Merge branch 'x86/core' into cpus4096

22 files changed, 967 insertions, 401 deletions

diff --git a/arch/x86/kernel/cpu/addon_cpuid_features.c b/arch/x86/kernel/cpu/addon_cpuid_features.c
index 2cf23634b6d9..6882a735d9c0 100644
--- a/arch/x86/kernel/cpu/addon_cpuid_features.c
+++ b/arch/x86/kernel/cpu/addon_cpuid_features.c
@@ -7,7 +7,7 @@
 #include <asm/pat.h>
 #include <asm/processor.h>
 
-#include <mach_apic.h>
+#include <asm/apic.h>
 
 struct cpuid_bit {
 	u16 feature;
@@ -69,7 +69,7 @@ void __cpuinit init_scattered_cpuid_features(struct cpuinfo_x86 *c)
  */
 void __cpuinit detect_extended_topology(struct cpuinfo_x86 *c)
 {
-#ifdef CONFIG_X86_SMP
+#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;
@@ -116,22 +116,14 @@ void __cpuinit detect_extended_topology(struct cpuinfo_x86 *c)
 
 	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)
+	c->cpu_core_id = apic->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);
+	c->phys_proc_id = apic->phys_pkg_id(c->initial_apicid, core_plus_mask_width);
 	/*
 	 * Reinit the apicid, now that we have extended initial_apicid.
 	 */
-	c->apicid = phys_pkg_id(c->initial_apicid, 0);
-#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);
-	/*
-	 * Reinit the apicid, now that we have extended initial_apicid.
-	 */
-	c->apicid = phys_pkg_id(0);
-#endif
+	c->apicid = apic->phys_pkg_id(c->initial_apicid, 0);
+
 	c->x86_max_cores = (core_level_siblings / smp_num_siblings);
 
 
@@ -143,37 +135,3 @@ void __cpuinit detect_extended_topology(struct cpuinfo_x86 *c)
 	return;
 #endif
 }
-
-#ifdef CONFIG_X86_PAT
-void __cpuinit validate_pat_support(struct cpuinfo_x86 *c)
-{
-	if (!cpu_has_pat)
-		pat_disable("PAT not supported by CPU.");
-
-	switch (c->x86_vendor) {
-	case X86_VENDOR_INTEL:
-		/*
-		 * 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 > 0x6 || (c->x86 == 6 && c->x86_model >= 15))
-			return;
-
-		pat_disable("PAT WC disabled due to known CPU erratum.");
-		return;
-
-	case X86_VENDOR_AMD:
-	case X86_VENDOR_CENTAUR:
-	case X86_VENDOR_TRANSMETA:
-		return;
-	}
-
-	pat_disable("PAT disabled. Not yet verified on this CPU type.");
-}
-#endif
diff --git a/arch/x86/kernel/cpu/amd.c b/arch/x86/kernel/cpu/amd.c
index 7c878f6aa919..f47df59016c5 100644
--- a/arch/x86/kernel/cpu/amd.c
+++ b/arch/x86/kernel/cpu/amd.c
@@ -5,6 +5,7 @@
 #include <asm/io.h>
 #include <asm/processor.h>
 #include <asm/apic.h>
+#include <asm/cpu.h>
 
 #ifdef CONFIG_X86_64
 # include <asm/numa_64.h>
@@ -12,8 +13,6 @@
 # include <asm/cacheflush.h>
 #endif
 
-#include <mach_apic.h>
-
 #include "cpu.h"
 
 #ifdef CONFIG_X86_32
@@ -143,6 +142,55 @@ static void __cpuinit init_amd_k6(struct cpuinfo_x86 *c)
 	}
 }
 
+static void __cpuinit amd_k7_smp_check(struct cpuinfo_x86 *c)
+{
+#ifdef CONFIG_SMP
+	/* calling is from identify_secondary_cpu() ? */
+	if (c->cpu_index == boot_cpu_id)
+		return;
+
+	/*
+	 * Certain Athlons might work (for various values of 'work') in SMP
+	 * but they are not certified as MP capable.
+	 */
+	/* Athlon 660/661 is valid. */
+	if ((c->x86_model == 6) && ((c->x86_mask == 0) ||
+	    (c->x86_mask == 1)))
+		goto valid_k7;
+
+	/* Duron 670 is valid */
+	if ((c->x86_model == 7) && (c->x86_mask == 0))
+		goto valid_k7;
+
+	/*
+	 * Athlon 662, Duron 671, and Athlon >model 7 have capability
+	 * bit. It's worth noting that the A5 stepping (662) of some
+	 * Athlon XP's have the MP bit set.
+	 * See http://www.heise.de/newsticker/data/jow-18.10.01-000 for
+	 * more.
+	 */
+	if (((c->x86_model == 6) && (c->x86_mask >= 2)) ||
+	    ((c->x86_model == 7) && (c->x86_mask >= 1)) ||
+	     (c->x86_model > 7))
+		if (cpu_has_mp)
+			goto valid_k7;
+
+	/* If we get here, not a certified SMP capable AMD system. */
+
+	/*
+	 * Don't taint if we are running SMP kernel on a single non-MP
+	 * approved Athlon
+	 */
+	WARN_ONCE(1, "WARNING: This combination of AMD"
+		"processors is not suitable for SMP.\n");
+	if (!test_taint(TAINT_UNSAFE_SMP))
+		add_taint(TAINT_UNSAFE_SMP);
+
+valid_k7:
+	;
+#endif
+}
+
 static void __cpuinit init_amd_k7(struct cpuinfo_x86 *c)
 {
 	u32 l, h;
@@ -177,6 +225,8 @@ static void __cpuinit init_amd_k7(struct cpuinfo_x86 *c)
 	}
 
 	set_cpu_cap(c, X86_FEATURE_K7);
+
+	amd_k7_smp_check(c);
 }
 #endif
 
diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c
index 83492b1f93b1..826d5c876278 100644
--- a/arch/x86/kernel/cpu/common.c
+++ b/arch/x86/kernel/cpu/common.c
@@ -21,14 +21,14 @@
 #include <asm/asm.h>
 #include <asm/numa.h>
 #include <asm/smp.h>
-#ifdef CONFIG_X86_LOCAL_APIC
-#include <asm/mpspec.h>
+#include <asm/cpu.h>
+#include <asm/cpumask.h>
 #include <asm/apic.h>
-#include <mach_apic.h>
-#include <asm/genapic.h>
+
+#ifdef CONFIG_X86_LOCAL_APIC
+#include <asm/uv/uv.h>
 #endif
 
-#include <asm/pda.h>
 #include <asm/pgtable.h>
 #include <asm/processor.h>
 #include <asm/desc.h>
@@ -37,6 +37,7 @@
 #include <asm/sections.h>
 #include <asm/setup.h>
 #include <asm/hypervisor.h>
+#include <asm/stackprotector.h>
 
 #include "cpu.h"
 
@@ -50,6 +51,15 @@ cpumask_var_t cpu_initialized_mask;
 /* representing cpus for which sibling maps can be computed */
 cpumask_var_t cpu_sibling_setup_mask;
 
+/* correctly size the local cpu masks */
+void __init setup_cpu_local_masks(void)
+{
+	alloc_bootmem_cpumask_var(&cpu_initialized_mask);
+	alloc_bootmem_cpumask_var(&cpu_callin_mask);
+	alloc_bootmem_cpumask_var(&cpu_callout_mask);
+	alloc_bootmem_cpumask_var(&cpu_sibling_setup_mask);
+}
+
 #else /* CONFIG_X86_32 */
 
 cpumask_t cpu_callin_map;
@@ -62,23 +72,23 @@ cpumask_t cpu_sibling_setup_map;
 
 static struct cpu_dev *this_cpu __cpuinitdata;
 
+DEFINE_PER_CPU_PAGE_ALIGNED(struct gdt_page, gdt_page) = { .gdt = {
 #ifdef CONFIG_X86_64
-/* We need valid kernel segments for data and code in long mode too
- * IRET will check the segment types  kkeil 2000/10/28
- * Also sysret mandates a special GDT layout
- */
-/* The TLS descriptors are currently at a different place compared to i386.
-   Hopefully nobody expects them at a fixed place (Wine?) */
-DEFINE_PER_CPU(struct gdt_page, gdt_page) = { .gdt = {
+	/*
+	 * We need valid kernel segments for data and code in long mode too
+	 * IRET will check the segment types  kkeil 2000/10/28
+	 * Also sysret mandates a special GDT layout
+	 *
+	 * The TLS descriptors are currently at a different place compared to i386.
+	 * Hopefully nobody expects them at a fixed place (Wine?)
+	 */
 	[GDT_ENTRY_KERNEL32_CS] = { { { 0x0000ffff, 0x00cf9b00 } } },
 	[GDT_ENTRY_KERNEL_CS] = { { { 0x0000ffff, 0x00af9b00 } } },
 	[GDT_ENTRY_KERNEL_DS] = { { { 0x0000ffff, 0x00cf9300 } } },
 	[GDT_ENTRY_DEFAULT_USER32_CS] = { { { 0x0000ffff, 0x00cffb00 } } },
 	[GDT_ENTRY_DEFAULT_USER_DS] = { { { 0x0000ffff, 0x00cff300 } } },
 	[GDT_ENTRY_DEFAULT_USER_CS] = { { { 0x0000ffff, 0x00affb00 } } },
-} };
 #else
-DEFINE_PER_CPU_PAGE_ALIGNED(struct gdt_page, gdt_page) = { .gdt = {
 	[GDT_ENTRY_KERNEL_CS] = { { { 0x0000ffff, 0x00cf9a00 } } },
 	[GDT_ENTRY_KERNEL_DS] = { { { 0x0000ffff, 0x00cf9200 } } },
 	[GDT_ENTRY_DEFAULT_USER_CS] = { { { 0x0000ffff, 0x00cffa00 } } },
@@ -110,9 +120,10 @@ DEFINE_PER_CPU_PAGE_ALIGNED(struct gdt_page, gdt_page) = { .gdt = {
 	[GDT_ENTRY_APMBIOS_BASE+2] = { { { 0x0000ffff, 0x00409200 } } },
 
 	[GDT_ENTRY_ESPFIX_SS] = { { { 0x00000000, 0x00c09200 } } },
-	[GDT_ENTRY_PERCPU] = { { { 0x00000000, 0x00000000 } } },
-} };
+	[GDT_ENTRY_PERCPU] = { { { 0x0000ffff, 0x00cf9200 } } },
+	GDT_STACK_CANARY_INIT
 #endif
+} };
 EXPORT_PER_CPU_SYMBOL_GPL(gdt_page);
 
 #ifdef CONFIG_X86_32
@@ -213,6 +224,49 @@ static inline void squash_the_stupid_serial_number(struct cpuinfo_x86 *c)
 #endif
 
 /*
+ * Some CPU features depend on higher CPUID levels, which may not always
+ * be available due to CPUID level capping or broken virtualization
+ * software.  Add those features to this table to auto-disable them.
+ */
+struct cpuid_dependent_feature {
+	u32 feature;
+	u32 level;
+};
+static const struct cpuid_dependent_feature __cpuinitconst
+cpuid_dependent_features[] = {
+	{ X86_FEATURE_MWAIT,		0x00000005 },
+	{ X86_FEATURE_DCA,		0x00000009 },
+	{ X86_FEATURE_XSAVE,		0x0000000d },
+	{ 0, 0 }
+};
+
+static void __cpuinit filter_cpuid_features(struct cpuinfo_x86 *c, bool warn)
+{
+	const struct cpuid_dependent_feature *df;
+	for (df = cpuid_dependent_features; df->feature; df++) {
+		/*
+		 * Note: cpuid_level is set to -1 if unavailable, but
+		 * extended_extended_level is set to 0 if unavailable
+		 * and the legitimate extended levels are all negative
+		 * when signed; hence the weird messing around with
+		 * signs here...
+		 */
+		if (cpu_has(c, df->feature) &&
+		    ((s32)df->level < 0 ?
+		     (u32)df->level > (u32)c->extended_cpuid_level :
+		     (s32)df->level > (s32)c->cpuid_level)) {
+			clear_cpu_cap(c, df->feature);
+			if (warn)
+				printk(KERN_WARNING
+				       "CPU: CPU feature %s disabled "
+				       "due to lack of CPUID level 0x%x\n",
+				       x86_cap_flags[df->feature],
+				       df->level);
+		}
+	}
+}
+
+/*
  * Naming convention should be: <Name> [(<Codename>)]
  * This table only is used unless init_<vendor>() below doesn't set it;
  * in particular, if CPUID levels 0x80000002..4 are supported, this isn't used
@@ -242,18 +296,29 @@ static char __cpuinit *table_lookup_model(struct cpuinfo_x86 *c)
 
