From 3819cd489ec5d18a4cbd2f05acdc516473caa105 Mon Sep 17 00:00:00 2001 From: Brian Gerst Date: Fri, 23 Jan 2009 11:03:29 +0900 Subject: x86: remove include of apic.h from hardirq_64.h Impact: cleanup APIC definitions aren't needed here. Remove the include and fix up the fallout. tj: added include to mce_intel_64.c. Signed-off-by: Brian Gerst Signed-off-by: Tejun Heo --- arch/x86/kernel/cpu/mcheck/mce_intel_64.c | 1 + 1 file changed, 1 insertion(+) (limited to 'arch/x86/kernel/cpu/mcheck/mce_intel_64.c') diff --git a/arch/x86/kernel/cpu/mcheck/mce_intel_64.c b/arch/x86/kernel/cpu/mcheck/mce_intel_64.c index 4b48f251fd3..5e8c79e748a 100644 --- a/arch/x86/kernel/cpu/mcheck/mce_intel_64.c +++ b/arch/x86/kernel/cpu/mcheck/mce_intel_64.c @@ -7,6 +7,7 @@ #include #include #include +#include #include #include #include -- cgit v1.2.2 From c1eeb2de41d7015678bdd412b48a5f071b84e29a Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Mon, 16 Feb 2009 23:02:14 -0800 Subject: x86: fold apic_ops into genapic Impact: cleanup make it simpler, don't need have one extra struct. v2: fix the sgi_uv build Signed-off-by: Yinghai Lu Cc: Andrew Morton Signed-off-by: Ingo Molnar --- arch/x86/kernel/cpu/mcheck/mce_intel_64.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86/kernel/cpu/mcheck/mce_intel_64.c') diff --git a/arch/x86/kernel/cpu/mcheck/mce_intel_64.c b/arch/x86/kernel/cpu/mcheck/mce_intel_64.c index 5e8c79e748a..42f090702f0 100644 --- a/arch/x86/kernel/cpu/mcheck/mce_intel_64.c +++ b/arch/x86/kernel/cpu/mcheck/mce_intel_64.c @@ -7,7 +7,7 @@ #include #include #include -#include +#include #include #include #include -- cgit v1.2.2 From 7b6aa335ca1a845c2262ec7a595b4521bca0f79d Mon Sep 17 00:00:00 2001 From: Ingo Molnar Date: Tue, 17 Feb 2009 13:58:15 +0100 Subject: x86, apic: remove genapic.h Impact: cleanup Remove genapic.h and remove all references to it. Signed-off-by: Ingo Molnar --- arch/x86/kernel/cpu/mcheck/mce_intel_64.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86/kernel/cpu/mcheck/mce_intel_64.c') diff --git a/arch/x86/kernel/cpu/mcheck/mce_intel_64.c b/arch/x86/kernel/cpu/mcheck/mce_intel_64.c index 42f090702f0..5e8c79e748a 100644 --- a/arch/x86/kernel/cpu/mcheck/mce_intel_64.c +++ b/arch/x86/kernel/cpu/mcheck/mce_intel_64.c @@ -7,7 +7,7 @@ #include #include #include -#include +#include #include #include #include -- cgit v1.2.2 From b5f2fa4ea00a179ac1c2ff342ceeee261dd75e53 Mon Sep 17 00:00:00 2001 From: Andi Kleen Date: Thu, 12 Feb 2009 13:43:22 +0100 Subject: x86, mce: factor out duplicated struct mce setup into one function Impact: cleanup This merely factors out duplicated code to set up the initial struct mce state into a single function. Signed-off-by: Andi Kleen Signed-off-by: H. Peter Anvin --- arch/x86/kernel/cpu/mcheck/mce_intel_64.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86/kernel/cpu/mcheck/mce_intel_64.c') diff --git a/arch/x86/kernel/cpu/mcheck/mce_intel_64.c b/arch/x86/kernel/cpu/mcheck/mce_intel_64.c index 4b48f251fd3..7f7f1015ef1 100644 --- a/arch/x86/kernel/cpu/mcheck/mce_intel_64.c +++ b/arch/x86/kernel/cpu/mcheck/mce_intel_64.c @@ -24,7 +24,7 @@ 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(); -- cgit v1.2.2 From ecab22aa6dc9d42ca52de2cad0854b4c6bd85ac9 Mon Sep 17 00:00:00 2001 From: Vegard Nossum Date: Fri, 20 Feb 2009 11:56:38 +0100 Subject: x86: use symbolic constants for MSR_IA32_MISC_ENABLE bits Impact: Cleanup. No functional changes. Signed-off-by: Vegard Nossum Signed-off-by: Ingo Molnar --- arch/x86/kernel/cpu/mcheck/mce_intel_64.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'arch/x86/kernel/cpu/mcheck/mce_intel_64.c') diff --git a/arch/x86/kernel/cpu/mcheck/mce_intel_64.c b/arch/x86/kernel/cpu/mcheck/mce_intel_64.c index 5e8c79e748a..ae00938ea50 100644 --- a/arch/x86/kernel/cpu/mcheck/mce_intel_64.c +++ b/arch/x86/kernel/cpu/mcheck/mce_intel_64.c @@ -49,13 +49,13 @@ 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) { @@ -73,7 +73,7 @@ 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); -- cgit v1.2.2 From 88ccbedd9ca85d1aca6a6f99df48dce87b7c02d4 Mon Sep 17 00:00:00 2001 From: Andi Kleen Date: Thu, 12 Feb 2009 13:49:36 +0100 Subject: x86, mce, cmci: add CMCI support Impact: Major new feature Intel CMCI (Corrected Machine Check Interrupt) is a new feature on Nehalem CPUs. It allows the CPU to trigger interrupts on corrected events, which allows faster reaction to them instead of with the traditional polling timer. Also use CMCI to discover shared