aboutsummaryrefslogtreecommitdiffstats
path: root/arch/x86/kernel/cpu/mce/core.c
diff options
context:
space:
mode:
Diffstat (limited to 'arch/x86/kernel/cpu/mce/core.c')
-rw-r--r--arch/x86/kernel/cpu/mce/core.c180
1 files changed, 124 insertions, 56 deletions
diff --git a/arch/x86/kernel/cpu/mce/core.c b/arch/x86/kernel/cpu/mce/core.c
index 5112a50e6486..743370ee4983 100644
--- a/arch/x86/kernel/cpu/mce/core.c
+++ b/arch/x86/kernel/cpu/mce/core.c
@@ -1,3 +1,4 @@
+// SPDX-License-Identifier: GPL-2.0-only
/*
* Machine check handler.
*
@@ -64,7 +65,23 @@ static DEFINE_MUTEX(mce_sysfs_mutex);
DEFINE_PER_CPU(unsigned, mce_exception_count);
-struct mce_bank *mce_banks __read_mostly;
+DEFINE_PER_CPU_READ_MOSTLY(unsigned int, mce_num_banks);
+
+struct mce_bank {
+ u64 ctl; /* subevents to enable */
+ bool init; /* initialise bank? */
+};
+static DEFINE_PER_CPU_READ_MOSTLY(struct mce_bank[MAX_NR_BANKS], mce_banks_array);
+
+#define ATTR_LEN 16
+/* One object for each MCE bank, shared by all CPUs */
+struct mce_bank_dev {
+ struct device_attribute attr; /* device attribute */
+ char attrname[ATTR_LEN]; /* attribute name */
+ u8 bank; /* bank number */
+};
+static struct mce_bank_dev mce_bank_devs[MAX_NR_BANKS];
+
struct mce_vendor_flags mce_flags __read_mostly;
struct mca_config mca_cfg __read_mostly = {
@@ -674,6 +691,7 @@ DEFINE_PER_CPU(unsigned, mce_poll_count);
*/
bool machine_check_poll(enum mcp_flags flags, mce_banks_t *b)
{
+ struct mce_bank *mce_banks = this_cpu_ptr(mce_banks_array);
bool error_seen = false;
struct mce m;
int i;
@@ -685,7 +703,7 @@ bool machine_check_poll(enum mcp_flags flags, mce_banks_t *b)
if (flags & MCP_TIMESTAMP)
m.tsc = rdtsc();
- for (i = 0; i < mca_cfg.banks; i++) {
+ for (i = 0; i < this_cpu_read(mce_num_banks); i++) {
if (!mce_banks[i].ctl || !test_bit(i, *b))
continue;
@@ -787,7 +805,7 @@ static int mce_no_way_out(struct mce *m, char **msg, unsigned long *validp,
char *tmp;
int i;
- for (i = 0; i < mca_cfg.banks; i++) {
+ for (i = 0; i < this_cpu_read(mce_num_banks); i++) {
m->status = mce_rdmsrl(msr_ops.status(i));
if (!(m->status & MCI_STATUS_VAL))
continue;
@@ -1067,7 +1085,7 @@ static void mce_clear_state(unsigned long *toclear)
{
int i;
- for (i = 0; i < mca_cfg.banks; i++) {
+ for (i = 0; i < this_cpu_read(mce_num_banks); i++) {
if (test_bit(i, toclear))
mce_wrmsrl(msr_ops.status(i), 0);
}
@@ -1121,10 +1139,11 @@ static void __mc_scan_banks(struct mce *m, struct mce *final,
unsigned long *toclear, unsigned long *valid_banks,
int no_way_out, int *worst)
{
+ struct mce_bank *mce_banks = this_cpu_ptr(mce_banks_array);
struct mca_config *cfg = &mca_cfg;
int severity, i;
- for (i = 0; i < cfg->banks; i++) {
+ for (i = 0; i < this_cpu_read(mce_num_banks); i++) {
__clear_bit(i, toclear);
if (!test_bit(i, valid_banks))
continue;
@@ -1329,7 +1348,7 @@ void do_machine_check(struct pt_regs *regs, long error_code)
local_irq_enable();
if (kill_it || do_memory_failure(&m))
- force_sig(SIGBUS, current);
+ force_sig(SIGBUS);
local_irq_disable();
ist_end_non_atomic();
} else {
@@ -1462,27 +1481,29 @@ int mce_notify_irq(void)
}
EXPORT_SYMBOL_GPL(mce_notify_irq);
-static int __mcheck_cpu_mce_banks_init(void)
+static void __mcheck_cpu_mce_banks_init(void)
{
+ struct mce_bank *mce_banks = this_cpu_ptr(mce_banks_array);
+ u8 n_banks = this_cpu_read(mce_num_banks);
int i;
- mce_banks = kcalloc(MAX_NR_BANKS, sizeof(struct mce_bank), GFP_KERNEL);
- if (!mce_banks)
- return -ENOMEM;
-
- for (i = 0; i < MAX_NR_BANKS; i++) {
+ for (i = 0; i < n_banks; i++) {
struct mce_bank *b = &mce_banks[i];
+ /*
+ * Init them all, __mcheck_cpu_apply_quirks() is going to apply
+ * the required vendor quirks before
+ * __mcheck_cpu_init_clear_banks() does the final bank setup.