 __u32 cleared_cpu_caps[NCAPINTS] __cpuinitdata;
 
+void load_percpu_segment(int cpu)
+{
+#ifdef CONFIG_X86_32
+	loadsegment(fs, __KERNEL_PERCPU);
+#else
+	loadsegment(gs, 0);
+	wrmsrl(MSR_GS_BASE, (unsigned long)per_cpu(irq_stack_union.gs_base, cpu));
+#endif
+	load_stack_canary_segment();
+}
+
 /* Current gdt points %fs at the "master" per-cpu area: after this,
  * it's on the real one. */
-void switch_to_new_gdt(void)
+void switch_to_new_gdt(int cpu)
 {
 	struct desc_ptr gdt_descr;
 
-	gdt_descr.address = (long)get_cpu_gdt_table(smp_processor_id());
+	gdt_descr.address = (long)get_cpu_gdt_table(cpu);
 	gdt_descr.size = GDT_SIZE - 1;
 	load_gdt(&gdt_descr);
-#ifdef CONFIG_X86_32
-	asm("mov %0, %%fs" : : "r" (__KERNEL_PERCPU) : "memory");
-#endif
+	/* Reload the per-cpu base */
+
+	load_percpu_segment(cpu);
 }
 
 static struct cpu_dev *cpu_devs[X86_VENDOR_NUM] = {};
@@ -383,11 +448,7 @@ void __cpuinit detect_ht(struct cpuinfo_x86 *c)
 		}
 
 		index_msb = get_count_order(smp_num_siblings);
-#ifdef CONFIG_X86_64
-		c->phys_proc_id = phys_pkg_id(index_msb);
-#else
-		c->phys_proc_id = phys_pkg_id(c->initial_apicid, index_msb);
-#endif
+		c->phys_proc_id = apic->phys_pkg_id(c->initial_apicid, index_msb);
 
 		smp_num_siblings = smp_num_siblings / c->x86_max_cores;
 
@@ -395,13 +456,8 @@ void __cpuinit detect_ht(struct cpuinfo_x86 *c)
 
 		core_bits = get_count_order(c->x86_max_cores);
 
-#ifdef CONFIG_X86_64
-		c->cpu_core_id = phys_pkg_id(index_msb) &
+		c->cpu_core_id = apic->phys_pkg_id(c->initial_apicid, index_msb) &
 					       ((1 << core_bits) - 1);
-#else
-		c->cpu_core_id = phys_pkg_id(c->initial_apicid, index_msb) &
-					       ((1 << core_bits) - 1);
-#endif
 	}
 
 out:
@@ -570,11 +626,10 @@ static void __init early_identify_cpu(struct cpuinfo_x86 *c)
 	if (this_cpu->c_early_init)
 		this_cpu->c_early_init(c);
 
-	validate_pat_support(c);
-
 #ifdef CONFIG_SMP
 	c->cpu_index = boot_cpu_id;
 #endif
+	filter_cpuid_features(c, false);
 }
 
 void __init early_cpu_init(void)
@@ -637,7 +692,7 @@ static void __cpuinit generic_identify(struct cpuinfo_x86 *c)
 		c->initial_apicid = (cpuid_ebx(1) >> 24) & 0xFF;
 #ifdef CONFIG_X86_32
 # ifdef CONFIG_X86_HT
-		c->apicid = phys_pkg_id(c->initial_apicid, 0);
+		c->apicid = apic->phys_pkg_id(c->initial_apicid, 0);
 # else
 		c->apicid = c->initial_apicid;
 # endif
@@ -684,7 +739,7 @@ static void __cpuinit identify_cpu(struct cpuinfo_x86 *c)
 		this_cpu->c_identify(c);
 
 #ifdef CONFIG_X86_64
-	c->apicid = phys_pkg_id(0);
+	c->apicid = apic->phys_pkg_id(c->initial_apicid, 0);
 #endif
 
 	/*
@@ -708,6 +763,9 @@ static void __cpuinit identify_cpu(struct cpuinfo_x86 *c)
 	 * we do "generic changes."
 	 */
 
+	/* Filter out anything that depends on CPUID levels we don't have */
+	filter_cpuid_features(c, true);
+
 	/* If the model name is still unset, do table lookup. */
 	if (!c->x86_model_id[0]) {
 		char *p;
@@ -877,54 +935,22 @@ static __init int setup_disablecpuid(char *arg)
 __setup("clearcpuid=", setup_disablecpuid);
 
 #ifdef CONFIG_X86_64
-struct x8664_pda **_cpu_pda __read_mostly;
-EXPORT_SYMBOL(_cpu_pda);
-
 struct desc_ptr idt_descr = { 256 * 16 - 1, (unsigned long) idt_table };
 
-static char boot_cpu_stack[IRQSTACKSIZE] __page_aligned_bss;
+DEFINE_PER_CPU_FIRST(union irq_stack_union,
+		     irq_stack_union) __aligned(PAGE_SIZE);
+DEFINE_PER_CPU(char *, irq_stack_ptr) =
+	init_per_cpu_var(irq_stack_union.irq_stack) + IRQ_STACK_SIZE - 64;
 
-void __cpuinit pda_init(int cpu)
-{
-	struct x8664_pda *pda = cpu_pda(cpu);
+DEFINE_PER_CPU(unsigned long, kernel_stack) =
+	(unsigned long)&init_thread_union - KERNEL_STACK_OFFSET + THREAD_SIZE;
+EXPORT_PER_CPU_SYMBOL(kernel_stack);
 
-	/* Setup up data that may be needed in __get_free_pages early */
-	loadsegment(fs, 0);
-	loadsegment(gs, 0);
-	/* Memory clobbers used to order PDA accessed */
-	mb();
-	wrmsrl(MSR_GS_BASE, pda);
-	mb();
-
-	pda->cpunumber = cpu;
-	pda->irqcount = -1;
-	pda->kernelstack = (unsigned long)stack_thread_info() -
-				 PDA_STACKOFFSET + THREAD_SIZE;
-	pda->active_mm = &init_mm;
-	pda->mmu_state = 0;
-
-	if (cpu == 0) {
-		/* others are initialized in smpboot.c */
-		pda->pcurrent = &init_task;
-		pda->irqstackptr = boot_cpu_stack;
-		pda->irqstackptr += IRQSTACKSIZE - 64;
-	} else {
-		if (!pda->irqstackptr) {
-			pda->irqstackptr = (char *)
-				__get_free_pages(GFP_ATOMIC, IRQSTACK_ORDER);
-			if (!pda->irqstackptr)
-				panic("cannot allocate irqstack for cpu %d",
-				      cpu);
-			pda->irqstackptr += IRQSTACKSIZE - 64;
-		}
+DEFINE_PER_CPU(unsigned int, irq_count) = -1;
 
-		if (pda->nodenumber == 0 && cpu_to_node(cpu) != NUMA_NO_NODE)
-			pda->nodenumber = cpu_to_node(cpu);
-	}
-}
-
-static char boot_exception_stacks[(N_EXCEPTION_STACKS - 1) * EXCEPTION_STKSZ +
-				  DEBUG_STKSZ] __page_aligned_bss;
+static DEFINE_PER_CPU_PAGE_ALIGNED(char, exception_stacks
+	[(N_EXCEPTION_STACKS - 1) * EXCEPTION_STKSZ + DEBUG_STKSZ])
+	__aligned(PAGE_SIZE);
 
 extern asmlinkage void ignore_sysret(void);
 
@@ -957,16 +983,21 @@ unsigned long kernel_eflags;
  */
 DEFINE_PER_CPU(struct orig_ist, orig_ist);
 
-#else
+#else	/* x86_64 */
 
-/* Make sure %fs is initialized properly in idle threads */
+#ifdef CONFIG_CC_STACKPROTECTOR
+DEFINE_PER_CPU(unsigned long, stack_canary);
+#endif
+
+/* Make sure %fs and %gs are initialized properly in idle threads */
 struct pt_regs * __cpuinit idle_regs(struct pt_regs *regs)
 {
 	memset(regs, 0, sizeof(struct pt_regs));
 	regs->fs = __KERNEL_PERCPU;
+	regs->gs = __KERNEL_STACK_CANARY;
 	return regs;
 }
-#endif
+#endif	/* x86_64 */
 
 /*
  * cpu_init() initializes state that is per-CPU. Some data is already
@@ -982,15 +1013,14 @@ void __cpuinit cpu_init(void)
 	struct tss_struct *t = &per_cpu(init_tss, cpu);
 	struct orig_ist *orig_ist = &per_cpu(orig_ist, cpu);
 	unsigned long v;
-	char *estacks = NULL;
 	struct task_struct *me;
 	int i;
 
-	/* CPU 0 is initialised in head64.c */
-	if (cpu != 0)
-		pda_init(cpu);
-	else
-		estacks = boot_exception_stacks;
+#ifdef CONFIG_NUMA
+	if (cpu != 0 && percpu_read(node_number) == 0 &&
+	    cpu_to_node(cpu) != NUMA_NO_NODE)
+		percpu_write(node_number, cpu_to_node(cpu));
+#endif
 
 	me = current;
 
@@ -1006,7 +1036,9 @@ void __cpuinit cpu_init(void)
 	 * and set up the GDT descriptor:
 	 */
 
-	switch_to_new_gdt();
+	switch_to_new_gdt(cpu);
+	loadsegment(fs, 0);
+
 	load_idt((const struct desc_ptr *)&idt_descr);
 
 	memset(me->thread.tls_array, 0, GDT_ENTRY_TLS_ENTRIES * 8);
@@ -1017,25 +1049,20 @@ void __cpuinit cpu_init(void)
 	barrier();
 
 	check_efer();
-	if (cpu != 0 && x2apic)
+	if (cpu != 0)
 		enable_x2apic();
 
 	/*
 	 * set up and load the per-CPU TSS
 	 */
 	if (!orig_ist->ist[0]) {
-		static const unsigned int order[N_EXCEPTION_STACKS] = {
-		  [0 ... N_EXCEPTION_STACKS - 1] = EXCEPTION_STACK_ORDER,
-		  [DEBUG_STACK - 1] = DEBUG_STACK_ORDER
+		static const unsigned int sizes[N_EXCEPTION_STACKS] = {
+		  [0 ... N_EXCEPTION_STACKS - 1] = EXCEPTION_STKSZ,
+		  [DEBUG_STACK - 1] = DEBUG_STKSZ
 		};
+		char *estacks = per_cpu(exception_stacks, cpu);
 		for (v = 0; v < N_EXCEPTION_STACKS; v++) {
-			if (cpu) {
-				estacks = (char *)__get_free_pages(GFP_ATOMIC, order[v]);
-				if (!estacks)
-					panic("Cannot allocate exception "
-					      "stack %ld %d\n", v, cpu);
-			}
-			estacks += PAGE_SIZE << order[v];
+			estacks += sizes[v];
 			orig_ist->ist[v] = t->x86_tss.ist[v] =
 					(unsigned long)estacks;
 		}
@@ -1069,22 +1096,19 @@ void __cpuinit cpu_init(void)
 	 */
 	if (kgdb_connected && arch_kgdb_ops.correct_hw_break)
 		arch_kgdb_ops.correct_hw_break();
-	else {
+	else
 #endif
-	/*
-	 * Clear all 6 debug registers:
-	 */
-
-	set_debugreg(0UL, 0);
-	set_debugreg(0UL, 1);
-	set_debugreg(0UL, 2);
-	set_debugreg(0UL, 3);
-	set_debugreg(0UL, 6);
-	set_debugreg(0UL, 7);
-#ifdef CONFIG_KGDB
-	/* If the kgdb is connected no debug regs should be altered. */
+	{
+		/*
+		 * Clear all 6 debug registers:
+		 */
+		set_debugreg(0UL, 0);
+		set_debugreg(0UL, 1);
+		set_debugreg(0UL, 2);
+		set_debugreg(0UL, 3);
+		set_debugreg(0UL, 6);
+		set_debugreg(0UL, 7);
 	}
-#endif
 
 	fpu_init();
 