banks. Machine check banks can be shared by CPU threads or even cores. Using the CMCI enable bit it is possible to detect the fact that another CPU already saw a specific bank. Use this to assign shared banks only to one CPU to avoid reporting duplicated events. On CPU hot unplug bank sharing is re discovered. This is done using a thread that cycles through all the CPUs. To avoid races between the poller and CMCI we only poll for banks that are not CMCI capable and only check CMCI owned banks on a interrupt. The shared banks ownership information is currently only used for CMCI interrupts, not polled banks. The sharing discovery code follows the algorithm recommended in the IA32 SDM Vol3a 14.5.2.1 The CMCI interrupt handler just calls the machine check poller to pick up the machine check event that caused the interrupt. I decided not to implement a separate threshold event like the AMD version has, because the threshold is always one currently and adding another event didn't seem to add any value. Some code inspired by Yunhong Jiang's Xen implementation, which was in term inspired by a earlier CMCI implementation by me. Signed-off-by: Andi Kleen Signed-off-by: H. Peter Anvin --- arch/x86/kernel/cpu/mcheck/mce_intel_64.c | 205 ++++++++++++++++++++++++++++++ 1 file changed, 205 insertions(+) (limited to 'arch/x86/kernel/cpu/mcheck/mce_intel_64.c') diff --git a/arch/x86/kernel/cpu/mcheck/mce_intel_64.c b/arch/x86/kernel/cpu/mcheck/mce_intel_64.c index 1b1491a76b5..a518ec8c6f8 100644 --- a/arch/x86/kernel/cpu/mcheck/mce_intel_64.c +++ b/arch/x86/kernel/cpu/mcheck/mce_intel_64.c @@ -1,6 +1,8 @@ /* * Intel specific MCE features. * Copyright 2004 Zwane Mwaikambo + * Copyright (C) 2008, 2009 Intel Corporation + * Author: Andi Kleen */ #include @@ -12,6 +14,7 @@ #include #include #include +#include asmlinkage void smp_thermal_interrupt(void) { @@ -84,7 +87,209 @@ 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 __cpuinit 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 __cpuinit 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. + */ +__cpuinit void cmci_recheck(void) +{ + unsigned long flags; + int banks; + + if (!mce_available(¤t_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 __cpuexit 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 __cpuexit 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, ¤t->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(); } -- cgit v1.2.2 From df20e2eb3e59b8625021a1bc8b1b53a4edc6008b Mon Sep 17 00:00:00 2001 From: "H. Peter Anvin" Date: Tue, 24 Feb 2009 13:19:02 -0800 Subject: x86, mce, cmci: remove incorrect __cpuinit/__cpuexit annotations Impact: Bug fix on UP The MCE code is reinitialized from resume, so we can't use __cpuinit/__cpuexit for most of the code. Remove those annotations for anything downstream of mce_init(). Signed-off-by: H. Peter Anvin --- arch/x86/kernel/cpu/mcheck/mce_intel_64.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) (limited to 'arch/x86/kernel/cpu/mcheck/mce_intel_64.c') diff --git a/arch/x86/kernel/cpu/mcheck/mce_intel_64.c b/arch/x86/kernel/cpu/mcheck/mce_intel_64.c index a518ec8c6f8..7a2e10fcfa3 100644 --- a/arch/x86/kernel/cpu/mcheck/mce_intel_64.c +++ b/arch/x86/kernel/cpu/mcheck/mce_intel_64.c @@ -104,7 +104,7 @@ static DEFINE_SPINLOCK(cmci_discover_lock); #define CMCI_THRESHOLD 1 -static __cpuinit int cmci_supported(int *banks) +static int cmci_supported(int *banks) { u64 cap; @@ -147,7 +147,7 @@ static void print_update(char *type, int *hdr, int num) * on this CPU. Use the algorithm recommended in the SDM to discover shared * banks. */ -static __cpuinit void cmci_discover(int banks, int boot) +static void cmci_discover(int banks, int boot) { unsigned long *owned = (void *)&__get_cpu_var(mce_banks_owned); int hdr = 0; @@ -192,7 +192,7 @@ static __cpuinit void cmci_discover(int banks, int boot) * Just in case we missed an event during initialization check * all the CMCI owned banks. */ -__cpuinit void cmci_recheck(void) +void cmci_recheck(void) { unsigned long flags; int banks; @@ -208,7 +208,7 @@ __cpuinit void cmci_recheck(void) * 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 __cpuexit cmci_clear(void) +void cmci_clear(void) { int i; int banks; @@ -233,7 +233,7 @@ void __cpuexit cmci_clear(void) * After a CPU went down cycle through all the others and rediscover * Must run in process context. */ -void __cpuexit cmci_rediscover(int dying) +void cmci_rediscover(int dying) { int banks; int cpu; -- cgit v1.2.2