#include <linux/sched.h>
#include <linux/sysfs.h>
#include <linux/types.h>
+#include <linux/slab.h>
#include <linux/init.h>
#include <linux/kmod.h>
#include <linux/poll.h>
#include <linux/smp.h>
#include <linux/fs.h>
#include <linux/mm.h>
+#include <linux/debugfs.h>
#include <asm/processor.h>
#include <asm/hw_irq.h>
#include "mce-internal.h"
+static DEFINE_MUTEX(mce_read_mutex);
+
+#define rcu_dereference_check_mce(p) \
+ rcu_dereference_check((p), \
+ rcu_read_lock_sched_held() || \
+ lockdep_is_held(&mce_read_mutex))
+
+#define CREATE_TRACE_POINTS
+#include <trace/events/mce.h>
+
int mce_disabled __read_mostly;
#define MISC_MCELOG_MINOR 227
static DEFINE_PER_CPU(struct mce, mces_seen);
static int cpu_missing;
+/*
+ * CPU/chipset specific EDAC code can register a notifier call here to print
+ * MCE errors in a human-readable form.
+ */
+ATOMIC_NOTIFIER_HEAD(x86_mce_decoder_chain);
+EXPORT_SYMBOL_GPL(x86_mce_decoder_chain);
+
+static int default_decode_mce(struct notifier_block *nb, unsigned long val,
+ void *data)
+{
+ pr_emerg("No human readable MCE decoding support on this CPU type.\n");
+ pr_emerg("Run the message through 'mcelog --ascii' to decode.\n");
+
+ return NOTIFY_STOP;
+}
+
+static struct notifier_block mce_dec_nb = {
+ .notifier_call = default_decode_mce,
+ .priority = -1,
+};
/* MCA banks polled by the period polling timer for corrected events */
DEFINE_PER_CPU(mce_banks_t, mce_poll_banks) = {
{
unsigned next, entry;
+ /* Emit the trace record: */
+ trace_mce_record(mce);
+
mce->finished = 0;
wmb();
for (;;) {
- entry = rcu_dereference(mcelog.next);
+ entry = rcu_dereference_check_mce(mcelog.next);
for (;;) {
/*
* When the buffer fills up discard new entries.
static void print_mce(struct mce *m)
{
- printk(KERN_EMERG
- "CPU %d: Machine Check Exception: %16Lx Bank %d: %016Lx\n",
+ pr_emerg("CPU %d: Machine Check Exception: %16Lx Bank %d: %016Lx\n",
m->extcpu, m->mcgstatus, m->bank, m->status);
+
if (m->ip) {
- printk(KERN_EMERG "RIP%s %02x:<%016Lx> ",
- !(m->mcgstatus & MCG_STATUS_EIPV) ? " !INEXACT!" : "",
- m->cs, m->ip);
+ pr_emerg("RIP%s %02x:<%016Lx> ",
+ !(m->mcgstatus & MCG_STATUS_EIPV) ? " !INEXACT!" : "",
+ m->cs, m->ip);
+
if (m->cs == __KERNEL_CS)
print_symbol("{%s}", m->ip);
- printk("\n");
+ pr_cont("\n");
}
- printk(KERN_EMERG "TSC %llx ", m->tsc);
+
+ pr_emerg("TSC %llx ", m->tsc);
if (m->addr)
- printk("ADDR %llx ", m->addr);
+ pr_cont("ADDR %llx ", m->addr);
if (m->misc)
- printk("MISC %llx ", m->misc);
- printk("\n");
- printk(KERN_EMERG "PROCESSOR %u:%x TIME %llu SOCKET %u APIC %x\n",
- m->cpuvendor, m->cpuid, m->time, m->socketid,
- m->apicid);
+ pr_cont("MISC %llx ", m->misc);
+
+ pr_cont("\n");
+ pr_emerg("PROCESSOR %u:%x TIME %llu SOCKET %u APIC %x\n",
+ m->cpuvendor, m->cpuid, m->time, m->socketid, m->apicid);
+
+ /*
+ * Print out human-readable details about the MCE error,
+ * (if the CPU has an implementation for that)
+ */
+ atomic_notifier_call_chain(&x86_mce_decoder_chain, 0, m);
}
static void print_mce_head(void)
{
- printk(KERN_EMERG "\n" KERN_EMERG "HARDWARE ERROR\n");
+ pr_emerg("\nHARDWARE ERROR\n");
}
static void print_mce_tail(void)
{
- printk(KERN_EMERG "This is not a software problem!\n"
- KERN_EMERG "Run through mcelog --ascii to decode and contact your hardware vendor\n");
+ pr_emerg("This is not a software problem!\n");
}
#define PANIC_TIMEOUT 5 /* 5 seconds */
static atomic_t mce_paniced;
+static int fake_panic;
+static atomic_t mce_fake_paniced;
+
/* Panic in progress. Enable interrupts and wait for final IPI */
