4 files changed, 96 insertions, 42 deletions

diff --git a/arch/x86/kernel/cpu/mcheck/mce.c b/arch/x86/kernel/cpu/mcheck/mce.c
index 4d5419b249d..9a79c8dbd8e 100644
--- a/arch/x86/kernel/cpu/mcheck/mce.c
+++ b/arch/x86/kernel/cpu/mcheck/mce.c
@@ -60,8 +60,6 @@ static DEFINE_MUTEX(mce_chrdev_read_mutex);
 
 #define SPINUNIT 100	/* 100ns */
 
-atomic_t mce_entry;
-
 DEFINE_PER_CPU(unsigned, mce_exception_count);
 
 struct mce_bank *mce_banks __read_mostly;
@@ -89,6 +87,9 @@ static DECLARE_WAIT_QUEUE_HEAD(mce_chrdev_wait);
 static DEFINE_PER_CPU(struct mce, mces_seen);
 static int			cpu_missing;
 
+/* CMCI storm detection filter */
+static DEFINE_PER_CPU(unsigned long, mce_polled_error);
+
 /*
  * MCA banks polled by the period polling timer for corrected events.
  * With Intel CMCI, this only has MCA banks which do not support CMCI (if any).
@@ -614,6 +615,7 @@ void machine_check_poll(enum mcp_flags flags, mce_banks_t *b)
 		if (!(m.status & MCI_STATUS_VAL))
 			continue;
 
+		this_cpu_write(mce_polled_error, 1);
 		/*
 		 * Uncorrected or signalled events are handled by the exception
 		 * handler when it is enabled, so don't process those here.
@@ -700,8 +702,7 @@ static int mce_timed_out(u64 *t)
 	if (!mca_cfg.monarch_timeout)
 		goto out;
 	if ((s64)*t < SPINUNIT) {
-		/* CHECKME: Make panic default for 1 too? */
-		if (mca_cfg.tolerant < 1)
+		if (mca_cfg.tolerant <= 1)
 			mce_panic("Timeout synchronizing machine check over CPUs",
 				  NULL, NULL);
 		cpu_missing = 1;
@@ -1037,8 +1038,6 @@ void do_machine_check(struct pt_regs *regs, long error_code)
 	DECLARE_BITMAP(valid_banks, MAX_NR_BANKS);
 	char *msg = "Unknown";
 
-	atomic_inc(&mce_entry);
-
 	this_cpu_inc(mce_exception_count);
 
 	if (!cfg->banks)
@@ -1168,7 +1167,6 @@ void do_machine_check(struct pt_regs *regs, long error_code)
 		mce_report_event(regs);
 	mce_wrmsrl(MSR_IA32_MCG_STATUS, 0);
 out:
-	atomic_dec(&mce_entry);
 	sync_core();
 }
 EXPORT_SYMBOL_GPL(do_machine_check);
@@ -1278,10 +1276,18 @@ static unsigned long mce_adjust_timer_default(unsigned long interval)
 static unsigned long (*mce_adjust_timer)(unsigned long interval) =
 	mce_adjust_timer_default;
 
+static int cmc_error_seen(void)
+{
+	unsigned long *v = &__get_cpu_var(mce_polled_error);
+
+	return test_and_clear_bit(0, v);
+}
+
 static void mce_timer_fn(unsigned long data)
 {
 	struct timer_list *t = &__get_cpu_var(mce_timer);
 	unsigned long iv;
+	int notify;
 
 	WARN_ON(smp_processor_id() != data);
 
@@ -1296,7 +1302,9 @@ static void mce_timer_fn(unsigned long data)
 	 * polling interval, otherwise increase the polling interval.
 	 */
 	iv = __this_cpu_read(mce_next_interval);
-	if (mce_notify_irq()) {
+	notify = mce_notify_irq();
+	notify |= cmc_error_seen();
+	if (notify) {
 		iv = max(iv / 2, (unsigned long) HZ/100);
 	} else {
 		iv = min(iv * 2, round_jiffies_relative(check_interval * HZ));
@@ -2423,28 +2431,67 @@ static __init int mcheck_init_device(void)
 	int err;
 	int i = 0;
 
-	if (!mce_available(&boot_cpu_data))
-		return -EIO;
+	if (!mce_available(&boot_cpu_data)) {
+		err = -EIO;
+		goto err_out;
+	}
 
-	zalloc_cpumask_var(&mce_device_initialized, GFP_KERNEL);
+	if (!zalloc_cpumask_var(&mce_device_initialized, GFP_KERNEL)) {
+		err = -ENOMEM;
+		goto err_out;
+	}
 
 	mce_init_banks();
 
 	err = subsys_system_register(&mce_subsys, NULL);
 	if (err)
-		return err;
+		goto err_out_mem;
 
+	cpu_notifier_register_begin();
 	for_each_online_cpu(i) {
 		err = mce_device_create(i);
-		if (err)
-			return err;
+		if (err) {
+			/*
+			 * Register notifier anyway (and do not unreg it) so
+			 * that we don't leave undeleted timers, see notifier
+			 * callback above.
+			 */
+			__register_hotcpu_notifier(&mce_cpu_notifier);
+			cpu_notifier_register_done();
+			goto err_device_create;
+		}
 	}
 