@@ -1114,7 +1138,7 @@ void __cpuinit cpu_init(void)
 		clear_in_cr4(X86_CR4_VME|X86_CR4_PVI|X86_CR4_TSD|X86_CR4_DE);
 
 	load_idt(&idt_descr);
-	switch_to_new_gdt();
+	switch_to_new_gdt(cpu);
 
 	/*
 	 * Set up and load the per-CPU TSS and LDT
@@ -1135,9 +1159,6 @@ void __cpuinit cpu_init(void)
 	__set_tss_desc(cpu, GDT_ENTRY_DOUBLEFAULT_TSS, &doublefault_tss);
 #endif
 
-	/* Clear %gs. */
-	asm volatile ("mov %0, %%gs" : : "r" (0));
-
 	/* Clear all 6 debug registers: */
 	set_debugreg(0, 0);
 	set_debugreg(0, 1);
diff --git a/arch/x86/kernel/cpu/cpufreq/Kconfig b/arch/x86/kernel/cpu/cpufreq/Kconfig
index efae3b22a0ff..65792c2cc462 100644
--- a/arch/x86/kernel/cpu/cpufreq/Kconfig
+++ b/arch/x86/kernel/cpu/cpufreq/Kconfig
@@ -245,17 +245,6 @@ config X86_E_POWERSAVER
 
 comment "shared options"
 
-config X86_ACPI_CPUFREQ_PROC_INTF
-	bool "/proc/acpi/processor/../performance interface (deprecated)"
-	depends on PROC_FS
-	depends on X86_ACPI_CPUFREQ || X86_POWERNOW_K7_ACPI || X86_POWERNOW_K8_ACPI
-	help
-	  This enables the deprecated /proc/acpi/processor/../performance
-	  interface. While it is helpful for debugging, the generic,
-	  cross-architecture cpufreq interfaces should be used.
-
-	  If in doubt, say N.
-
 config X86_SPEEDSTEP_LIB
 	tristate
 	default (X86_SPEEDSTEP_ICH || X86_SPEEDSTEP_SMI || X86_P4_CLOCKMOD)
diff --git a/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c b/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c
index 4b1c319d30c3..22590cf688ae 100644
--- a/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c
+++ b/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c
@@ -601,7 +601,7 @@ static int acpi_cpufreq_cpu_init(struct cpufreq_policy *policy)
 	if (!data)
 		return -ENOMEM;
 
-	data->acpi_data = percpu_ptr(acpi_perf_data, cpu);
+	data->acpi_data = per_cpu_ptr(acpi_perf_data, cpu);
 	per_cpu(drv_data, cpu) = data;
 
 	if (cpu_has(c, X86_FEATURE_CONSTANT_TSC))
diff --git a/arch/x86/kernel/cpu/cpufreq/e_powersaver.c b/arch/x86/kernel/cpu/cpufreq/e_powersaver.c
index c2f930d86640..41ab3f064cb1 100644
--- a/arch/x86/kernel/cpu/cpufreq/e_powersaver.c
+++ b/arch/x86/kernel/cpu/cpufreq/e_powersaver.c
@@ -204,12 +204,12 @@ static int eps_cpu_init(struct cpufreq_policy *policy)
 	}
 	/* Enable Enhanced PowerSaver */
 	rdmsrl(MSR_IA32_MISC_ENABLE, val);
-	if (!(val & 1 << 16)) {
-		val |= 1 << 16;
+	if (!(val & MSR_IA32_MISC_ENABLE_ENHANCED_SPEEDSTEP)) {
+		val |= MSR_IA32_MISC_ENABLE_ENHANCED_SPEEDSTEP;
 		wrmsrl(MSR_IA32_MISC_ENABLE, val);
 		/* Can be locked at 0 */
 		rdmsrl(MSR_IA32_MISC_ENABLE, val);
-		if (!(val & 1 << 16)) {
+		if (!(val & MSR_IA32_MISC_ENABLE_ENHANCED_SPEEDSTEP)) {
 			printk(KERN_INFO "eps: Can't enable Enhanced PowerSaver\n");
 			return -ENODEV;
 		}
diff --git a/arch/x86/kernel/cpu/cpufreq/p4-clockmod.c b/arch/x86/kernel/cpu/cpufreq/p4-clockmod.c
index b585e04cbc9e..3178c3acd97e 100644
--- a/arch/x86/kernel/cpu/cpufreq/p4-clockmod.c
+++ b/arch/x86/kernel/cpu/cpufreq/p4-clockmod.c
@@ -277,7 +277,6 @@ static struct cpufreq_driver p4clockmod_driver = {
 	.name		= "p4-clockmod",
 	.owner		= THIS_MODULE,
 	.attr		= p4clockmod_attr,
-	.hide_interface	= 1,
 };
 
 
diff --git a/arch/x86/kernel/cpu/cpufreq/powernow-k8.c b/arch/x86/kernel/cpu/cpufreq/powernow-k8.c
index 5c28b37dea11..6428aa17b40e 100644
--- a/arch/x86/kernel/cpu/cpufreq/powernow-k8.c
+++ b/arch/x86/kernel/cpu/cpufreq/powernow-k8.c
@@ -939,10 +939,25 @@ static void powernow_k8_cpu_exit_acpi(struct powernow_k8_data *data)
 	free_cpumask_var(data->acpi_data.shared_cpu_map);
 }
 
+static int get_transition_latency(struct powernow_k8_data *data)
+{
+	int max_latency = 0;
+	int i;
+	for (i = 0; i < data->acpi_data.state_count; i++) {
+		int cur_latency = data->acpi_data.states[i].transition_latency
+			+ data->acpi_data.states[i].bus_master_latency;
+		if (cur_latency > max_latency)
+			max_latency = cur_latency;
+	}
+	/* value in usecs, needs to be in nanoseconds */
+	return 1000 * max_latency;
+}
+
 #else
 static int powernow_k8_cpu_init_acpi(struct powernow_k8_data *data) { return -ENODEV; }
 static void powernow_k8_cpu_exit_acpi(struct powernow_k8_data *data) { return; }
 static void powernow_k8_acpi_pst_values(struct powernow_k8_data *data, unsigned int index) { return; }
+static int get_transition_latency(struct powernow_k8_data *data) { return 0; }
 #endif /* CONFIG_X86_POWERNOW_K8_ACPI */
 
 /* Take a frequency, and issue the fid/vid transition command */
@@ -1142,8 +1157,7 @@ static int __cpuinit powernowk8_cpu_init(struct cpufreq_policy *pol)
 	data->cpu = pol->cpu;
 	data->currpstate = HW_PSTATE_INVALID;
 
-	rc = powernow_k8_cpu_init_acpi(data);
-	if (rc) {
+	if (powernow_k8_cpu_init_acpi(data)) {
 		/*
 		 * Use the PSB BIOS structure. This is only availabe on
 		 * an UP version, and is deprecated by AMD.
@@ -1161,19 +1175,28 @@ static int __cpuinit powernowk8_cpu_init(struct cpufreq_policy *pol)
 			       "ACPI maintainers and complain to your BIOS "
 			       "vendor.\n");
 #endif
-			goto err_out;
+			kfree(data);
+			return -ENODEV;
 		}
 		if (pol->cpu != 0) {
 			printk(KERN_ERR FW_BUG PFX "No ACPI _PSS objects for "
 			       "CPU other than CPU0. Complain to your BIOS "
 			       "vendor.\n");
-			goto err_out;
+			kfree(data);
+			return -ENODEV;
 		}
 		rc = find_psb_table(data);
 		if (rc) {
-			goto err_out;
+			kfree(data);
+			return -ENODEV;
 		}
-	}
+		/* Take a crude guess here.
+		 * That guess was in microseconds, so multiply with 1000 */
+		pol->cpuinfo.transition_latency = (
+			 ((data->rvo + 8) * data->vstable * VST_UNITS_20US) +
+			 ((1 << data->irt) * 30)) * 1000;
+	} else /* ACPI _PSS objects available */
+		pol->cpuinfo.transition_latency = get_transition_latency(data);
 
 	/* only run on specific CPU from here on */
 	oldmask = current->cpus_allowed;
@@ -1204,11 +1227,6 @@ static int __cpuinit powernowk8_cpu_init(struct cpufreq_policy *pol)
 		cpumask_copy(pol->cpus, &per_cpu(cpu_core_map, pol->cpu));
 	data->available_cores = pol->cpus;
 
-	/* Take a crude guess here.
-	 * That guess was in microseconds, so multiply with 1000 */
-	pol->cpuinfo.transition_latency = (((data->rvo + 8) * data->vstable * VST_UNITS_20US)
-	    + (3 * (1 << data->irt) * 10)) * 1000;
-
 	if (cpu_family == CPU_HW_PSTATE)
 		pol->cur = find_khz_freq_from_pstate(data->powernow_table, data->currpstate);
 	else
diff --git a/arch/x86/kernel/cpu/cpufreq/speedstep-centrino.c b/arch/x86/kernel/cpu/cpufreq/speedstep-centrino.c
index f08998278a3a..c9f1fdc02830 100644
--- a/arch/x86/kernel/cpu/cpufreq/speedstep-centrino.c
+++ b/arch/x86/kernel/cpu/cpufreq/speedstep-centrino.c
@@ -390,14 +390,14 @@ static int centrino_cpu_init(struct cpufreq_policy *policy)
 	   enable it if not. */
 	rdmsr(MSR_IA32_MISC_ENABLE, l, h);
 
-	if (!(l & (1<<16))) {
-		l |= (1<<16);
+	if (!(l & MSR_IA32_MISC_ENABLE_ENHANCED_SPEEDSTEP)) {
+		l |= MSR_IA32_MISC_ENABLE_ENHANCED_SPEEDSTEP;
 		dprintk("trying to enable Enhanced SpeedStep (%x)\n", l);
 		wrmsr(MSR_IA32_MISC_ENABLE, l, h);
 
 		/* check to see if it stuck */
 		rdmsr(MSR_IA32_MISC_ENABLE, l, h);
-		if (!(l & (1<<16))) {
+		if (!(l & MSR_IA32_MISC_ENABLE_ENHANCED_SPEEDSTEP)) {
 			printk(KERN_INFO PFX
 				"couldn't enable Enhanced SpeedStep\n");
 			return -ENODEV;
diff --git a/arch/x86/kernel/cpu/intel.c b/arch/x86/kernel/cpu/intel.c
index 8ea6929e974c..191117f1ad51 100644
--- a/arch/x86/kernel/cpu/intel.c
+++ b/arch/x86/kernel/cpu/intel.c
@@ -13,6 +13,7 @@
 #include <asm/uaccess.h>
 #include <asm/ds.h>
 #include <asm/bugs.h>
+#include <asm/cpu.h>
 
 #ifdef CONFIG_X86_64
 #include <asm/topology.h>
@@ -24,11 +25,23 @@
 #ifdef CONFIG_X86_LOCAL_APIC
 #include <asm/mpspec.h>
 #include <asm/apic.h>
-#include <mach_apic.h>
 #endif
 
 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);
@@ -50,6 +63,18 @@ static void __cpuinit early_init_intel(struct cpuinfo_x86 *c)
 		set_cpu_cap(c, X86_FEATURE_NONSTOP_TSC);
 	}
 
