summaryrefslogtreecommitdiff
path: root/arch/x86/kernel/cpu
diff options
context:
space:
mode:
authorIngo Molnar <mingo@elte.hu>2009-03-06 16:44:14 +0100
committerIngo Molnar <mingo@elte.hu>2009-03-06 16:45:01 +0100
commitf0ef03985130287c6c84ebe69416cf790e6cc00e (patch)
tree3ecb04cc4d82e5fc3ae5f1747e6da172ae8cbcb7 /arch/x86/kernel/cpu
parent16097439703bcd38e9fe5608c12add6dacb825ea (diff)
parent31bbed527e7039203920c51c9fb48c27aed0820c (diff)
Merge branch 'x86/core' into tracing/textedit
Conflicts: arch/x86/Kconfig block/blktrace.c kernel/irq/handle.c Semantic conflict: kernel/trace/blktrace.c Signed-off-by: Ingo Molnar <mingo@elte.hu>
Diffstat (limited to 'arch/x86/kernel/cpu')
-rw-r--r--arch/x86/kernel/cpu/addon_cpuid_features.c54
-rw-r--r--arch/x86/kernel/cpu/amd.c2
-rw-r--r--arch/x86/kernel/cpu/common.c257
-rw-r--r--arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c2
-rw-r--r--arch/x86/kernel/cpu/cpufreq/e_powersaver.c6
-rw-r--r--arch/x86/kernel/cpu/cpufreq/speedstep-centrino.c6
-rw-r--r--arch/x86/kernel/cpu/intel.c17
-rw-r--r--arch/x86/kernel/cpu/intel_cacheinfo.c63
-rw-r--r--arch/x86/kernel/cpu/mcheck/Makefile1
-rw-r--r--arch/x86/kernel/cpu/mcheck/mce_32.c14
-rw-r--r--arch/x86/kernel/cpu/mcheck/mce_64.c530
-rw-r--r--arch/x86/kernel/cpu/mcheck/mce_amd_64.c43
-rw-r--r--arch/x86/kernel/cpu/mcheck/mce_intel_64.c214
-rw-r--r--arch/x86/kernel/cpu/mcheck/p4.c4
-rw-r--r--arch/x86/kernel/cpu/mcheck/threshold.c29
-rw-r--r--arch/x86/kernel/cpu/perfctr-watchdog.c2
-rw-r--r--arch/x86/kernel/cpu/proc.c20
17 files changed, 873 insertions, 391 deletions
diff --git a/arch/x86/kernel/cpu/addon_cpuid_features.c b/arch/x86/kernel/cpu/addon_cpuid_features.c
index 2cf23634b6d..6882a735d9c 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 7c878f6aa91..25423a5b80e 100644
--- a/arch/x86/kernel/cpu/amd.c
+++ b/arch/x86/kernel/cpu/amd.c
@@ -12,8 +12,6 @@
# include <asm/cacheflush.h>
#endif
-#include <mach_apic.h>
-
#include "cpu.h"
#ifdef CONFIG_X86_32
diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c
index 83492b1f93b..826d5c87627 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/acpi-cpufreq.c b/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c
index c5d737cdb36..5e40f54171e 100644
--- a/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c
+++ b/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c
@@ -603,7 +603,7 @@ 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 c2f930d8664..41ab3f064cb 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/speedstep-centrino.c b/arch/x86/kernel/cpu/cpufreq/speedstep-centrino.c
index f08998278a3..c9f1fdc0283 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 5fff00c70de..1a89a2b68d1 100644
--- a/arch/x86/kernel/cpu/intel.c
+++ b/arch/x86/kernel/cpu/intel.c
@@ -25,7 +25,6 @@
#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)
@@ -69,6 +68,18 @@ static void __cpuinit early_init_intel(struct cpuinfo_x86 *c)
sched_clock_stable = 1;
}
+ /*
+ * There is a known erratum on Pentium III and Core Solo
+ * and Core Duo CPUs.
+ * " Page with PAT set to WC while associated MTRR is UC
+ * may consolidate to UC "
+ * Because of this erratum, it is better to stick with
+ * setting WC in MTRR rather than using PAT on these CPUs.
+ *
+ * Enable PAT WC only on P4, Core 2 or later CPUs.
