, int key)
+{
+ u32 mem = ramtop();
+ u32 root = power2(mem);
+ u32 base = root;
+ u32 top = root;
+ u32 floor = 0;
+ int ct = 0;
+
+ while(ct<nr)
+ {
+ u32 fspace = 0;
+
+ /*
+ * Find the largest block we will fill going upwards
+ */
+
+ u32 high = power2(mem-top);
+
+ /*
+ * Find the largest block we will fill going downwards
+ */
+
+ u32 low = base/2;
+
+ /*
+ * Don't fill below 1Mb going downwards as there
+ * is an ISA hole in the way.
+ */
+
+ if(base <= 1024*1024)
+ low = 0;
+
+ /*
+ * See how much space we could cover by filling below
+ * the ISA hole
+ */
+
+ if(floor == 0)
+ fspace = 512*1024;
+ else if(floor ==512*1024)
+ fspace = 128*1024;
+
+ /* And forget ROM space */
+
+ /*
+ * Now install the largest coverage we get
+ */
+
+ if(fspace > high && fspace > low)
+ {
+ centaur_mcr_insert(ct, floor, fspace, key);
+ floor += fspace;
+ }
+ else if(high > low)
+ {
+ centaur_mcr_insert(ct, top, high, key);
+ top += high;
+ }
+ else if(low > 0)
+ {
+ base -= low;
+ centaur_mcr_insert(ct, base, low, key);
+ }
+ else break;
+ ct++;
+ }
+ /*
+ * We loaded ct values. We now need to set the mask. The caller
+ * must do this bit.
+ */
+
+ return ct;
+}
+
+static void __cpuinit centaur_create_optimal_mcr(void)
+{
+ int i;
+ /*
+ * Allocate up to 6 mcrs to mark as much of ram as possible
+ * as write combining and weak write ordered.
+ *
+ * To experiment with: Linux never uses stack operations for
+ * mmio spaces so we could globally enable stack operation wc
+ *
+ * Load the registers with type 31 - full write combining, all
+ * writes weakly ordered.
+ */
+ int used = centaur_mcr_compute(6, 31);
+
+ /*
+ * Wipe unused MCRs
+ */
+
+ for(i=used;i<8;i++)
+ wrmsr(MSR_IDT_MCR0+i, 0, 0);
+}
+
+static void __cpuinit winchip2_create_optimal_mcr(void)
+{
+ u32 lo, hi;
+ int i;
+
+ /*
+ * Allocate up to 6 mcrs to mark as much of ram as possible
+ * as write combining, weak store ordered.
+ *
+ * Load the registers with type 25
+ * 8 - weak write ordering
+ * 16 - weak read ordering
+ * 1 - write combining
+ */
+
+ int used = centaur_mcr_compute(6, 25);
+
+ /*
+ * Mark the registers we are using.
+ */
+
+ rdmsr(MSR_IDT_MCR_CTRL, lo, hi);
+ for(i=0;i<used;i++)
+ lo|=1<<(9+i);
+ wrmsr(MSR_IDT_MCR_CTRL, lo, hi);
+
+ /*
+ * Wipe unused MCRs
+ */
+
+ for(i=used;i<8;i++)
+ wrmsr(MSR_IDT_MCR0+i, 0, 0);
+}
+
+/*
+ * Handle the MCR key on the Winchip 2.
+ */
+
+static void __cpuinit winchip2_unprotect_mcr(void)
+{
+ u32 lo, hi;
+ u32 key;
+
+ rdmsr(MSR_IDT_MCR_CTRL, lo, hi);
+ lo&=~0x1C0; /* blank bits 8-6 */
+ key = (lo>>17) & 7;
+ lo |= key<<6; /* replace with unlock key */
+ wrmsr(MSR_IDT_MCR_CTRL, lo, hi);
+}
+
+static void __cpuinit winchip2_protect_mcr(void)
+{
+ u32 lo, hi;
+
+ rdmsr(MSR_IDT_MCR_CTRL, lo, hi);
+ lo&=~0x1C0; /* blank bits 8-6 */
+ wrmsr(MSR_IDT_MCR_CTRL, lo, hi);
+}
+#endif /* CONFIG_X86_OOSTORE */
+
+#define ACE_PRESENT (1 << 6)
+#define ACE_ENABLED (1 << 7)
+#define ACE_FCR (1 << 28) /* MSR_VIA_FCR */
+
+#define RNG_PRESENT (1 << 2)
+#define RNG_ENABLED (1 << 3)
+#define RNG_ENABLE (1 << 6) /* MSR_VIA_RNG */
+
+static void __cpuinit 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);
+ printk(KERN_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);
+ printk(KERN_INFO "CPU: Enabled h/w RNG\n");
+ }
+
+ /* store Centaur Extended Feature Flags as
+ * word 5 of the CPU capability bit array
+ */
+ c->x86_capability[5] = cpuid_edx(0xC0000001);
+ }
+
+ /* Cyrix III family needs CX8 & PGE explicity enabled. */
+ if (c->x86_model >=6 && c->x86_model <= 9) {
+ rdmsr (MSR_VIA_FCR, lo, hi);
+ lo |= (1<<1 | 1<<7);
+ wrmsr (MSR_VIA_FCR, lo, hi);
+ set_bit(X86_FEATURE_CX8, c->x86_capability);
+ }
+
+ /* Before Nehemiah, the C3's had 3dNOW! */
+ if (c->x86_model >=6 && c->x86_model <9)
+ set_bit(X86_FEATURE_3DNOW, c->x86_capability);