+ */
b->ctl = -1ULL;
b->init = 1;
}
- return 0;
}
/*
* Initialize Machine Checks for a CPU.
*/
-static int __mcheck_cpu_cap_init(void)
+static void __mcheck_cpu_cap_init(void)
{
u64 cap;
u8 b;
@@ -1490,16 +1511,16 @@ static int __mcheck_cpu_cap_init(void)
rdmsrl(MSR_IA32_MCG_CAP, cap);
b = cap & MCG_BANKCNT_MASK;
- if (WARN_ON_ONCE(b > MAX_NR_BANKS))
+
+ if (b > MAX_NR_BANKS) {
+ pr_warn("CPU%d: Using only %u machine check banks out of %u\n",
+ smp_processor_id(), MAX_NR_BANKS, b);
b = MAX_NR_BANKS;
+ }
- mca_cfg.banks = max(mca_cfg.banks, b);
+ this_cpu_write(mce_num_banks, b);
- if (!mce_banks) {
- int err = __mcheck_cpu_mce_banks_init();
- if (err)
- return err;
- }
+ __mcheck_cpu_mce_banks_init();
/* Use accurate RIP reporting if available. */
if ((cap & MCG_EXT_P) && MCG_EXT_CNT(cap) >= 9)
@@ -1507,8 +1528,6 @@ static int __mcheck_cpu_cap_init(void)
if (cap & MCG_SER_P)
mca_cfg.ser = 1;
-
- return 0;
}
static void __mcheck_cpu_init_generic(void)
@@ -1535,9 +1554,10 @@ static void __mcheck_cpu_init_generic(void)
static void __mcheck_cpu_init_clear_banks(void)
{
+ struct mce_bank *mce_banks = this_cpu_ptr(mce_banks_array);
int i;
- for (i = 0; i < mca_cfg.banks; i++) {
+ for (i = 0; i < this_cpu_read(mce_num_banks); i++) {
struct mce_bank *b = &mce_banks[i];
if (!b->init)
@@ -1548,6 +1568,33 @@ static void __mcheck_cpu_init_clear_banks(void)
}
/*
+ * Do a final check to see if there are any unused/RAZ banks.
+ *
+ * This must be done after the banks have been initialized and any quirks have
+ * been applied.
+ *
+ * Do not call this from any user-initiated flows, e.g. CPU hotplug or sysfs.
+ * Otherwise, a user who disables a bank will not be able to re-enable it
+ * without a system reboot.