+	__register_hotcpu_notifier(&mce_cpu_notifier);
+	cpu_notifier_register_done();
+
 	register_syscore_ops(&mce_syscore_ops);
-	register_hotcpu_notifier(&mce_cpu_notifier);
 
 	/* register character device /dev/mcelog */
-	misc_register(&mce_chrdev_device);
+	err = misc_register(&mce_chrdev_device);
+	if (err)
+		goto err_register;
+
+	return 0;
+
+err_register:
+	unregister_syscore_ops(&mce_syscore_ops);
+
+err_device_create:
+	/*
+	 * We didn't keep track of which devices were created above, but
+	 * even if we had, the set of online cpus might have changed.
+	 * Play safe and remove for every possible cpu, since
+	 * mce_device_remove() will do the right thing.
+	 */
+	for_each_possible_cpu(i)
+		mce_device_remove(i);
+
+err_out_mem:
+	free_cpumask_var(mce_device_initialized);
+
+err_out:
+	pr_err("Unable to init device /dev/mcelog (rc: %d)\n", err);
 
 	return err;
 }
diff --git a/arch/x86/kernel/cpu/mcheck/mce_intel.c b/arch/x86/kernel/cpu/mcheck/mce_intel.c
index fb6156fee6f..9a316b21df8 100644
--- a/arch/x86/kernel/cpu/mcheck/mce_intel.c
+++ b/arch/x86/kernel/cpu/mcheck/mce_intel.c
@@ -9,6 +9,7 @@
 #include <linux/interrupt.h>
 #include <linux/percpu.h>
 #include <linux/sched.h>
+#include <linux/cpumask.h>
 #include <asm/apic.h>
 #include <asm/processor.h>
 #include <asm/msr.h>
@@ -41,7 +42,7 @@ 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_RAW_SPINLOCK(cmci_discover_lock);
+static DEFINE_SPINLOCK(cmci_discover_lock);
 
 #define CMCI_THRESHOLD		1
 #define CMCI_POLL_INTERVAL	(30 * HZ)
@@ -137,6 +138,22 @@ unsigned long mce_intel_adjust_timer(unsigned long interval)
 	}
 }
 
+static void cmci_storm_disable_banks(void)
+{
+	unsigned long flags, *owned;
+	int bank;
+	u64 val;
+
+	spin_lock_irqsave(&cmci_discover_lock, flags);
+	owned = __get_cpu_var(mce_banks_owned);
+	for_each_set_bit(bank, owned, MAX_NR_BANKS) {
+		rdmsrl(MSR_IA32_MCx_CTL2(bank), val);
+		val &= ~MCI_CTL2_CMCI_EN;
+		wrmsrl(MSR_IA32_MCx_CTL2(bank), val);
+	}
+	spin_unlock_irqrestore(&cmci_discover_lock, flags);
+}
+
 static bool cmci_storm_detect(void)
 {
 	unsigned int cnt = __this_cpu_read(cmci_storm_cnt);
@@ -158,7 +175,7 @@ static bool cmci_storm_detect(void)
 	if (cnt <= CMCI_STORM_THRESHOLD)
 		return false;
 
-	cmci_clear();
+	cmci_storm_disable_banks();
 	__this_cpu_write(cmci_storm_state, CMCI_STORM_ACTIVE);
 	r = atomic_add_return(1, &cmci_storm_on_cpus);
 	mce_timer_kick(CMCI_POLL_INTERVAL);
@@ -194,7 +211,7 @@ static void cmci_discover(int banks)
 	int i;
 	int bios_wrong_thresh = 0;
 
-	raw_spin_lock_irqsave(&cmci_discover_lock, flags);
+	spin_lock_irqsave(&cmci_discover_lock, flags);
 	for (i = 0; i < banks; i++) {
 		u64 val;
 		int bios_zero_thresh = 0;
@@ -249,7 +266,7 @@ static void cmci_discover(int banks)
 			WARN_ON(!test_bit(i, __get_cpu_var(mce_poll_banks)));
 		}
 	}
-	raw_spin_unlock_irqrestore(&cmci_discover_lock, flags);
+	spin_unlock_irqrestore(&cmci_discover_lock, flags);
 	if (mca_cfg.bios_cmci_threshold && bios_wrong_thresh) {
 		pr_info_once(
 			"bios_cmci_threshold: Some banks do not have valid thresholds set\n");
@@ -299,10 +316,10 @@ void cmci_clear(void)
 
 	if (!cmci_supported(&banks))
 		return;
-	raw_spin_lock_irqsave(&cmci_discover_lock, flags);
+	spin_lock_irqsave(&cmci_discover_lock, flags);
 	for (i = 0; i < banks; i++)
 		__cmci_disable_bank(i);
-	raw_spin_unlock_irqrestore(&cmci_discover_lock, flags);
+	spin_unlock_irqrestore(&cmci_discover_lock, flags);
 }
 
 static void cmci_rediscover_work_func(void *arg)
@@ -343,9 +360,9 @@ void cmci_disable_bank(int bank)
 	if (!cmci_supported(&banks))
 		return;
 