static void wait_for_panic(void)
{
long timeout = PANIC_TIMEOUT*USEC_PER_SEC;
+
preempt_disable();
local_irq_enable();
while (timeout-- > 0)
{
int i;
- /*
- * Make sure only one CPU runs in machine check panic
- */
- if (atomic_inc_return(&mce_paniced) > 1)
- wait_for_panic();
- barrier();
+ if (!fake_panic) {
+ /*
+ * Make sure only one CPU runs in machine check panic
+ */
+ if (atomic_inc_return(&mce_paniced) > 1)
+ wait_for_panic();
+ barrier();
- bust_spinlocks(1);
- console_verbose();
+ bust_spinlocks(1);
+ console_verbose();
+ } else {
+ /* Don't log too much for fake panic */
+ if (atomic_inc_return(&mce_fake_paniced) > 1)
+ return;
+ }
print_mce_head();
/* First print corrected ones that are still unlogged */
for (i = 0; i < MCE_LOG_LEN; i++) {
print_mce_tail();
if (exp)
printk(KERN_EMERG "Machine check: %s\n", exp);
- if (panic_timeout == 0)
- panic_timeout = mce_panic_timeout;
- panic(msg);
+ if (!fake_panic) {
+ if (panic_timeout == 0)
+ panic_timeout = mce_panic_timeout;
+ panic(msg);
+ } else
+ printk(KERN_EMERG "Fake kernel panic: %s\n", msg);
}
/* Support code for software error injection */
static int msr_to_offset(u32 msr)
{
unsigned bank = __get_cpu_var(injectm.bank);
+
if (msr == rip_msr)
return offsetof(struct mce, ip);
if (msr == MSR_IA32_MCx_STATUS(bank))
static u64 mce_rdmsrl(u32 msr)
{
u64 v;
+
if (__get_cpu_var(injectm).finished) {
int offset = msr_to_offset(msr);
+
if (offset < 0)
return 0;
return *(u64 *)((char *)&__get_cpu_var(injectm) + offset);
}
- rdmsrl(msr, v);
+
+ if (rdmsrl_safe(msr, &v)) {
+ WARN_ONCE(1, "mce: Unable to read msr %d!\n", msr);
+ /*
+ * Return zero in case the access faulted. This should
+ * not happen normally but can happen if the CPU does
+ * something weird, or if the code is buggy.
+ */
+ v = 0;
+ }
+
return v;
}
{
if (__get_cpu_var(injectm).finished) {
int offset = msr_to_offset(msr);
+
if (offset >= 0)
*(u64 *)((char *)&__get_cpu_var(injectm) + offset) = v;
return;
m->ip = mce_rdmsrl(rip_msr);
}
-#ifdef CONFIG_X86_LOCAL_APIC
+#ifdef CONFIG_X86_LOCAL_APIC
/*
* Called after interrupts have been reenabled again
* when a MCE happened during an interrupts off region
* This way we prevent any potential data corruption in a unrecoverable case
* and also makes sure always all CPU's errors are examined.
*
- * Also this detects the case of an machine check event coming from outer
+ * Also this detects the case of a machine check event coming from outer
* space (not detected by any CPUs) In this case some external agent wants
* us to shut down, so panic too.
*
* No machine check event found. Must be some external
* source or one CPU is hung. Panic.
*/
- if (!m && tolerant < 3)
+ if (global_worst <= MCE_KEEP_SEVERITY && tolerant < 3)
mce_panic("Machine check from unknown source", NULL, NULL);
/*
mce_setup(&m);
m.mcgstatus = mce_rdmsrl(MSR_IA32_MCG_STATUS);
- no_way_out = mce_no_way_out(&m, &msg);
-
final = &__get_cpu_var(mces_seen);
*final = m;
+ no_way_out = mce_no_way_out(&m, &msg);
+
barrier();
/*
*/
static int check_interval = 5 * 60; /* 5 minutes */
-static DEFINE_PER_CPU(int, next_interval); /* in jiffies */
+static DEFINE_PER_CPU(int, mce_next_interval); /* in jiffies */
static DEFINE_PER_CPU(struct timer_list, mce_timer);
-static void mcheck_timer(unsigned long data)
+static void mce_start_timer(unsigned long data)
{
struct timer_list *t = &per_cpu(mce_timer, data);
int *n;
* Alert userspace if needed. If we logged an MCE, reduce the
* polling interval, otherwise increase the polling interval.