+
+ get_model_name(c);
+ display_cacheinfo(c);
+}
+
+static void __cpuinit init_centaur(struct cpuinfo_x86 *c)
+{
+ enum {
+ ECX8=1<<1,
+ EIERRINT=1<<2,
+ DPM=1<<3,
+ DMCE=1<<4,
+ DSTPCLK=1<<5,
+ ELINEAR=1<<6,
+ DSMC=1<<7,
+ DTLOCK=1<<8,
+ EDCTLB=1<<8,
+ EMMX=1<<9,
+ DPDC=1<<11,
+ EBRPRED=1<<12,
+ DIC=1<<13,
+ DDC=1<<14,
+ DNA=1<<15,
+ ERETSTK=1<<16,
+ E2MMX=1<<19,
+ EAMD3D=1<<20,
+ };
+
+ 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_bit(0*32+31, c->x86_capability);
+
+ switch (c->x86) {
+
+ case 5:
+ switch(c->x86_model) {
+ case 4:
+ name="C6";
+ fcr_set=ECX8|DSMC|EDCTLB|EMMX|ERETSTK;
+ fcr_clr=DPDC;
+ printk(KERN_NOTICE "Disabling bugged TSC.\n");
+ clear_bit(X86_FEATURE_TSC, c->x86_capability);
+#ifdef CONFIG_X86_OOSTORE
+ centaur_create_optimal_mcr();
+ /* Enable
+ write combining on non-stack, non-string
+ write combining on string, all types
+ weak write ordering
+
+ The C6 original lacks weak read order
+
+ Note 0x120 is write only on Winchip 1 */
+
+ wrmsr(MSR_IDT_MCR_CTRL, 0x01F0001F, 0);
+#endif
+ 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;
+#ifdef CONFIG_X86_OOSTORE
+ winchip2_unprotect_mcr();
+ winchip2_create_optimal_mcr();
+ rdmsr(MSR_IDT_MCR_CTRL, lo, hi);
+ /* Enable
+ write combining on non-stack, non-string
+ write combining on string, all types
+ weak write ordering
+ */
+ lo|=31;
+ wrmsr(MSR_IDT_MCR_CTRL, lo, hi);
+ winchip2_protect_mcr();
+#endif
+ break;
+ case 9:
+ name="3";
+ fcr_set=ECX8|DSMC|DTLOCK|EMMX|EBRPRED|ERETSTK|E2MMX|EAMD3D;
+ fcr_clr=DPDC;
+#ifdef CONFIG_X86_OOSTORE
+ winchip2_unprotect_mcr();
+ winchip2_create_optimal_mcr();
+ rdmsr(MSR_IDT_MCR_CTRL, lo, hi);
+ /* Enable
+ write combining on non-stack, non-string
+ write combining on string, all types
+ weak write ordering
+ */
+ lo|=31;
+ wrmsr(MSR_IDT_MCR_CTRL, lo, hi);
+ winchip2_protect_mcr();
+#endif
+ break;
+ default:
+ name="??";
+ }
+
+ rdmsr(MSR_IDT_FCR1, lo, hi);
+ newlo=(lo|fcr_set) & (~fcr_clr);
+
+ if (newlo!=lo) {
+ printk(KERN_INFO "Centaur FCR was 0x%X now 0x%X\n", lo, newlo );
+ wrmsr(MSR_IDT_FCR1, newlo, hi );
+ } else {
+ printk(KERN_INFO "Centaur FCR is 0x%X\n",lo);
+ }
+ /* Emulate MTRRs using Centaur's MCR. */
+ set_bit(X86_FEATURE_CENTAUR_MCR, c->x86_capability);
+ /* Report CX8 */
+ set_bit(X86_FEATURE_CX8, c->x86_capability);
+ /* Set 3DNow! on Winchip 2 and above. */
+ if (c->x86_model >=8)
+ set_bit(X86_FEATURE_3DNOW, c->x86_capability);
+ /* 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;
+
+ case 6:
+ init_c3(c);
+ break;
+ }
+}
+
+static unsigned int __cpuinit centaur_size_cache(struct cpuinfo_x86 * c, unsigned int size)
+{
+ /* VIA C3 CPUs (670-68F) need further shifting. */
+ if ((c->x86 == 6) && ((c->x86_model == 7) || (c->x86_model == 8)))
+ size >>= 8;
+
+ /* VIA also screwed up Nehemiah stepping 1, and made
+ it return '65KB' instead of '64KB'
+ - Note, it seems this may only be in engineering samples. */
+ if ((c->x86==6) && (c->x86_model==9) && (c->x86_mask==1) && (size==65))
+ size -=1;
+
+ return size;
+}
+
+static struct cpu_dev centaur_cpu_dev __cpuinitdata = {
+ .c_vendor = "Centaur",
+ .c_ident = { "CentaurHauls" },
+ .c_init = init_centaur,
+ .c_size_cache = centaur_size_cache,
+};
+
+int __init centaur_init_cpu(void)
+{
+ cpu_devs[X86_VENDOR_CENTAUR] = &centaur_cpu_dev;
+ return 0;
+}
diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c
new file mode 100644
index 00000000000..d506201d397
--- /dev/null
+++ b/arch/x86/kernel/cpu/common.c
@@ -0,0 +1,733 @@
+#include <linux/init.h>
+#include <linux/string.h>
+#include <linux/delay.h>
+#include <linux/smp.h>
+#include <linux/module.h>
+#include <linux/percpu.h>
+#include <linux/bootmem.h>
+#include <asm/semaphore.h>
+#include <asm/processor.h>
+#include <asm/i387.h>
+#include <asm/msr.h>