+ */
+ if (c->x86 == 6 && c->x86_model < 15)
+ clear_cpu_cap(c, X86_FEATURE_PAT);
}
#ifdef CONFIG_X86_32
@@ -141,10 +152,10 @@ 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);
}
}
diff --git a/arch/x86/kernel/cpu/intel_cacheinfo.c b/arch/x86/kernel/cpu/intel_cacheinfo.c
index da299eb85fc..7293508d8f5 100644
--- a/arch/x86/kernel/cpu/intel_cacheinfo.c
+++ b/arch/x86/kernel/cpu/intel_cacheinfo.c
@@ -147,7 +147,16 @@ struct _cpuid4_info {
union _cpuid4_leaf_ecx ecx;
unsigned long size;
unsigned long can_disable;
- cpumask_t shared_cpu_map; /* future?: only cpus/node is needed */
+ DECLARE_BITMAP(shared_cpu_map, NR_CPUS);
+};
+
+/* subset of above _cpuid4_info w/o shared_cpu_map */
+struct _cpuid4_info_regs {
+ union _cpuid4_leaf_eax eax;
+ union _cpuid4_leaf_ebx ebx;
+ union _cpuid4_leaf_ecx ecx;
+ unsigned long size;
+ unsigned long can_disable;
};
#ifdef CONFIG_PCI
@@ -278,7 +287,7 @@ amd_cpuid4(int leaf, union _cpuid4_leaf_eax *eax,
}
static void __cpuinit
-amd_check_l3_disable(int index, struct _cpuid4_info *this_leaf)
+amd_check_l3_disable(int index, struct _cpuid4_info_regs *this_leaf)
{
if (index < 3)
return;
@@ -286,7 +295,8 @@ amd_check_l3_disable(int index, struct _cpuid4_info *this_leaf)
}
static int
-__cpuinit cpuid4_cache_lookup(int index, struct _cpuid4_info *this_leaf)
+__cpuinit cpuid4_cache_lookup_regs(int index,
+ struct _cpuid4_info_regs *this_leaf)
{
union _cpuid4_leaf_eax eax;
union _cpuid4_leaf_ebx ebx;
@@ -314,6 +324,15 @@ __cpuinit cpuid4_cache_lookup(int index, struct _cpuid4_info *this_leaf)
return 0;
}
+static int
+__cpuinit cpuid4_cache_lookup(int index, struct _cpuid4_info *this_leaf)
+{
+ struct _cpuid4_info_regs *leaf_regs =
+ (struct _cpuid4_info_regs *)this_leaf;
+
+ return cpuid4_cache_lookup_regs(index, leaf_regs);
+}
+
static int __cpuinit find_num_cache_leaves(void)
{
unsigned int eax, ebx, ecx, edx;
@@ -353,11 +372,10 @@ unsigned int __cpuinit init_intel_cacheinfo(struct cpuinfo_x86 *c)
* parameters cpuid leaf to find the cache details
*/
for (i = 0; i < num_cache_leaves; i++) {
- struct _cpuid4_info this_leaf;
-
+ struct _cpuid4_info_regs this_leaf;
int retval;
- retval = cpuid4_cache_lookup(i, &this_leaf);
+ retval = cpuid4_cache_lookup_regs(i, &this_leaf);
if (retval >= 0) {
switch(this_leaf.eax.split.level) {
case 1:
@@ -506,17 +524,20 @@ static void __cpuinit cache_shared_cpu_map_setup(unsigned int cpu, int index)
num_threads_sharing = 1 + this_leaf->eax.split.num_threads_sharing;
if (num_threads_sharing == 1)
- cpu_set(cpu, this_leaf->shared_cpu_map);
+ cpumask_set_cpu(cpu, to_cpumask(this_leaf->shared_cpu_map));
else {
index_msb = get_count_order(num_threads_sharing);
for_each_online_cpu(i) {
if (cpu_data(i).apicid >> index_msb ==
c->apicid >> index_msb) {
- cpu_set(i, this_leaf->shared_cpu_map);
+ cpumask_set_cpu(i,
+ to_cpumask(this_leaf->shared_cpu_map));
if (i != cpu && per_cpu(cpuid4_info, i)) {
- sibling_leaf = CPUID4_INFO_IDX(i, index);
- cpu_set(cpu, sibling_leaf->shared_cpu_map);
+ sibling_leaf =
+ CPUID4_INFO_IDX(i, index);
+ cpumask_set_cpu(cpu, to_cpumask(
+ sibling_leaf->shared_cpu_map));
}
}
}
@@ -528,9 +549,10 @@ static void __cpuinit cache_remove_shared_cpu_map(unsigned int cpu, int index)
int sibling;
this_leaf = CPUID4_INFO_IDX(cpu, index);
- for_each_cpu_mask_nr(sibling, this_leaf->shared_cpu_map) {
+ for_each_cpu(sibling, to_cpumask(this_leaf->shared_cpu_map)) {
sibling_leaf = CPUID4_INFO_IDX(sibling, index);
- cpu_clear(cpu, sibling_leaf->shared_cpu_map);
+ cpumask_clear_cpu(cpu,
+ to_cpumask(sibling_leaf->shared_cpu_map));
}
}
#else
@@ -635,8 +657,9 @@ static ssize_t show_shared_cpu_map_func(struct _cpuid4_info *this_leaf,
int n = 0;
if (len > 1) {
- cpumask_t *mask = &this_leaf->shared_cpu_map;
+ const struct cpumask *mask;
+ mask = to_cpumask(this_leaf->shared_cpu_map);
n = type?