-	raw_spin_lock_irqsave(&cmci_discover_lock, flags);
+	spin_lock_irqsave(&cmci_discover_lock, flags);
 	__cmci_disable_bank(bank);
-	raw_spin_unlock_irqrestore(&cmci_discover_lock, flags);
+	spin_unlock_irqrestore(&cmci_discover_lock, flags);
 }
 
 static void intel_init_cmci(void)
diff --git a/arch/x86/kernel/cpu/mcheck/therm_throt.c b/arch/x86/kernel/cpu/mcheck/therm_throt.c
index 3eec7de76ef..36a1bb6d1ee 100644
--- a/arch/x86/kernel/cpu/mcheck/therm_throt.c
+++ b/arch/x86/kernel/cpu/mcheck/therm_throt.c
@@ -271,9 +271,6 @@ static void thermal_throttle_remove_dev(struct device *dev)
 	sysfs_remove_group(&dev->kobj, &thermal_attr_group);
 }
 
-/* Mutex protecting device creation against CPU hotplug: */
-static DEFINE_MUTEX(therm_cpu_lock);
-
 /* Get notified when a cpu comes on/off. Be hotplug friendly. */
 static int
 thermal_throttle_cpu_callback(struct notifier_block *nfb,
@@ -289,18 +286,14 @@ thermal_throttle_cpu_callback(struct notifier_block *nfb,
 	switch (action) {
 	case CPU_UP_PREPARE:
 	case CPU_UP_PREPARE_FROZEN:
-		mutex_lock(&therm_cpu_lock);
 		err = thermal_throttle_add_dev(dev, cpu);
-		mutex_unlock(&therm_cpu_lock);
 		WARN_ON(err);
 		break;
 	case CPU_UP_CANCELED:
 	case CPU_UP_CANCELED_FROZEN:
 	case CPU_DEAD:
 	case CPU_DEAD_FROZEN:
-		mutex_lock(&therm_cpu_lock);
 		thermal_throttle_remove_dev(dev);
-		mutex_unlock(&therm_cpu_lock);
 		break;
 	}
 	return notifier_from_errno(err);
@@ -319,19 +312,16 @@ static __init int thermal_throttle_init_device(void)
 	if (!atomic_read(&therm_throt_en))
 		return 0;
 
-	register_hotcpu_notifier(&thermal_throttle_cpu_notifier);
+	cpu_notifier_register_begin();
 
-#ifdef CONFIG_HOTPLUG_CPU
-	mutex_lock(&therm_cpu_lock);
-#endif
 	/* connect live CPUs to sysfs */
 	for_each_online_cpu(cpu) {
 		err = thermal_throttle_add_dev(get_cpu_device(cpu), cpu);
 		WARN_ON(err);
 	}
-#ifdef CONFIG_HOTPLUG_CPU
-	mutex_unlock(&therm_cpu_lock);
-#endif
+
+	__register_hotcpu_notifier(&thermal_throttle_cpu_notifier);
+	cpu_notifier_register_done();
 
 	return 0;
 }
@@ -439,14 +429,14 @@ static inline void __smp_thermal_interrupt(void)
 	smp_thermal_vector();
 }
 
-asmlinkage void smp_thermal_interrupt(struct pt_regs *regs)
+asmlinkage __visible void smp_thermal_interrupt(struct pt_regs *regs)
 {
 	entering_irq();
 	__smp_thermal_interrupt();
 	exiting_ack_irq();
 }
 
-asmlinkage void smp_trace_thermal_interrupt(struct pt_regs *regs)
+asmlinkage __visible void smp_trace_thermal_interrupt(struct pt_regs *regs)
 {
 	entering_irq();
 	trace_thermal_apic_entry(THERMAL_APIC_VECTOR);
diff --git a/arch/x86/kernel/cpu/mcheck/threshold.c b/arch/x86/kernel/cpu/mcheck/threshold.c
index fe6b1c86645..7245980186e 100644
--- a/arch/x86/kernel/cpu/mcheck/threshold.c
+++ b/arch/x86/kernel/cpu/mcheck/threshold.c
@@ -24,14 +24,14 @@ static inline void __smp_threshold_interrupt(void)
 	mce_threshold_vector();
 }
 
-asmlinkage void smp_threshold_interrupt(void)
+asmlinkage __visible void smp_threshold_interrupt(void)
 {
 	entering_irq();
 	__smp_threshold_interrupt();
 	exiting_ack_irq();
 }
 
-asmlinkage void smp_trace_threshold_interrupt(void)
+asmlinkage __visible void smp_trace_threshold_interrupt(void)
 {
 	entering_irq();
 	trace_threshold_apic_entry(THRESHOLD_APIC_VECTOR);