*/
- n = &__get_cpu_var(next_interval);
+ n = &__get_cpu_var(mce_next_interval);
if (mce_notify_irq())
*n = max(*n/2, HZ/100);
else
*n = min(*n*2, (int)round_jiffies_relative(check_interval*HZ));
t->expires = jiffies + *n;
- add_timer(t);
+ add_timer_on(t, smp_processor_id());
}
static void mce_do_trigger(struct work_struct *work)
}
EXPORT_SYMBOL_GPL(mce_notify_irq);
-static int mce_banks_init(void)
+static int __cpuinit __mcheck_cpu_mce_banks_init(void)
{
int i;
return -ENOMEM;
for (i = 0; i < banks; i++) {
struct mce_bank *b = &mce_banks[i];
+
b->ctl = -1ULL;
b->init = 1;
}
/*
* Initialize Machine Checks for a CPU.
*/
-static int mce_cap_init(void)
+static int __cpuinit __mcheck_cpu_cap_init(void)
{
unsigned b;
u64 cap;
rdmsrl(MSR_IA32_MCG_CAP, cap);
b = cap & MCG_BANKCNT_MASK;
- printk(KERN_INFO "mce: CPU supports %d MCE banks\n", b);
+ if (!banks)
+ printk(KERN_INFO "mce: CPU supports %d MCE banks\n", b);
if (b > MAX_NR_BANKS) {
printk(KERN_WARNING
WARN_ON(banks != 0 && b != banks);
banks = b;
if (!mce_banks) {
- int err = mce_banks_init();
+ int err = __mcheck_cpu_mce_banks_init();
+
if (err)
return err;
}
return 0;
}
-static void mce_init(void)
+static void __mcheck_cpu_init_generic(void)
{
mce_banks_t all_banks;
u64 cap;
for (i = 0; i < banks; i++) {
struct mce_bank *b = &mce_banks[i];
+
if (!b->init)
continue;
wrmsrl(MSR_IA32_MCx_CTL(i), b->ctl);
}
/* Add per CPU specific workarounds here */
-static void mce_cpu_quirks(struct cpuinfo_x86 *c)
+static int __cpuinit __mcheck_cpu_apply_quirks(struct cpuinfo_x86 *c)
{
+ if (c->x86_vendor == X86_VENDOR_UNKNOWN) {
+ pr_info("MCE: unknown CPU type - not enabling MCE support.\n");
+ return -EOPNOTSUPP;
+ }
+
/* This should be disabled by the BIOS, but isn't always */
if (c->x86_vendor == X86_VENDOR_AMD) {
if (c->x86 == 15 && banks > 4) {
monarch_timeout < 0)
monarch_timeout = USEC_PER_SEC;
- /* There are also broken BIOSes on some Pentium M systems. */
- if (c->x86 == 6 && c->x86_model == 13 && mce_bootlog < 0)
+ /*
+ * There are also broken BIOSes on some Pentium M and
+ * earlier systems:
+ */
+ if (c->x86 == 6 && c->x86_model <= 13 && mce_bootlog < 0)
mce_bootlog = 0;
}
if (monarch_timeout < 0)
monarch_timeout = 0;
if (mce_bootlog != 0)
mce_panic_timeout = 30;
+
+ return 0;
}
-static void __cpuinit mce_ancient_init(struct cpuinfo_x86 *c)
+static void __cpuinit __mcheck_cpu_ancient_init(struct cpuinfo_x86 *c)
{
if (c->x86 != 5)
return;
}
}
-static void mce_cpu_features(struct cpuinfo_x86 *c)
+static void __mcheck_cpu_init_vendor(struct cpuinfo_x86 *c)
{
switch (c->x86_vendor) {
case X86_VENDOR_INTEL:
}
}
-static void mce_init_timer(void)
+static void __mcheck_cpu_init_timer(void)
{
struct timer_list *t = &__get_cpu_var(mce_timer);
- int *n = &__get_cpu_var(next_interval);
+ int *n = &__get_cpu_var(mce_next_interval);
+
+ setup_timer(t, mce_start_timer, smp_processor_id());
if (mce_ignore_ce)
return;
*n = check_interval * HZ;
if (!*n)
return;
- setup_timer(t, mcheck_timer, smp_processor_id());
t->expires = round_jiffies(jiffies + *n);
- add_timer(t);
+ add_timer_on(t, smp_processor_id());
}
/* Handle unconfigured int18 (should never happen) */
* Called for each booted CPU to set up machine checks.