+	/*
+	 * 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_X86_32
@@ -86,6 +111,28 @@ static void __cpuinit trap_init_f00f_bug(void)
 }
 #endif
 
+static void __cpuinit intel_smp_check(struct cpuinfo_x86 *c)
+{
+#ifdef CONFIG_SMP
+	/* calling is from identify_secondary_cpu() ? */
+	if (c->cpu_index == boot_cpu_id)
+		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");
+	}
+#endif
+}
+
 static void __cpuinit intel_workarounds(struct cpuinfo_x86 *c)
 {
 	unsigned long lo, hi;
@@ -122,10 +169,10 @@ static void __cpuinit intel_workarounds(struct cpuinfo_x86 *c)
 	 */
 	if ((c->x86 == 15) && (c->x86_model == 1) && (c->x86_mask == 1)) {
 		rdmsr(MSR_IA32_MISC_ENABLE, lo, hi);
-		if ((lo & (1<<9)) == 0) {
+		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 |= (1<<9);	/* Disable hw prefetching */
+			lo |= MSR_IA32_MISC_ENABLE_PREFETCH_DISABLE;
 			wrmsr (MSR_IA32_MISC_ENABLE, lo, hi);
 		}
 	}
@@ -162,6 +209,8 @@ static void __cpuinit intel_workarounds(struct cpuinfo_x86 *c)
 #ifdef CONFIG_X86_NUMAQ
 	numaq_tsc_disable();
 #endif
+
+	intel_smp_check(c);
 }
 #else
 static void __cpuinit intel_workarounds(struct cpuinfo_x86 *c)
@@ -278,6 +327,9 @@ static void __cpuinit init_intel(struct cpuinfo_x86 *c)
 		ds_init_intel(c);
 	}
 
+	if (c->x86 == 6 && c->x86_model == 29 && cpu_has_clflush)
+		set_cpu_cap(c, X86_FEATURE_CLFLUSH_MONITOR);
+
 #ifdef CONFIG_X86_64
 	if (c->x86 == 15)
 		c->x86_cache_alignment = c->x86_clflush_size * 2;
diff --git a/arch/x86/kernel/cpu/intel_cacheinfo.c b/arch/x86/kernel/cpu/intel_cacheinfo.c
index 03f93c5dcfb3..8e6ce2c146d6 100644
--- a/arch/x86/kernel/cpu/intel_cacheinfo.c
+++ b/arch/x86/kernel/cpu/intel_cacheinfo.c
@@ -36,8 +36,11 @@ static struct _cache_table cache_table[] __cpuinitdata =
 {
 	{ 0x06, LVL_1_INST, 8 },	/* 4-way set assoc, 32 byte line size */
 	{ 0x08, LVL_1_INST, 16 },	/* 4-way set assoc, 32 byte line size */
+	{ 0x09, LVL_1_INST, 32 },	/* 4-way set assoc, 64 byte line size */
 	{ 0x0a, LVL_1_DATA, 8 },	/* 2 way set assoc, 32 byte line size */
 	{ 0x0c, LVL_1_DATA, 16 },	/* 4-way set assoc, 32 byte line size */
+	{ 0x0d, LVL_1_DATA, 16 },	/* 4-way set assoc, 64 byte line size */
+	{ 0x21, LVL_2,      256 },	/* 8-way set assoc, 64 byte line size */
 	{ 0x22, LVL_3,      512 },	/* 4-way set assoc, sectored cache, 64 byte line size */
 	{ 0x23, LVL_3,      1024 },	/* 8-way set assoc, sectored cache, 64 byte line size */
 	{ 0x25, LVL_3,      2048 },	/* 8-way set assoc, sectored cache, 64 byte line size */
@@ -85,6 +88,18 @@ static struct _cache_table cache_table[] __cpuinitdata =
 	{ 0x85, LVL_2,    2048 },	/* 8-way set assoc, 32 byte line size */
 	{ 0x86, LVL_2,     512 },	/* 4-way set assoc, 64 byte line size */
 	{ 0x87, LVL_2,    1024 },	/* 8-way set assoc, 64 byte line size */
+	{ 0xd0, LVL_3,     512 },	/* 4-way set assoc, 64 byte line size */
+	{ 0xd1, LVL_3,    1024 },	/* 4-way set assoc, 64 byte line size */
+	{ 0xd2, LVL_3,    2048 },	/* 4-way set assoc, 64 byte line size */
+	{ 0xd6, LVL_3,    1024 },	/* 8-way set assoc, 64 byte line size */
+	{ 0xd7, LVL_3,    2038 },	/* 8-way set assoc, 64 byte line size */
+	{ 0xd8, LVL_3,    4096 },	/* 12-way set assoc, 64 byte line size */
+	{ 0xdc, LVL_3,    2048 },	/* 12-way set assoc, 64 byte line size */
+	{ 0xdd, LVL_3,    4096 },	/* 12-way set assoc, 64 byte line size */
+	{ 0xde, LVL_3,    8192 },	/* 12-way set assoc, 64 byte line size */
+	{ 0xe2, LVL_3,    2048 },	/* 16-way set assoc, 64 byte line size */
+	{ 0xe3, LVL_3,    4096 },	/* 16-way set assoc, 64 byte line size */
+	{ 0xe4, LVL_3,    8192 },	/* 16-way set assoc, 64 byte line size */
 	{ 0x00, 0, 0}
 };
 
diff --git a/arch/x86/kernel/cpu/mcheck/Makefile b/arch/x86/kernel/cpu/mcheck/Makefile
index d7d2323bbb69..b2f89829bbe8 100644
--- a/arch/x86/kernel/cpu/mcheck/Makefile
+++ b/arch/x86/kernel/cpu/mcheck/Makefile
@@ -4,3 +4,4 @@ obj-$(CONFIG_X86_32)		+= k7.o p4.o p5.o p6.o winchip.o
 obj-$(CONFIG_X86_MCE_INTEL)	+= mce_intel_64.o
 obj-$(CONFIG_X86_MCE_AMD)	+= mce_amd_64.o
 obj-$(CONFIG_X86_MCE_NONFATAL)	+= non-fatal.o
+obj-$(CONFIG_X86_MCE_THRESHOLD) += threshold.o
diff --git a/arch/x86/kernel/cpu/mcheck/mce_32.c b/arch/x86/kernel/cpu/mcheck/mce_32.c
index dfaebce3633e..3552119b091d 100644
--- a/arch/x86/kernel/cpu/mcheck/mce_32.c
+++ b/arch/x86/kernel/cpu/mcheck/mce_32.c
@@ -60,20 +60,6 @@ void mcheck_init(struct cpuinfo_x86 *c)
 	}
 }
 
-static unsigned long old_cr4 __initdata;
-
-void __init stop_mce(void)
-{
-	old_cr4 = read_cr4();
-	clear_in_cr4(X86_CR4_MCE);
-}
-
-void __init restart_mce(void)
-{
-	if (old_cr4 & X86_CR4_MCE)
-		set_in_cr4(X86_CR4_MCE);
-}
-
 static int __init mcheck_disable(char *str)
 {
 	mce_disabled = 1;
diff --git a/arch/x86/kernel/cpu/mcheck/mce_64.c b/arch/x86/kernel/cpu/mcheck/mce_64.c
index 1c838032fd37..ca14604611ec 100644
--- a/arch/x86/kernel/cpu/mcheck/mce_64.c
+++ b/arch/x86/kernel/cpu/mcheck/mce_64.c
@@ -3,6 +3,8 @@
  * K8 parts Copyright 2002,2003 Andi Kleen, SuSE Labs.
  * Rest from unknown author(s).
  * 2004 Andi Kleen. Rewrote most of it.
+ * Copyright 2008 Intel Corporation
+ * Author: Andi Kleen
  */
 
 #include <linux/init.h>
@@ -24,6 +26,9 @@
 #include <linux/ctype.h>
 #include <linux/kmod.h>
 #include <linux/kdebug.h>
+#include <linux/kobject.h>
+#include <linux/sysfs.h>
+#include <linux/ratelimit.h>
 #include <asm/processor.h>
 #include <asm/msr.h>
 #include <asm/mce.h>
@@ -32,7 +37,6 @@
 #include <asm/idle.h>
 
 #define MISC_MCELOG_MINOR 227
-#define NR_SYSFS_BANKS 6
 
 atomic_t mce_entry;
 
@@ -47,7 +51,7 @@ static int mce_dont_init;
  */
 static int tolerant = 1;
 static int banks;
-static unsigned long bank[NR_SYSFS_BANKS] = { [0 ... NR_SYSFS_BANKS-1] = ~0UL };
+static u64 *bank;
 static unsigned long notify_user;
 static int rip_msr;
 static int mce_bootlog = -1;
@@ -58,6 +62,19 @@ static char *trigger_argv[2] = { trigger, NULL };
 
 static DECLARE_WAIT_QUEUE_HEAD(mce_wait);
 
+/* MCA banks polled by the period polling timer for corrected events */
+DEFINE_PER_CPU(mce_banks_t, mce_poll_banks) = {
+	[0 ... BITS_TO_LONGS(MAX_NR_BANKS)-1] = ~0UL
+};
+
+/* Do initial initialization of a struct mce */
+void mce_setup(struct mce *m)
+{
+	memset(m, 0, sizeof(struct mce));
+	m->cpu = smp_processor_id();
+	rdtscll(m->tsc);
+}
+
 /*
  * Lockless MCE logging infrastructure.
  * This avoids deadlocks on printk locks without having to break locks. Also
@@ -119,11 +136,11 @@ static void print_mce(struct mce *m)
 			print_symbol("{%s}", m->ip);
 		printk("\n");
 	}
-	printk(KERN_EMERG "TSC %Lx ", m->tsc);
+	printk(KERN_EMERG "TSC %llx ", m->tsc);
 	if (m->addr)
-		printk("ADDR %Lx ", m->addr);
+		printk("ADDR %llx ", m->addr);
 	if (m->misc)
-		printk("MISC %Lx ", m->misc);
+		printk("MISC %llx ", m->misc);
 	printk("\n");
 	printk(KERN_EMERG "This is not a software problem!\n");
 	printk(KERN_EMERG "Run through mcelog --ascii to decode "
@@ -149,8 +166,10 @@ static void mce_panic(char *msg, struct mce *backup, unsigned long start)
 	panic(msg);
 }
 
-static int mce_available(struct cpuinfo_x86 *c)
+int mce_available(struct cpuinfo_x86 *c)
 {
+	if (mce_dont_init)
+		return 0;
 	return cpu_has(c, X86_FEATURE_MCE) && cpu_has(c, X86_FEATURE_MCA);
 }
 
@@ -172,7 +191,77 @@ static inline void mce_get_rip(struct mce *m, struct pt_regs *regs)
 }
 
 /*
- * The actual machine check handler
+ * Poll for corrected events or events that happened before reset.
+ * Those are just logged through /dev/mcelog.
+ *
+ * This is executed in standard interrupt context.
+ */
+void machine_check_poll(enum mcp_flags flags, mce_banks_t *b)
+{
+	struct mce m;
+	int i;
+
+	mce_setup(&m);
+
+	rdmsrl(MSR_IA32_MCG_STATUS, m.mcgstatus);
+	for (i = 0; i < banks; i++) {
+		if (!bank[i] || !test_bit(i, *b))
+			continue;
+
+		m.misc = 0;
+		m.addr = 0;
+		m.bank = i;
+		m.tsc = 0;
+
+		barrier();
+		rdmsrl(MSR_IA32_MC0_STATUS + i*4, m.status);
+		if (!(m.status & MCI_STATUS_VAL))
+			continue;
+
+		/*
+		 * Uncorrected events are handled by the exception handler
+		 * when it is enabled. But when the exception is disabled log
+		 * everything.
+		 *
+		 * TBD do the same check for MCI_STATUS_EN here?
+		 */
+		if ((m.status & MCI_STATUS_UC) && !(flags & MCP_UC))
+			continue;
+
+		if (m.status & MCI_STATUS_MISCV)
+			rdmsrl(MSR_IA32_MC0_MISC + i*4, m.misc);
+		if (m.status & MCI_STATUS_ADDRV)
+			rdmsrl(MSR_IA32_MC0_ADDR + i*4, m.addr);
+
+		if (!(flags & MCP_TIMESTAMP))
+			m.tsc = 0;
+		/*
+		 * Don't get the IP here because it's unlikely to
+		 * have anything to do with the actual error location.
+		 */
+
+		mce_log(&m);
+		add_taint(TAINT_MACHINE_CHECK);
+
+		/*
+		 * Clear state for this bank.
+		 */
+		wrmsrl(MSR_IA32_MC0_STATUS+4*i, 0);
+	}
+
+	/*
+	 * Don't clear MCG_STATUS here because it's only defined for
+	 * exceptions.
+	 */
+}
+
+/*
+ * The actual machine check handler. This only handles real
+ * exceptions when something got corrupted coming in through int 18.
+ *
+ * This is executed in NMI context not subject to normal locking rules. This
+ * implies that most kernel services cannot be safely used. Don't even
+ * think about putting a printk in there!
  */
 void do_machine_check(struct pt_regs * regs, long error_code)
 {
@@ -190,17 +279,18 @@ void do_machine_check(struct pt_regs * regs, long error_code)
 	 * error.
 	 */
 	int kill_it = 0;
+	DECLARE_BITMAP(toclear, MAX_NR_BANKS);
 
 	atomic_inc(&mce_entry);
 