cpulist_scnprintf(buf, len-2, mask) :
cpumask_scnprintf(buf, len-2, mask);
@@ -699,7 +722,8 @@ static struct pci_dev *get_k8_northbridge(int node)
static ssize_t show_cache_disable(struct _cpuid4_info *this_leaf, char *buf)
{
- int node = cpu_to_node(first_cpu(this_leaf->shared_cpu_map));
+ const struct cpumask *mask = to_cpumask(this_leaf->shared_cpu_map);
+ int node = cpu_to_node(cpumask_first(mask));
struct pci_dev *dev = NULL;
ssize_t ret = 0;
int i;
@@ -733,7 +757,8 @@ static ssize_t
store_cache_disable(struct _cpuid4_info *this_leaf, const char *buf,
size_t count)
{
- int node = cpu_to_node(first_cpu(this_leaf->shared_cpu_map));
+ const struct cpumask *mask = to_cpumask(this_leaf->shared_cpu_map);
+ int node = cpu_to_node(cpumask_first(mask));
struct pci_dev *dev = NULL;
unsigned int ret, index, val;
@@ -878,7 +903,7 @@ err_out:
return -ENOMEM;
}
-static cpumask_t cache_dev_map = CPU_MASK_NONE;
+static DECLARE_BITMAP(cache_dev_map, NR_CPUS);
/* Add/Remove cache interface for CPU device */
static int __cpuinit cache_add_dev(struct sys_device * sys_dev)
@@ -918,7 +943,7 @@ static int __cpuinit cache_add_dev(struct sys_device * sys_dev)
}
kobject_uevent(&(this_object->kobj), KOBJ_ADD);
}
- cpu_set(cpu, cache_dev_map);
+ cpumask_set_cpu(cpu, to_cpumask(cache_dev_map));
kobject_uevent(per_cpu(cache_kobject, cpu), KOBJ_ADD);
return 0;
@@ -931,9 +956,9 @@ static void __cpuinit cache_remove_dev(struct sys_device * sys_dev)
if (per_cpu(cpuid4_info, cpu) == NULL)
return;
- if (!cpu_isset(cpu, cache_dev_map))
+ if (!cpumask_test_cpu(cpu, to_cpumask(cache_dev_map)))
return;
- cpu_clear(cpu, cache_dev_map);
+ cpumask_clear_cpu(cpu, to_cpumask(cache_dev_map));
for (i = 0; i < num_cache_leaves; i++)
kobject_put(&(INDEX_KOBJECT_PTR(cpu,i)->kobj));
diff --git a/arch/x86/kernel/cpu/mcheck/Makefile b/arch/x86/kernel/cpu/mcheck/Makefile
index d7d2323bbb6..b2f89829bbe 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 dfaebce3633..3552119b091 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 fe79985ce0f..bfbd5323a63 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)
@@ -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. */
@@ -504,20 +629,38 @@ static void 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_relative(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. */
@@ -728,6 +863,29 @@ __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. */
@@ -738,20 +896,24 @@ static int mce_resume(struct sys_device *dev)
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",
};
@@ -778,16 +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)
@@ -814,8 +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
};
@@ -845,11 +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]);
}
@@ -868,15 +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:
@@ -891,6 +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_relative(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;
}
@@ -899,6 +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;
@@ -906,6 +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 f2ee0ae29bd..c5a32f92d07 100644
--- a/arch/x86/kernel/cpu/mcheck/mce_amd_64.c
+++ b/arch/x86/kernel/cpu/mcheck/mce_amd_64.c
@@ -67,7 +67,7 @@ static struct threshold_block threshold_defaults = {
struct threshold_bank {
struct kobject *kobj;