* Must be called with preempt off:
*/
-void __cpuinit mcheck_init(struct cpuinfo_x86 *c)
+void __cpuinit mcheck_cpu_init(struct cpuinfo_x86 *c)
{
if (mce_disabled)
return;
- mce_ancient_init(c);
+ __mcheck_cpu_ancient_init(c);
if (!mce_available(c))
return;
- if (mce_cap_init() < 0) {
+ if (__mcheck_cpu_cap_init() < 0 || __mcheck_cpu_apply_quirks(c) < 0) {
mce_disabled = 1;
return;
}
- mce_cpu_quirks(c);
machine_check_vector = do_machine_check;
- mce_init();
- mce_cpu_features(c);
- mce_init_timer();
+ __mcheck_cpu_init_generic();
+ __mcheck_cpu_init_vendor(c);
+ __mcheck_cpu_init_timer();
INIT_WORK(&__get_cpu_var(mce_work), mce_process_work);
+
}
/*
rdtscll(cpu_tsc[smp_processor_id()]);
}
-static DEFINE_MUTEX(mce_read_mutex);
-
static ssize_t mce_read(struct file *filp, char __user *ubuf, size_t usize,
loff_t *off)
{
return -ENOMEM;
mutex_lock(&mce_read_mutex);
- next = rcu_dereference(mcelog.next);
+ next = rcu_dereference_check_mce(mcelog.next);
/* Only supports full reads right now */
if (*off != 0 || usize < MCE_LOG_LEN*sizeof(struct mce)) {
static unsigned int mce_poll(struct file *file, poll_table *wait)
{
poll_wait(file, &mce_wait, wait);
- if (rcu_dereference(mcelog.next))
+ if (rcu_dereference_check_mce(mcelog.next))
return POLLIN | POLLRDNORM;
return 0;
}
*/
static int __init mcheck_enable(char *str)
{
- if (*str == 0)
+ if (*str == 0) {
enable_p5_mce();
+ return 1;
+ }
if (*str == '=')
str++;
if (!strcmp(str, "off"))
}
__setup("mce", mcheck_enable);
+int __init mcheck_init(void)
+{
+ atomic_notifier_chain_register(&x86_mce_decoder_chain, &mce_dec_nb);
+
+ mcheck_intel_therm_init();
+
+ return 0;
+}
+
/*
* Sysfs support
*/
* Disable machine checks on suspend and shutdown. We can't really handle
* them later.
*/
-static int mce_disable(void)
+static int mce_disable_error_reporting(void)
{
int i;
for (i = 0; i < banks; i++) {
struct mce_bank *b = &mce_banks[i];
+
if (b->init)
wrmsrl(MSR_IA32_MCx_CTL(i), 0);
}
static int mce_suspend(struct sys_device *dev, pm_message_t state)
{
- return mce_disable();
+ return mce_disable_error_reporting();
}
static int mce_shutdown(struct sys_device *dev)
{
- return mce_disable();
+ return mce_disable_error_reporting();
}
/*
*/
static int mce_resume(struct sys_device *dev)
{
- mce_init();
- mce_cpu_features(¤t_cpu_data);
+ __mcheck_cpu_init_generic();
+ __mcheck_cpu_init_vendor(¤t_cpu_data);
return 0;
}
del_timer_sync(&__get_cpu_var(mce_timer));
if (!mce_available(¤t_cpu_data))
return;
- mce_init();
- mce_init_timer();
+ __mcheck_cpu_init_generic();
+ __mcheck_cpu_init_timer();
}
/* Reinit MCEs after user configuration changes */
cmci_reenable();
cmci_recheck();
if (all)
- mce_init_timer();
+ __mcheck_cpu_init_timer();
}
static struct sysdev_class mce_sysclass = {
const char *buf, size_t siz)
{
char *p;
- int len;
strncpy(mce_helper, buf, sizeof(mce_helper));
mce_helper[sizeof(mce_helper)-1] = 0;
- len = strlen(mce_helper);
p = strchr(mce_helper, '\n');
- if (*p)
+ if (p)