-	if ((regs
-	     && notify_die(DIE_NMI, "machine check", regs, error_code,
+	if (notify_die(DIE_NMI, "machine check", regs, error_code,
 			   18, SIGKILL) == NOTIFY_STOP)
-	    || !banks)
+		goto out2;
+	if (!banks)
 		goto out2;
 
-	memset(&m, 0, sizeof(struct mce));
-	m.cpu = smp_processor_id();
+	mce_setup(&m);
+
 	rdmsrl(MSR_IA32_MCG_STATUS, m.mcgstatus);
 	/* if the restart IP is not valid, we're done for */
 	if (!(m.mcgstatus & MCG_STATUS_RIPV))
@@ -210,18 +300,32 @@ void do_machine_check(struct pt_regs * regs, long error_code)
 	barrier();
 
 	for (i = 0; i < banks; i++) {
-		if (i < NR_SYSFS_BANKS && !bank[i])
+		__clear_bit(i, toclear);
+		if (!bank[i])
 			continue;
 
 		m.misc = 0;
 		m.addr = 0;
 		m.bank = i;
-		m.tsc = 0;
 
 		rdmsrl(MSR_IA32_MC0_STATUS + i*4, m.status);
 		if ((m.status & MCI_STATUS_VAL) == 0)
 			continue;
 
+		/*
+		 * Non uncorrected errors are handled by machine_check_poll
+		 * Leave them alone.
+		 */
+		if ((m.status & MCI_STATUS_UC) == 0)
+			continue;
+
+		/*
+		 * Set taint even when machine check was not enabled.
+		 */
+		add_taint(TAINT_MACHINE_CHECK);
+
+		__set_bit(i, toclear);
+
 		if (m.status & MCI_STATUS_EN) {
 			/* if PCC was set, there's no way out */
 			no_way_out |= !!(m.status & MCI_STATUS_PCC);
@@ -235,6 +339,12 @@ void do_machine_check(struct pt_regs * regs, long error_code)
 					no_way_out = 1;
 				kill_it = 1;
 			}
+		} else {
+			/*
+			 * Machine check event was not enabled. Clear, but
+			 * ignore.
+			 */
+			continue;
 		}
 
 		if (m.status & MCI_STATUS_MISCV)
@@ -243,10 +353,7 @@ void do_machine_check(struct pt_regs * regs, long error_code)
 			rdmsrl(MSR_IA32_MC0_ADDR + i*4, m.addr);
 
 		mce_get_rip(&m, regs);
-		if (error_code >= 0)
-			rdtscll(m.tsc);
-		if (error_code != -2)
-			mce_log(&m);
+		mce_log(&m);
 
 		/* Did this bank cause the exception? */
 		/* Assume that the bank with uncorrectable errors did it,
@@ -255,14 +362,8 @@ void do_machine_check(struct pt_regs * regs, long error_code)
 			panicm = m;
 			panicm_found = 1;
 		}
-
-		add_taint(TAINT_MACHINE_CHECK);
 	}
 
-	/* Never do anything final in the polling timer */
-	if (!regs)
-		goto out;
-
 	/* If we didn't find an uncorrectable error, pick
 	   the last one (shouldn't happen, just being safe). */
 	if (!panicm_found)
@@ -295,11 +396,11 @@ void do_machine_check(struct pt_regs * regs, long error_code)
 		 * If we know that the error was in user space, send a
 		 * SIGBUS.  Otherwise, panic if tolerance is low.
 		 *
-		 * do_exit() takes an awful lot of locks and has a slight
+		 * force_sig() takes an awful lot of locks and has a slight
 		 * risk of deadlocking.
 		 */
 		if (user_space) {
-			do_exit(SIGBUS);
+			force_sig(SIGBUS, current);
 		} else if (panic_on_oops || tolerant < 2) {
 			mce_panic("Uncorrected machine check",
 				&panicm, mcestart);
@@ -309,10 +410,11 @@ void do_machine_check(struct pt_regs * regs, long error_code)
 	/* notify userspace ASAP */
 	set_thread_flag(TIF_MCE_NOTIFY);
 
- out:
 	/* the last thing we do is clear state */
-	for (i = 0; i < banks; i++)
-		wrmsrl(MSR_IA32_MC0_STATUS+4*i, 0);
+	for (i = 0; i < banks; i++) {
+		if (test_bit(i, toclear))
+			wrmsrl(MSR_IA32_MC0_STATUS+4*i, 0);
+	}
 	wrmsrl(MSR_IA32_MCG_STATUS, 0);
  out2:
 	atomic_dec(&mce_entry);
@@ -332,15 +434,13 @@ void do_machine_check(struct pt_regs * regs, long error_code)
  * and historically has been the register value of the
  * MSR_IA32_THERMAL_STATUS (Intel) msr.
  */
-void mce_log_therm_throt_event(unsigned int cpu, __u64 status)
+void mce_log_therm_throt_event(__u64 status)
 {
 	struct mce m;
 
-	memset(&m, 0, sizeof(m));
-	m.cpu = cpu;
+	mce_setup(&m);
 	m.bank = MCE_THERMAL_BANK;
 	m.status = status;
-	rdtscll(m.tsc);
 	mce_log(&m);
 }
 #endif /* CONFIG_X86_MCE_INTEL */
@@ -353,18 +453,18 @@ void mce_log_therm_throt_event(unsigned int cpu, __u64 status)
 
 static int check_interval = 5 * 60; /* 5 minutes */
 static int next_interval; /* in jiffies */
-static void mcheck_timer(struct work_struct *work);
-static DECLARE_DELAYED_WORK(mcheck_work, mcheck_timer);
+static void mcheck_timer(unsigned long);
+static DEFINE_PER_CPU(struct timer_list, mce_timer);
 
-static void mcheck_check_cpu(void *info)
+static void mcheck_timer(unsigned long data)
 {
-	if (mce_available(&current_cpu_data))
-		do_machine_check(NULL, 0);
-}
+	struct timer_list *t = &per_cpu(mce_timer, data);
 
-static void mcheck_timer(struct work_struct *work)
-{
-	on_each_cpu(mcheck_check_cpu, NULL, 1);
+	WARN_ON(smp_processor_id() != data);
+
+	if (mce_available(&current_cpu_data))
+		machine_check_poll(MCP_TIMESTAMP,
+				&__get_cpu_var(mce_poll_banks));
 
 	/*
 	 * Alert userspace if needed.  If we logged an MCE, reduce the
@@ -377,31 +477,41 @@ static void mcheck_timer(struct work_struct *work)
 				(int)round_jiffies_relative(check_interval*HZ));
 	}
 
-	schedule_delayed_work(&mcheck_work, next_interval);
+	t->expires = jiffies + next_interval;
+	add_timer(t);
+}
+
+static void mce_do_trigger(struct work_struct *work)
+{
+	call_usermodehelper(trigger, trigger_argv, NULL, UMH_NO_WAIT);
 }
 
+static DECLARE_WORK(mce_trigger_work, mce_do_trigger);
+
 /*
- * This is only called from process context.  This is where we do
- * anything we need to alert userspace about new MCEs.  This is called
- * directly from the poller and also from entry.S and idle, thanks to
- * TIF_MCE_NOTIFY.
+ * Notify the user(s) about new machine check events.
+ * Can be called from interrupt context, but not from machine check/NMI
+ * context.
  */
 int mce_notify_user(void)
 {
+	/* Not more than two messages every minute */
+	static DEFINE_RATELIMIT_STATE(ratelimit, 60*HZ, 2);
+
 	clear_thread_flag(TIF_MCE_NOTIFY);
 	if (test_and_clear_bit(0, &notify_user)) {
-		static unsigned long last_print;
-		unsigned long now = jiffies;
-
 		wake_up_interruptible(&mce_wait);
-		if (trigger[0])
-			call_usermodehelper(trigger, trigger_argv, NULL,
-						UMH_NO_WAIT);
 
-		if (time_after_eq(now, last_print + (check_interval*HZ))) {
-			last_print = now;
+		/*
+		 * There is no risk of missing notifications because
+		 * work_pending is always cleared before the function is
+		 * executed.
+		 */
+		if (trigger[0] && !work_pending(&mce_trigger_work))
+			schedule_work(&mce_trigger_work);
+
+		if (__ratelimit(&ratelimit))
 			printk(KERN_INFO "Machine check events logged\n");
-		}
 
 		return 1;
 	}
@@ -425,63 +535,78 @@ static struct notifier_block mce_idle_notifier = {
 
 static __init int periodic_mcheck_init(void)
 {
-	next_interval = check_interval * HZ;
-	if (next_interval)
-		schedule_delayed_work(&mcheck_work,
-				      round_jiffies_relative(next_interval));
-	idle_notifier_register(&mce_idle_notifier);
-	return 0;
+       idle_notifier_register(&mce_idle_notifier);
+       return 0;
 }
 __initcall(periodic_mcheck_init);
 
-
 /*
  * Initialize Machine Checks for a CPU.
  */
-static void mce_init(void *dummy)
+static int mce_cap_init(void)
 {
 	u64 cap;
-	int i;
+	unsigned b;
 
 	rdmsrl(MSR_IA32_MCG_CAP, cap);
-	banks = cap & 0xff;
-	if (banks > MCE_EXTENDED_BANK) {
-		banks = MCE_EXTENDED_BANK;
-		printk(KERN_INFO "MCE: warning: using only %d banks\n",
-		       MCE_EXTENDED_BANK);
+	b = cap & 0xff;
+	if (b > MAX_NR_BANKS) {
+		printk(KERN_WARNING
+		       "MCE: Using only %u machine check banks out of %u\n",
+			MAX_NR_BANKS, b);
+		b = MAX_NR_BANKS;
 	}
+
+	/* Don't support asymmetric configurations today */
+	WARN_ON(banks != 0 && b != banks);
+	banks = b;
+	if (!bank) {
+		bank = kmalloc(banks * sizeof(u64), GFP_KERNEL);
+		if (!bank)
+			return -ENOMEM;
+		memset(bank, 0xff, banks * sizeof(u64));
+	}
+
 	/* Use accurate RIP reporting if available. */
 	if ((cap & (1<<9)) && ((cap >> 16) & 0xff) >= 9)
 		rip_msr = MSR_IA32_MCG_EIP;
 
-	/* Log the machine checks left over from the previous reset.
-	   This also clears all registers */
-	do_machine_check(NULL, mce_bootlog ? -1 : -2);
+	return 0;
+}
+
+static void mce_init(void *dummy)
+{
+	u64 cap;
+	int i;
+	mce_banks_t all_banks;
+
+	/*
+	 * Log the machine checks left over from the previous reset.
+	 */
+	bitmap_fill(all_banks, MAX_NR_BANKS);
+	machine_check_poll(MCP_UC, &all_banks);
 
 	set_in_cr4(X86_CR4_MCE);
 