+ */
+static void __mcheck_cpu_check_banks(void)
+{
+ struct mce_bank *mce_banks = this_cpu_ptr(mce_banks_array);
+ u64 msrval;
+ int i;
+
+ for (i = 0; i < this_cpu_read(mce_num_banks); i++) {
+ struct mce_bank *b = &mce_banks[i];
+
+ if (!b->init)
+ continue;
+
+ rdmsrl(msr_ops.ctl(i), msrval);
+ b->init = !!msrval;
+ }
+}
+
+/*
* During IFU recovery Sandy Bridge -EP4S processors set the RIPV and
* EIPV bits in MCG_STATUS to zero on the affected logical processor (SDM
* Vol 3B Table 15-20). But this confuses both the code that determines
@@ -1578,6 +1625,7 @@ static void quirk_sandybridge_ifu(int bank, struct mce *m, struct pt_regs *regs)
/* Add per CPU specific workarounds here */
static int __mcheck_cpu_apply_quirks(struct cpuinfo_x86 *c)
{
+ struct mce_bank *mce_banks = this_cpu_ptr(mce_banks_array);
struct mca_config *cfg = &mca_cfg;
if (c->x86_vendor == X86_VENDOR_UNKNOWN) {
@@ -1587,7 +1635,7 @@ static int __mcheck_cpu_apply_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 && cfg->banks > 4) {
+ if (c->x86 == 15 && this_cpu_read(mce_num_banks) > 4) {
/*
* disable GART TBL walk error reporting, which
* trips off incorrectly with the IOMMU & 3ware
@@ -1606,7 +1654,7 @@ static int __mcheck_cpu_apply_quirks(struct cpuinfo_x86 *c)
* Various K7s with broken bank 0 around. Always disable
* by default.
*/
- if (c->x86 == 6 && cfg->banks > 0)
+ if (c->x86 == 6 && this_cpu_read(mce_num_banks) > 0)
mce_banks[0].ctl = 0;
/*
@@ -1628,7 +1676,7 @@ static int __mcheck_cpu_apply_quirks(struct cpuinfo_x86 *c)
* valid event later, merely don't write CTL0.
*/
- if (c->x86 == 6 && c->x86_model < 0x1A && cfg->banks > 0)
+ if (c->x86 == 6 && c->x86_model < 0x1A && this_cpu_read(mce_num_banks) > 0)
mce_banks[0].init = 0;
/*
@@ -1814,7 +1862,9 @@ void mcheck_cpu_init(struct cpuinfo_x86 *c)
if (!mce_available(c))
return;
- if (__mcheck_cpu_cap_init() < 0 || __mcheck_cpu_apply_quirks(c) < 0) {
+ __mcheck_cpu_cap_init();
+
+ if (__mcheck_cpu_apply_quirks(c) < 0) {
mca_cfg.disabled = 1;
return;
}
@@ -1831,6 +1881,7 @@ void mcheck_cpu_init(struct cpuinfo_x86 *c)
__mcheck_cpu_init_generic();
__mcheck_cpu_init_vendor(c);
__mcheck_cpu_init_clear_banks();
+ __mcheck_cpu_check_banks();
__mcheck_cpu_setup_timer();
}
@@ -1862,7 +1913,7 @@ static void __mce_disable_bank(void *arg)
void mce_disable_bank(int bank)
{
- if (bank >= mca_cfg.banks) {
+ if (bank >= this_cpu_read(mce_num_banks)) {
pr_warn(FW_BUG
"Ignoring request to disable invalid MCA bank %d.\n",
bank);
@@ -1948,9 +1999,10 @@ int __init mcheck_init(void)
*/
static void mce_disable_error_reporting(void)
{
+ struct mce_bank *mce_banks = this_cpu_ptr(mce_banks_array);
int i;
- for (i = 0; i < mca_cfg.banks; i++) {
+ for (i = 0; i < this_cpu_read(mce_num_banks); i++) {
struct mce_bank *b = &mce_banks[i];
if (b->init)
@@ -2050,26 +2102,47 @@ static struct bus_type mce_subsys = {
DEFINE_PER_CPU(struct device *, mce_device);
-static inline struct mce_bank *attr_to_bank(struct device_attribute *attr)
+static inline struct mce_bank_dev *attr_to_bank(struct device_attribute *attr)
{
- return container_of(attr, struct mce_bank, attr);
+ return container_of(attr, struct mce_bank_dev, attr);
}
static ssize_t show_bank(struct device *s, struct device_attribute *attr,