*p = 0;
- return len;
+ return strlen(mce_helper) + !!p;
}
static ssize_t set_ignore_ce(struct sys_device *s,
sysdev_remove_file(&per_cpu(mce_dev, cpu), &mce_banks[j].attr);
error:
while (--i >= 0)
- sysdev_remove_file(&per_cpu(mce_dev, cpu), &mce_banks[i].attr);
+ sysdev_remove_file(&per_cpu(mce_dev, cpu), mce_attrs[i]);
sysdev_unregister(&per_cpu(mce_dev, cpu));
}
/* Make sure there are no machine checks on offlined CPUs. */
-static void mce_disable_cpu(void *h)
+static void __cpuinit mce_disable_cpu(void *h)
{
unsigned long action = *(unsigned long *)h;
int i;
if (!mce_available(¤t_cpu_data))
return;
+
if (!(action & CPU_TASKS_FROZEN))
cmci_clear();
for (i = 0; i < banks; i++) {
struct mce_bank *b = &mce_banks[i];
+
if (b->init)
wrmsrl(MSR_IA32_MCx_CTL(i), 0);
}
}
-static void mce_reenable_cpu(void *h)
+static void __cpuinit mce_reenable_cpu(void *h)
{
unsigned long action = *(unsigned long *)h;
int i;
cmci_reenable();
for (i = 0; i < banks; i++) {
struct mce_bank *b = &mce_banks[i];
+
if (b->init)
wrmsrl(MSR_IA32_MCx_CTL(i), b->ctl);
}
break;
case CPU_DOWN_FAILED:
case CPU_DOWN_FAILED_FROZEN:
- t->expires = round_jiffies(jiffies +
- __get_cpu_var(next_interval));
- add_timer_on(t, cpu);
+ if (!mce_ignore_ce && check_interval) {
+ t->expires = round_jiffies(jiffies +
+ __get_cpu_var(mce_next_interval));
+ add_timer_on(t, cpu);
+ }
smp_call_function_single(cpu, mce_reenable_cpu, &action, 1);
break;
case CPU_POST_DEAD:
struct mce_bank *b = &mce_banks[i];
struct sysdev_attribute *a = &b->attr;
+ sysfs_attr_init(&a->attr);
a->attr.name = b->attrname;
snprintf(b->attrname, ATTR_LEN, "bank%d", i);
}
}
-static __init int mce_init_device(void)
+static __init int mcheck_init_device(void)
{
int err;
int i = 0;
return err;
}
-device_initcall(mce_init_device);
+device_initcall(mcheck_init_device);
/*
* Old style boot options parsing. Only for compatibility.
return 1;
}
__setup("nomce", mcheck_disable);
+
+#ifdef CONFIG_DEBUG_FS
+struct dentry *mce_get_debugfs_dir(void)
+{
+ static struct dentry *dmce;
+
+ if (!dmce)
+ dmce = debugfs_create_dir("mce", NULL);
+
+ return dmce;
+}
+
+static void mce_reset(void)
+{
+ cpu_missing = 0;
+ atomic_set(&mce_fake_paniced, 0);
+ atomic_set(&mce_executing, 0);
+ atomic_set(&mce_callin, 0);
+ atomic_set(&global_nwo, 0);
+}
+
+static int fake_panic_get(void *data, u64 *val)
+{
+ *val = fake_panic;
+ return 0;
+}
+
+static int fake_panic_set(void *data, u64 val)
+{
+ mce_reset();
+ fake_panic = val;
+ return 0;
+}
+
+DEFINE_SIMPLE_ATTRIBUTE(fake_panic_fops, fake_panic_get,
+ fake_panic_set, "%llu\n");
+
+static int __init mcheck_debugfs_init(void)
+{
+ struct dentry *dmce, *ffake_panic;
+
+ dmce = mce_get_debugfs_dir();
+ if (!dmce)
+ return -ENOMEM;
+ ffake_panic = debugfs_create_file("fake_panic", 0444, dmce, NULL,
+ &fake_panic_fops);
+ if (!ffake_panic)
+ return -ENOMEM;
+
+ return 0;
+}
+late_initcall(mcheck_debugfs_init);
+#endif
git://ftp.safe.ca / safe / jmp / linux-2.6 / blobdiff