+	rdmsrl(MSR_IA32_MCG_CAP, cap);
 	if (cap & MCG_CTL_P)
 		wrmsr(MSR_IA32_MCG_CTL, 0xffffffff, 0xffffffff);
 
 	for (i = 0; i < banks; i++) {
-		if (i < NR_SYSFS_BANKS)
-			wrmsrl(MSR_IA32_MC0_CTL+4*i, bank[i]);
-		else
-			wrmsrl(MSR_IA32_MC0_CTL+4*i, ~0UL);
-
+		wrmsrl(MSR_IA32_MC0_CTL+4*i, bank[i]);
 		wrmsrl(MSR_IA32_MC0_STATUS+4*i, 0);
 	}
 }
 
 /* Add per CPU specific workarounds here */
-static void __cpuinit mce_cpu_quirks(struct cpuinfo_x86 *c)
+static void mce_cpu_quirks(struct cpuinfo_x86 *c)
 {
 	/* This should be disabled by the BIOS, but isn't always */
 	if (c->x86_vendor == X86_VENDOR_AMD) {
-		if(c->x86 == 15)
+		if (c->x86 == 15 && banks > 4)
 			/* disable GART TBL walk error reporting, which trips off
 			   incorrectly with the IOMMU & 3ware & Cerberus. */
-			clear_bit(10, &bank[4]);
+			clear_bit(10, (unsigned long *)&bank[4]);
 		if(c->x86 <= 17 && mce_bootlog < 0)
 			/* Lots of broken BIOS around that don't clear them
 			   by default and leave crap in there. Don't log. */
@@ -490,7 +615,7 @@ static void __cpuinit mce_cpu_quirks(struct cpuinfo_x86 *c)
 
 }
 
-static void __cpuinit mce_cpu_features(struct cpuinfo_x86 *c)
+static void mce_cpu_features(struct cpuinfo_x86 *c)
 {
 	switch (c->x86_vendor) {
 	case X86_VENDOR_INTEL:
@@ -504,20 +629,38 @@ static void __cpuinit mce_cpu_features(struct cpuinfo_x86 *c)
 	}
 }
 
+static void mce_init_timer(void)
+{
+	struct timer_list *t = &__get_cpu_var(mce_timer);
+
+	/* data race harmless because everyone sets to the same value */
+	if (!next_interval)
+		next_interval = check_interval * HZ;
+	if (!next_interval)
+		return;
+	setup_timer(t, mcheck_timer, smp_processor_id());
+	t->expires = round_jiffies(jiffies + next_interval);
+	add_timer(t);
+}
+
 /*
  * Called for each booted CPU to set up machine checks.
  * Must be called with preempt off.
  */
 void __cpuinit mcheck_init(struct cpuinfo_x86 *c)
 {
-	mce_cpu_quirks(c);
+	if (!mce_available(c))
+		return;
 
-	if (mce_dont_init ||
-	    !mce_available(c))
+	if (mce_cap_init() < 0) {
+		mce_dont_init = 1;
 		return;
+	}
+	mce_cpu_quirks(c);
 
 	mce_init(NULL);
 	mce_cpu_features(c);
+	mce_init_timer();
 }
 
 /*
@@ -573,7 +716,7 @@ static ssize_t mce_read(struct file *filp, char __user *ubuf, size_t usize,
 {
 	unsigned long *cpu_tsc;
 	static DEFINE_MUTEX(mce_read_mutex);
-	unsigned next;
+	unsigned prev, next;
 	char __user *buf = ubuf;
 	int i, err;
 
@@ -592,25 +735,32 @@ static ssize_t mce_read(struct file *filp, char __user *ubuf, size_t usize,
 	}
 
 	err = 0;
-	for (i = 0; i < next; i++) {
-		unsigned long start = jiffies;
-
-		while (!mcelog.entry[i].finished) {
-			if (time_after_eq(jiffies, start + 2)) {
-				memset(mcelog.entry + i,0, sizeof(struct mce));
-				goto timeout;
+	prev = 0;
+	do {
+		for (i = prev; i < next; i++) {
+			unsigned long start = jiffies;
+
+			while (!mcelog.entry[i].finished) {
+				if (time_after_eq(jiffies, start + 2)) {
+					memset(mcelog.entry + i, 0,
+					       sizeof(struct mce));
+					goto timeout;
+				}
+				cpu_relax();
 			}
-			cpu_relax();
+			smp_rmb();
+			err |= copy_to_user(buf, mcelog.entry + i,
+					    sizeof(struct mce));
+			buf += sizeof(struct mce);
+timeout:
+			;
 		}
-		smp_rmb();
-		err |= copy_to_user(buf, mcelog.entry + i, sizeof(struct mce));
-		buf += sizeof(struct mce);
- timeout:
-		;
-	}
 
-	memset(mcelog.entry, 0, next * sizeof(struct mce));
-	mcelog.next = 0;
+		memset(mcelog.entry + prev, 0,
+		       (next - prev) * sizeof(struct mce));
+		prev = next;
+		next = cmpxchg(&mcelog.next, prev, 0);
+	} while (next != prev);
 
 	synchronize_sched();
 
@@ -680,20 +830,6 @@ static struct miscdevice mce_log_device = {
 	&mce_chrdev_ops,
 };
 
-static unsigned long old_cr4 __initdata;
-
-void __init stop_mce(void)
-{
-	old_cr4 = read_cr4();
-	clear_in_cr4(X86_CR4_MCE);
-}
-
-void __init restart_mce(void)
-{
-	if (old_cr4 & X86_CR4_MCE)
-		set_in_cr4(X86_CR4_MCE);
-}
-
 /*
  * Old style boot options parsing. Only for compatibility.
  */
@@ -703,8 +839,7 @@ static int __init mcheck_disable(char *str)
 	return 1;
 }
 
-/* mce=off disables machine check. Note you can re-enable it later
-   using sysfs.
+/* mce=off disables machine check.
    mce=TOLERANCELEVEL (number, see above)
    mce=bootlog Log MCEs from before booting. Disabled by default on AMD.
    mce=nobootlog Don't log MCEs from before booting. */
@@ -728,29 +863,57 @@ __setup("mce=", mcheck_enable);
  * Sysfs support
  */
 
+/*
+ * Disable machine checks on suspend and shutdown. We can't really handle
+ * them later.
+ */
+static int mce_disable(void)
+{
+	int i;
+
+	for (i = 0; i < banks; i++)
+		wrmsrl(MSR_IA32_MC0_CTL + i*4, 0);
+	return 0;
+}
+
+static int mce_suspend(struct sys_device *dev, pm_message_t state)
+{
+	return mce_disable();
+}
+
+static int mce_shutdown(struct sys_device *dev)
+{
+	return mce_disable();
+}
+
 /* On resume clear all MCE state. Don't want to see leftovers from the BIOS.
    Only one CPU is active at this time, the others get readded later using
    CPU hotplug. */
 static int mce_resume(struct sys_device *dev)
 {
 	mce_init(NULL);
+	mce_cpu_features(&current_cpu_data);
 	return 0;
 }
 
+static void mce_cpu_restart(void *data)
+{
+	del_timer_sync(&__get_cpu_var(mce_timer));
+	if (mce_available(&current_cpu_data))
+		mce_init(NULL);
+	mce_init_timer();
+}
+
 /* Reinit MCEs after user configuration changes */
 static void mce_restart(void)
 {
-	if (next_interval)
-		cancel_delayed_work(&mcheck_work);
-	/* Timer race is harmless here */
-	on_each_cpu(mce_init, NULL, 1);
 	next_interval = check_interval * HZ;
-	if (next_interval)
-		schedule_delayed_work(&mcheck_work,
-				      round_jiffies_relative(next_interval));
+	on_each_cpu(mce_cpu_restart, NULL, 1);
 }
 
 static struct sysdev_class mce_sysclass = {
+	.suspend = mce_suspend,
+	.shutdown = mce_shutdown,
 	.resume = mce_resume,
 	.name = "machinecheck",
 };
@@ -777,16 +940,26 @@ void (*threshold_cpu_callback)(unsigned long action, unsigned int cpu) __cpuinit
 	}								\
 	static SYSDEV_ATTR(name, 0644, show_ ## name, set_ ## name);
 
-/*
- * TBD should generate these dynamically based on number of available banks.
- * Have only 6 contol banks in /sysfs until then.
- */
-ACCESSOR(bank0ctl,bank[0],mce_restart())
-ACCESSOR(bank1ctl,bank[1],mce_restart())
-ACCESSOR(bank2ctl,bank[2],mce_restart())
-ACCESSOR(bank3ctl,bank[3],mce_restart())
-ACCESSOR(bank4ctl,bank[4],mce_restart())
-ACCESSOR(bank5ctl,bank[5],mce_restart())
+static struct sysdev_attribute *bank_attrs;
+
+static ssize_t show_bank(struct sys_device *s, struct sysdev_attribute *attr,
+			 char *buf)
+{
+	u64 b = bank[attr - bank_attrs];
+	return sprintf(buf, "%llx\n", b);
+}
+
+static ssize_t set_bank(struct sys_device *s, struct sysdev_attribute *attr,
+			const char *buf, size_t siz)
+{
+	char *end;
+	u64 new = simple_strtoull(buf, &end, 0);
+	if (end == buf)
+		return -EINVAL;
+	bank[attr - bank_attrs] = new;
+	mce_restart();
+	return end-buf;
+}
 
 static ssize_t show_trigger(struct sys_device *s, struct sysdev_attribute *attr,
 				char *buf)
@@ -813,8 +986,6 @@ static SYSDEV_ATTR(trigger, 0644, show_trigger, set_trigger);
 static SYSDEV_INT_ATTR(tolerant, 0644, tolerant);
 ACCESSOR(check_interval,check_interval,mce_restart())
 static struct sysdev_attribute *mce_attributes[] = {
-	&attr_bank0ctl, &attr_bank1ctl, &attr_bank2ctl,
-	&attr_bank3ctl, &attr_bank4ctl, &attr_bank5ctl,
 	&attr_tolerant.attr, &attr_check_interval, &attr_trigger,
 	NULL
 };
@@ -844,11 +1015,22 @@ static __cpuinit int mce_create_device(unsigned int cpu)
 		if (err)
 			goto error;
 	}
+	for (i = 0; i < banks; i++) {
+		err = sysdev_create_file(&per_cpu(device_mce, cpu),
+					&bank_attrs[i]);
+		if (err)
+			goto error2;
+	}
 	cpu_set(cpu, mce_device_initialized);
 
 	return 0;
+error2:
+	while (--i >= 0) {
+		sysdev_remove_file(&per_cpu(device_mce, cpu),
+					&bank_attrs[i]);
+	}
 error:
-	while (i--) {
+	while (--i >= 0) {
 		sysdev_remove_file(&per_cpu(device_mce,cpu),
 				   mce_attributes[i]);
 	}
@@ -867,15 +1049,46 @@ static __cpuinit void mce_remove_device(unsigned int cpu)
 	for (i = 0; mce_attributes[i]; i++)
 		sysdev_remove_file(&per_cpu(device_mce,cpu),
 			mce_attributes[i]);
+	for (i = 0; i < banks; i++)
+		sysdev_remove_file(&per_cpu(device_mce, cpu),
+			&bank_attrs[i]);
 	sysdev_unregister(&per_cpu(device_mce,cpu));
 	cpu_clear(cpu, mce_device_initialized);
 }
 