struct threshold_block *blocks;
- cpumask_t cpus;
+ cpumask_var_t cpus;
};
static DEFINE_PER_CPU(struct threshold_bank *, threshold_banks[NR_BANKS]);
@@ -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
*/
@@ -174,6 +176,8 @@ void 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 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();
}
/*
@@ -481,7 +477,7 @@ static __cpuinit int threshold_create_bank(unsigned int cpu, unsigned int bank)
#ifdef CONFIG_SMP
if (cpu_data(cpu).cpu_core_id && shared_bank[bank]) { /* symlink */
- i = first_cpu(per_cpu(cpu_core_map, cpu));
+ i = cpumask_first(&per_cpu(cpu_core_map, cpu));
/* first core not up yet */
if (cpu_data(i).cpu_core_id)
@@ -501,7 +497,7 @@ static __cpuinit int threshold_create_bank(unsigned int cpu, unsigned int bank)
if (err)
goto out;
- b->cpus = per_cpu(cpu_core_map, cpu);
+ cpumask_copy(b->cpus, &per_cpu(cpu_core_map, cpu));
per_cpu(threshold_banks, cpu)[bank] = b;
goto out;
}
@@ -512,15 +508,20 @@ static __cpuinit int threshold_create_bank(unsigned int cpu, unsigned int bank)
err = -ENOMEM;
goto out;
}
+ if (!alloc_cpumask_var(&b->cpus, GFP_KERNEL)) {
+ kfree(b);
+ err = -ENOMEM;
+ goto out;
+ }
b->kobj = kobject_create_and_add(name, &per_cpu(device_mce, cpu).kobj);
if (!b->kobj)
goto out_free;
#ifndef CONFIG_SMP
- b->cpus = CPU_MASK_ALL;
+ cpumask_setall(b->cpus);
#else
- b->cpus = per_cpu(cpu_core_map, cpu);
+ cpumask_copy(b->cpus, &per_cpu(cpu_core_map, cpu));
#endif
per_cpu(threshold_banks, cpu)[bank] = b;
@@ -529,7 +530,7 @@ static __cpuinit int threshold_create_bank(unsigned int cpu, unsigned int bank)
if (err)
goto out_free;
- for_each_cpu_mask_nr(i, b->cpus) {
+ for_each_cpu(i, b->cpus) {
if (i == cpu)
continue;
@@ -545,6 +546,7 @@ static __cpuinit int threshold_create_bank(unsigned int cpu, unsigned int bank)
out_free:
per_cpu(threshold_banks, cpu)[bank] = NULL;
+ free_cpumask_var(b->cpus);
kfree(b);
out:
return err;
@@ -619,7 +621,7 @@ static void threshold_remove_bank(unsigned int cpu, int bank)
#endif
/* remove all sibling symlinks before unregistering */
- for_each_cpu_mask_nr(i, b->cpus) {
+ for_each_cpu(i, b->cpus) {
if (i == cpu)
continue;
@@ -632,6 +634,7 @@ static void threshold_remove_bank(unsigned int cpu, int bank)
free_out:
kobject_del(b->kobj);
kobject_put(b->kobj);
+ free_cpumask_var(b->cpus);
kfree(b);
per_cpu(threshold_banks, cpu)[bank] = NULL;
}
diff --git a/arch/x86/kernel/cpu/mcheck/mce_intel_64.c b/arch/x86/kernel/cpu/mcheck/mce_intel_64.c
index f44c3662436..aaa7d973093 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)
{
@@ -24,7 +28,7 @@ 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();
@@ -48,13 +52,13 @@ 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;
}
- 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 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 intel_init_thermal(struct cpuinfo_x86 *c)
return;
}
+/*
+ * 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 9b60fce09f7..f53bdcbaf38 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 00000000000..23ee9e730f7
--- /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/perfctr-watchdog.c b/arch/x86/kernel/cpu/perfctr-watchdog.c
index 9abd48b2267..f6c70a164e3 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 01b1244ef1c..d67e0e48bc2 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,