+/* Make sure there are no machine checks on offlined CPUs. */
+static void mce_disable_cpu(void *h)
+{
+	int i;
+	unsigned long action = *(unsigned long *)h;
+
+	if (!mce_available(&current_cpu_data))
+		return;
+	if (!(action & CPU_TASKS_FROZEN))
+		cmci_clear();
+	for (i = 0; i < banks; i++)
+		wrmsrl(MSR_IA32_MC0_CTL + i*4, 0);
+}
+
+static void mce_reenable_cpu(void *h)
+{
+	int i;
+	unsigned long action = *(unsigned long *)h;
+
+	if (!mce_available(&current_cpu_data))
+		return;
+	if (!(action & CPU_TASKS_FROZEN))
+		cmci_reenable();
+	for (i = 0; i < banks; i++)
+		wrmsrl(MSR_IA32_MC0_CTL + i*4, bank[i]);
+}
+
 /* Get notified when a cpu comes on/off. Be hotplug friendly. */
 static int __cpuinit mce_cpu_callback(struct notifier_block *nfb,
 				      unsigned long action, void *hcpu)
 {
 	unsigned int cpu = (unsigned long)hcpu;
+	struct timer_list *t = &per_cpu(mce_timer, cpu);
 
 	switch (action) {
 	case CPU_ONLINE:
@@ -890,6 +1103,21 @@ static int __cpuinit mce_cpu_callback(struct notifier_block *nfb,
 			threshold_cpu_callback(action, cpu);
 		mce_remove_device(cpu);
 		break;
+	case CPU_DOWN_PREPARE:
+	case CPU_DOWN_PREPARE_FROZEN:
+		del_timer_sync(t);
+		smp_call_function_single(cpu, mce_disable_cpu, &action, 1);
+		break;
+	case CPU_DOWN_FAILED:
+	case CPU_DOWN_FAILED_FROZEN:
+		t->expires = round_jiffies(jiffies + next_interval);
+		add_timer_on(t, cpu);
+		smp_call_function_single(cpu, mce_reenable_cpu, &action, 1);
+		break;
+	case CPU_POST_DEAD:
+		/* intentionally ignoring frozen here */
+		cmci_rediscover(cpu);
+		break;
 	}
 	return NOTIFY_OK;
 }
@@ -898,6 +1126,34 @@ static struct notifier_block mce_cpu_notifier __cpuinitdata = {
 	.notifier_call = mce_cpu_callback,
 };
 
+static __init int mce_init_banks(void)
+{
+	int i;
+
+	bank_attrs = kzalloc(sizeof(struct sysdev_attribute) * banks,
+				GFP_KERNEL);
+	if (!bank_attrs)
+		return -ENOMEM;
+
+	for (i = 0; i < banks; i++) {
+		struct sysdev_attribute *a = &bank_attrs[i];
+		a->attr.name = kasprintf(GFP_KERNEL, "bank%d", i);
+		if (!a->attr.name)
+			goto nomem;
+		a->attr.mode = 0644;
+		a->show = show_bank;
+		a->store = set_bank;
+	}
+	return 0;
+
+nomem:
+	while (--i >= 0)
+		kfree(bank_attrs[i].attr.name);
+	kfree(bank_attrs);
+	bank_attrs = NULL;
+	return -ENOMEM;
+}
+
 static __init int mce_init_device(void)
 {
 	int err;
@@ -905,6 +1161,11 @@ static __init int mce_init_device(void)
 
 	if (!mce_available(&boot_cpu_data))
 		return -EIO;
+
+	err = mce_init_banks();
+	if (err)
+		return err;
+
 	err = sysdev_class_register(&mce_sysclass);
 	if (err)
 		return err;
diff --git a/arch/x86/kernel/cpu/mcheck/mce_amd_64.c b/arch/x86/kernel/cpu/mcheck/mce_amd_64.c
index 4772e91e8246..c5a32f92d07e 100644
--- a/arch/x86/kernel/cpu/mcheck/mce_amd_64.c
+++ b/arch/x86/kernel/cpu/mcheck/mce_amd_64.c
@@ -79,6 +79,8 @@ static unsigned char shared_bank[NR_BANKS] = {
 
 static DEFINE_PER_CPU(unsigned char, bank_map);	/* see which banks are on */
 
+static void amd_threshold_interrupt(void);
+
 /*
  * CPU Initialization
  */
@@ -121,7 +123,7 @@ static long threshold_restart_bank(void *_tr)
 }
 
 /* cpu init entry point, called from mce.c with preempt off */
-void __cpuinit mce_amd_feature_init(struct cpuinfo_x86 *c)
+void mce_amd_feature_init(struct cpuinfo_x86 *c)
 {
 	unsigned int bank, block;
 	unsigned int cpu = smp_processor_id();
@@ -174,6 +176,8 @@ void __cpuinit mce_amd_feature_init(struct cpuinfo_x86 *c)
 			tr.reset = 0;
 			tr.old_limit = 0;
 			threshold_restart_bank(&tr);
+
+			mce_threshold_vector = amd_threshold_interrupt;
 		}
 	}
 }
@@ -187,19 +191,13 @@ void __cpuinit mce_amd_feature_init(struct cpuinfo_x86 *c)
  * the interrupt goes off when error_count reaches threshold_limit.
  * the handler will simply log mcelog w/ software defined bank number.
  */
-asmlinkage void mce_threshold_interrupt(void)
+static void amd_threshold_interrupt(void)
 {
 	unsigned int bank, block;
 	struct mce m;
 	u32 low = 0, high = 0, address = 0;
 
-	ack_APIC_irq();
-	exit_idle();
-	irq_enter();
-
-	memset(&m, 0, sizeof(m));
-	rdtscll(m.tsc);
-	m.cpu = smp_processor_id();
+	mce_setup(&m);
 
 	/* assume first bank caused it */
 	for (bank = 0; bank < NR_BANKS; ++bank) {
@@ -233,7 +231,8 @@ asmlinkage void mce_threshold_interrupt(void)
 
 			/* Log the machine check that caused the threshold
 			   event. */
-			do_machine_check(NULL, 0);
+			machine_check_poll(MCP_TIMESTAMP,
+					&__get_cpu_var(mce_poll_banks));
 
 			if (high & MASK_OVERFLOW_HI) {
 				rdmsrl(address, m.misc);
@@ -243,13 +242,10 @@ asmlinkage void mce_threshold_interrupt(void)
 				       + bank * NR_BLOCKS
 				       + block;
 				mce_log(&m);
-				goto out;
+				return;
 			}
 		}
 	}
-out:
-	inc_irq_stat(irq_threshold_count);
-	irq_exit();
 }
 
 /*
diff --git a/arch/x86/kernel/cpu/mcheck/mce_intel_64.c b/arch/x86/kernel/cpu/mcheck/mce_intel_64.c
index 4b48f251fd39..aaa7d9730938 100644
--- a/arch/x86/kernel/cpu/mcheck/mce_intel_64.c
+++ b/arch/x86/kernel/cpu/mcheck/mce_intel_64.c
@@ -1,17 +1,21 @@
 /*
  * Intel specific MCE features.
  * Copyright 2004 Zwane Mwaikambo <zwane@linuxpower.ca>
+ * Copyright (C) 2008, 2009 Intel Corporation
+ * Author: Andi Kleen
  */
 
 #include <linux/init.h>
 #include <linux/interrupt.h>
 #include <linux/percpu.h>
 #include <asm/processor.h>
+#include <asm/apic.h>
 #include <asm/msr.h>
 #include <asm/mce.h>
 #include <asm/hw_irq.h>
 #include <asm/idle.h>
 #include <asm/therm_throt.h>
+#include <asm/apic.h>
 
 asmlinkage void smp_thermal_interrupt(void)
 {
@@ -24,13 +28,13 @@ asmlinkage void smp_thermal_interrupt(void)
 
 	rdmsrl(MSR_IA32_THERM_STATUS, msr_val);
 	if (therm_throt_process(msr_val & 1))
-		mce_log_therm_throt_event(smp_processor_id(), msr_val);
+		mce_log_therm_throt_event(msr_val);
 
 	inc_irq_stat(irq_thermal_count);
 	irq_exit();
 }
 
-static void __cpuinit intel_init_thermal(struct cpuinfo_x86 *c)
+static void intel_init_thermal(struct cpuinfo_x86 *c)
 {
 	u32 l, h;
 	int tm2 = 0;
@@ -48,13 +52,13 @@ static void __cpuinit intel_init_thermal(struct cpuinfo_x86 *c)
 	 */
 	rdmsr(MSR_IA32_MISC_ENABLE, l, h);
 	h = apic_read(APIC_LVTTHMR);
-	if ((l & (1 << 3)) && (h & APIC_DM_SMI)) {
+	if ((l & MSR_IA32_MISC_ENABLE_TM1) && (h & APIC_DM_SMI)) {
 		printk(KERN_DEBUG
 		       "CPU%d: Thermal monitoring handled by SMI\n", cpu);
 		return;
 	}
 
-	if (cpu_has(c, X86_FEATURE_TM2) && (l & (1 << 13)))
+	if (cpu_has(c, X86_FEATURE_TM2) && (l & MSR_IA32_MISC_ENABLE_TM2))
 		tm2 = 1;
 
 	if (h & APIC_VECTOR_MASK) {
@@ -72,7 +76,7 @@ static void __cpuinit intel_init_thermal(struct cpuinfo_x86 *c)
 	wrmsr(MSR_IA32_THERM_INTERRUPT, l | 0x03, h);
 
 	rdmsr(MSR_IA32_MISC_ENABLE, l, h);
-	wrmsr(MSR_IA32_MISC_ENABLE, l | (1 << 3), h);
+	wrmsr(MSR_IA32_MISC_ENABLE, l | MSR_IA32_MISC_ENABLE_TM1, h);
 
 	l = apic_read(APIC_LVTTHMR);
 	apic_write(APIC_LVTTHMR, l & ~APIC_LVT_MASKED);
@@ -84,7 +88,209 @@ static void __cpuinit intel_init_thermal(struct cpuinfo_x86 *c)
 	return;
 }
 
-void __cpuinit mce_intel_feature_init(struct cpuinfo_x86 *c)
+/*
+ * Support for Intel Correct Machine Check Interrupts. This allows
+ * the CPU to raise an interrupt when a corrected machine check happened.
+ * Normally we pick those up using a regular polling timer.
+ * Also supports reliable discovery of shared banks.
+ */
+
+static DEFINE_PER_CPU(mce_banks_t, mce_banks_owned);
+
+/*
+ * cmci_discover_lock protects against parallel discovery attempts
+ * which could race against each other.
+ */
+static DEFINE_SPINLOCK(cmci_discover_lock);
+
+#define CMCI_THRESHOLD 1
+
+static int cmci_supported(int *banks)
+{
+	u64 cap;
+
+	/*
+	 * Vendor check is not strictly needed, but the initial
+	 * initialization is vendor keyed and this
+	 * makes sure none of the backdoors are entered otherwise.
+	 */
+	if (boot_cpu_data.x86_vendor != X86_VENDOR_INTEL)
+		return 0;
+	if (!cpu_has_apic || lapic_get_maxlvt() < 6)
+		return 0;
+	rdmsrl(MSR_IA32_MCG_CAP, cap);
+	*banks = min_t(unsigned, MAX_NR_BANKS, cap & 0xff);
+	return !!(cap & MCG_CMCI_P);
+}
+
+/*
+ * The interrupt handler. This is called on every event.
+ * Just call the poller directly to log any events.
+ * This could in theory increase the threshold under high load,
+ * but doesn't for now.
+ */
+static void intel_threshold_interrupt(void)
+{
+	machine_check_poll(MCP_TIMESTAMP, &__get_cpu_var(mce_banks_owned));
+	mce_notify_user();
+}
+
+static void print_update(char *type, int *hdr, int num)
+{
+	if (*hdr == 0)
+		printk(KERN_INFO "CPU %d MCA banks", smp_processor_id());
+	*hdr = 1;
+	printk(KERN_CONT " %s:%d", type, num);
+}
+
+/*
+ * Enable CMCI (Corrected Machine Check Interrupt) for available MCE banks
+ * on this CPU. Use the algorithm recommended in the SDM to discover shared
+ * banks.
+ */
+static void cmci_discover(int banks, int boot)
+{
+	unsigned long *owned = (void *)&__get_cpu_var(mce_banks_owned);
+	int hdr = 0;
+	int i;
+
+	spin_lock(&cmci_discover_lock);
+	for (i = 0; i < banks; i++) {
+		u64 val;
+
+		if (test_bit(i, owned))
+			continue;
+
+		rdmsrl(MSR_IA32_MC0_CTL2 + i, val);
+
+		/* Already owned by someone else? */
+		if (val & CMCI_EN) {
+			if (test_and_clear_bit(i, owned) || boot)
+				print_update("SHD", &hdr, i);
+			__clear_bit(i, __get_cpu_var(mce_poll_banks));
+			continue;
+		}
+
+		val |= CMCI_EN | CMCI_THRESHOLD;
+		wrmsrl(MSR_IA32_MC0_CTL2 + i, val);
+		rdmsrl(MSR_IA32_MC0_CTL2 + i, val);
+
+		/* Did the enable bit stick? -- the bank supports CMCI */
+		if (val & CMCI_EN) {
+			if (!test_and_set_bit(i, owned) || boot)
+				print_update("CMCI", &hdr, i);
+			__clear_bit(i, __get_cpu_var(mce_poll_banks));
+		} else {
+			WARN_ON(!test_bit(i, __get_cpu_var(mce_poll_banks)));
+		}
+	}
+	spin_unlock(&cmci_discover_lock);
+	if (hdr)
+		printk(KERN_CONT "\n");
+}
+
+/*
+ * Just in case we missed an event during initialization check
+ * all the CMCI owned banks.
+ */
+void cmci_recheck(void)
+{
+	unsigned long flags;
+	int banks;
+
+	if (!mce_available(&current_cpu_data) || !cmci_supported(&banks))
+		return;
+	local_irq_save(flags);
+	machine_check_poll(MCP_TIMESTAMP, &__get_cpu_var(mce_banks_owned));
+	local_irq_restore(flags);
+}
+
+/*
+ * Disable CMCI on this CPU for all banks it owns when it goes down.
+ * This allows other CPUs to claim the banks on rediscovery.
+ */
+void cmci_clear(void)
+{
+	int i;
+	int banks;
+	u64 val;
+
+	if (!cmci_supported(&banks))
+		return;
+	spin_lock(&cmci_discover_lock);
+	for (i = 0; i < banks; i++) {
+		if (!test_bit(i, __get_cpu_var(mce_banks_owned)))
+			continue;
+		/* Disable CMCI */
+		rdmsrl(MSR_IA32_MC0_CTL2 + i, val);
+		val &= ~(CMCI_EN|CMCI_THRESHOLD_MASK);
+		wrmsrl(MSR_IA32_MC0_CTL2 + i, val);
+		__clear_bit(i, __get_cpu_var(mce_banks_owned));
+	}
+	spin_unlock(&cmci_discover_lock);
+}
+
+/*
+ * After a CPU went down cycle through all the others and rediscover
+ * Must run in process context.
+ */
+void cmci_rediscover(int dying)
+{
+	int banks;
+	int cpu;
+	cpumask_var_t old;
+
+	if (!cmci_supported(&banks))
+		return;
+	if (!alloc_cpumask_var(&old, GFP_KERNEL))
+		return;
+	cpumask_copy(old, &current->cpus_allowed);
+
+	for_each_online_cpu (cpu) {
+		if (cpu == dying)
+			continue;
+		if (set_cpus_allowed_ptr(current, &cpumask_of_cpu(cpu)))
+			continue;
+		/* Recheck banks in case CPUs don't all have the same */
+		if (cmci_supported(&banks))
+			cmci_discover(banks, 0);
+	}
+
+	set_cpus_allowed_ptr(current, old);
+	free_cpumask_var(old);
+}
+
+/*
+ * Reenable CMCI on this CPU in case a CPU down failed.
+ */
+void cmci_reenable(void)
+{
+	int banks;
+	if (cmci_supported(&banks))
+		cmci_discover(banks, 0);
+}
+
+static __cpuinit void intel_init_cmci(void)
+{
+	int banks;
+
+	if (!cmci_supported(&banks))
+		return;
+
+	mce_threshold_vector = intel_threshold_interrupt;
+	cmci_discover(banks, 1);
+	/*
+	 * For CPU #0 this runs with still disabled APIC, but that's
+	 * ok because only the vector is set up. We still do another
+	 * check for the banks later for CPU #0 just to make sure
+	 * to not miss any events.
+	 */
+	apic_write(APIC_LVTCMCI, THRESHOLD_APIC_VECTOR|APIC_DM_FIXED);
+	cmci_recheck();
+}
+
+void mce_intel_feature_init(struct cpuinfo_x86 *c)
 {
 	intel_init_thermal(c);
+	intel_init_cmci();
 }
diff --git a/arch/x86/kernel/cpu/mcheck/p4.c b/arch/x86/kernel/cpu/mcheck/p4.c
index 9b60fce09f75..f53bdcbaf382 100644
--- a/arch/x86/kernel/cpu/mcheck/p4.c
+++ b/arch/x86/kernel/cpu/mcheck/p4.c
@@ -85,7 +85,7 @@ static void intel_init_thermal(struct cpuinfo_x86 *c)
 	 */
 	rdmsr(MSR_IA32_MISC_ENABLE, l, h);
 	h = apic_read(APIC_LVTTHMR);
-	if ((l & (1<<3)) && (h & APIC_DM_SMI)) {
+	if ((l & MSR_IA32_MISC_ENABLE_TM1) && (h & APIC_DM_SMI)) {
 		printk(KERN_DEBUG "CPU%d: Thermal monitoring handled by SMI\n",
 				cpu);
 		return; /* -EBUSY */
@@ -111,7 +111,7 @@ static void intel_init_thermal(struct cpuinfo_x86 *c)
 	vendor_thermal_interrupt = intel_thermal_interrupt;
 
 	rdmsr(MSR_IA32_MISC_ENABLE, l, h);
-	wrmsr(MSR_IA32_MISC_ENABLE, l | (1<<3), h);
+	wrmsr(MSR_IA32_MISC_ENABLE, l | MSR_IA32_MISC_ENABLE_TM1, h);
 
 	l = apic_read(APIC_LVTTHMR);
 	apic_write(APIC_LVTTHMR, l & ~APIC_LVT_MASKED);
diff --git a/arch/x86/kernel/cpu/mcheck/threshold.c b/arch/x86/kernel/cpu/mcheck/threshold.c
new file mode 100644
index 000000000000..23ee9e730f78
--- /dev/null
+++ b/arch/x86/kernel/cpu/mcheck/threshold.c
@@ -0,0 +1,29 @@
+/*
+ * Common corrected MCE threshold handler code:
+ */
+#include <linux/interrupt.h>
+#include <linux/kernel.h>
+
+#include <asm/irq_vectors.h>
+#include <asm/apic.h>
+#include <asm/idle.h>
+#include <asm/mce.h>
+
+static void default_threshold_interrupt(void)
+{
+	printk(KERN_ERR "Unexpected threshold interrupt at vector %x\n",
+			 THRESHOLD_APIC_VECTOR);
+}
+
+void (*mce_threshold_vector)(void) = default_threshold_interrupt;
+
+asmlinkage void mce_threshold_interrupt(void)
+{
+	exit_idle();
+	irq_enter();
+	inc_irq_stat(irq_threshold_count);
+	mce_threshold_vector();
+	irq_exit();
+	/* Ack only at the end to avoid potential reentry */
+	ack_APIC_irq();
+}
diff --git a/arch/x86/kernel/cpu/mtrr/generic.c b/arch/x86/kernel/cpu/mtrr/generic.c
index b59ddcc88cd8..0c0a455fe95c 100644
--- a/arch/x86/kernel/cpu/mtrr/generic.c
+++ b/arch/x86/kernel/cpu/mtrr/generic.c
@@ -33,11 +33,13 @@ u64 mtrr_tom2;
 struct mtrr_state_type mtrr_state = {};
 EXPORT_SYMBOL_GPL(mtrr_state);
 
-#undef MODULE_PARAM_PREFIX
-#define MODULE_PARAM_PREFIX "mtrr."
-
-static int mtrr_show;
-module_param_named(show, mtrr_show, bool, 0);
+static int __initdata mtrr_show;
+static int __init mtrr_debug(char *opt)
+{
+	mtrr_show = 1;
+	return 0;
+}
+early_param("mtrr.show", mtrr_debug);
 
 /*
  * Returns the effective MTRR type for the region
diff --git a/arch/x86/kernel/cpu/mtrr/main.c b/arch/x86/kernel/cpu/mtrr/main.c
index d259e5d2e054..236a401b8259 100644
--- a/arch/x86/kernel/cpu/mtrr/main.c
+++ b/arch/x86/kernel/cpu/mtrr/main.c
@@ -1594,8 +1594,7 @@ int __init mtrr_trim_uncached_memory(unsigned long end_pfn)
 
 	/* kvm/qemu doesn't have mtrr set right, don't trim them all */
 	if (!highest_pfn) {
-		WARN(!kvm_para_available(), KERN_WARNING
-				"WARNING: strange, CPU MTRRs all blank?\n");
+		printk(KERN_INFO "CPU MTRRs all blank - virtualized system.\n");
 		return 0;
 	}
 
diff --git a/arch/x86/kernel/cpu/perfctr-watchdog.c b/arch/x86/kernel/cpu/perfctr-watchdog.c
index 9abd48b22674..f6c70a164e32 100644
--- a/arch/x86/kernel/cpu/perfctr-watchdog.c
+++ b/arch/x86/kernel/cpu/perfctr-watchdog.c
@@ -19,7 +19,7 @@
 #include <linux/nmi.h>
 #include <linux/kprobes.h>
 
-#include <asm/apic.h>
+#include <asm/genapic.h>
 #include <asm/intel_arch_perfmon.h>
 
 struct nmi_watchdog_ctlblk {
diff --git a/arch/x86/kernel/cpu/proc.c b/arch/x86/kernel/cpu/proc.c
index 01b1244ef1c0..d67e0e48bc2d 100644
--- a/arch/x86/kernel/cpu/proc.c
+++ b/arch/x86/kernel/cpu/proc.c
@@ -7,11 +7,10 @@
 /*
  *	Get CPU information for use by the procfs.
  */
-#ifdef CONFIG_X86_32
 static void show_cpuinfo_core(struct seq_file *m, struct cpuinfo_x86 *c,
 			      unsigned int cpu)
 {
-#ifdef CONFIG_X86_HT
+#ifdef CONFIG_SMP
 	if (c->x86_max_cores * smp_num_siblings > 1) {
 		seq_printf(m, "physical id\t: %d\n", c->phys_proc_id);
 		seq_printf(m, "siblings\t: %d\n",
@@ -24,6 +23,7 @@ static void show_cpuinfo_core(struct seq_file *m, struct cpuinfo_x86 *c,
 #endif
 }
 
+#ifdef CONFIG_X86_32
 static void show_cpuinfo_misc(struct seq_file *m, struct cpuinfo_x86 *c)
 {
 	/*
@@ -50,22 +50,6 @@ static void show_cpuinfo_misc(struct seq_file *m, struct cpuinfo_x86 *c)
 		   c->wp_works_ok ? "yes" : "no");
 }
 #else
-static void show_cpuinfo_core(struct seq_file *m, struct cpuinfo_x86 *c,
-			      unsigned int cpu)
-{
-#ifdef CONFIG_SMP
-	if (c->x86_max_cores * smp_num_siblings > 1) {
-		seq_printf(m, "physical id\t: %d\n", c->phys_proc_id);
-		seq_printf(m, "siblings\t: %d\n",
-			   cpus_weight(per_cpu(cpu_core_map, cpu)));
-		seq_printf(m, "core id\t\t: %d\n", c->cpu_core_id);
-		seq_printf(m, "cpu cores\t: %d\n", c->booted_cores);
-		seq_printf(m, "apicid\t\t: %d\n", c->apicid);
-		seq_printf(m, "initial apicid\t: %d\n", c->initial_apicid);
-	}
-#endif
-}
-
 static void show_cpuinfo_misc(struct seq_file *m, struct cpuinfo_x86 *c)
 {
 	seq_printf(m,