char *buf)
{
- return sprintf(buf, "%llx\n", attr_to_bank(attr)->ctl);
+ u8 bank = attr_to_bank(attr)->bank;
+ struct mce_bank *b;
+
+ if (bank >= per_cpu(mce_num_banks, s->id))
+ return -EINVAL;
+
+ b = &per_cpu(mce_banks_array, s->id)[bank];
+
+ if (!b->init)
+ return -ENODEV;
+
+ return sprintf(buf, "%llx\n", b->ctl);
}
static ssize_t set_bank(struct device *s, struct device_attribute *attr,
const char *buf, size_t size)
{
+ u8 bank = attr_to_bank(attr)->bank;
+ struct mce_bank *b;
u64 new;
if (kstrtou64(buf, 0, &new) < 0)
return -EINVAL;
- attr_to_bank(attr)->ctl = new;
+ if (bank >= per_cpu(mce_num_banks, s->id))
+ return -EINVAL;
+
+ b = &per_cpu(mce_banks_array, s->id)[bank];
+
+ if (!b->init)
+ return -ENODEV;
+
+ b->ctl = new;
mce_restart();
return size;
@@ -2184,7 +2257,7 @@ static void mce_device_release(struct device *dev)
kfree(dev);
}
-/* Per cpu device init. All of the cpus still share the same ctrl bank: */
+/* Per CPU device init. All of the CPUs still share the same bank device: */
static int mce_device_create(unsigned int cpu)
{
struct device *dev;
@@ -2216,8 +2289,8 @@ static int mce_device_create(unsigned int cpu)
if (err)
goto error;
}
- for (j = 0; j < mca_cfg.banks; j++) {
- err = device_create_file(dev, &mce_banks[j].attr);
+ for (j = 0; j < per_cpu(mce_num_banks, cpu); j++) {
+ err = device_create_file(dev, &mce_bank_devs[j].attr);
if (err)
goto error2;
}
@@ -2227,7 +2300,7 @@ static int mce_device_create(unsigned int cpu)
return 0;
error2:
while (--j >= 0)
- device_remove_file(dev, &mce_banks[j].attr);
+ device_remove_file(dev, &mce_bank_devs[j].attr);
error:
while (--i >= 0)
device_remove_file(dev, mce_device_attrs[i]);
@@ -2248,8 +2321,8 @@ static void mce_device_remove(unsigned int cpu)
for (i = 0; mce_device_attrs[i]; i++)
device_remove_file(dev, mce_device_attrs[i]);
- for (i = 0; i < mca_cfg.banks; i++)
- device_remove_file(dev, &mce_banks[i].attr);
+ for (i = 0; i < per_cpu(mce_num_banks, cpu); i++)
+ device_remove_file(dev, &mce_bank_devs[i].attr);
device_unregister(dev);
cpumask_clear_cpu(cpu, mce_device_initialized);
@@ -2270,6 +2343,7 @@ static void mce_disable_cpu(void)
static void mce_reenable_cpu(void)
{
+ struct mce_bank *mce_banks = this_cpu_ptr(mce_banks_array);
int i;
if (!mce_available(raw_cpu_ptr(&cpu_info)))
@@ -2277,7 +2351,7 @@ static void mce_reenable_cpu(void)
if (!cpuhp_tasks_frozen)
cmci_reenable();
- for (i = 0; i < mca_cfg.banks; i++) {
+ for (i = 0; i < this_cpu_read(mce_num_banks); i++) {
struct mce_bank *b = &mce_banks[i];
if (b->init)
@@ -2327,10 +2401,12 @@ static __init void mce_init_banks(void)
{
int i;
- for (i = 0; i < mca_cfg.banks; i++) {
- struct mce_bank *b = &mce_banks[i];
+ for (i = 0; i < MAX_NR_BANKS; i++) {
+ struct mce_bank_dev *b = &mce_bank_devs[i];
struct device_attribute *a = &b->attr;
+ b->bank = i;
+
sysfs_attr_init(&a->attr);
a->attr.name = b->attrname;
snprintf(b->attrname, ATTR_LEN, "bank%d", i);
@@ -2440,22 +2516,16 @@ static int fake_panic_set(void *data, u64 val)
DEFINE_DEBUGFS_ATTRIBUTE(fake_panic_fops, fake_panic_get, fake_panic_set,
"%llu\n");
-static int __init mcheck_debugfs_init(void)
+static void __init mcheck_debugfs_init(void)
{
- struct dentry *dmce, *ffake_panic;
+ struct dentry *dmce;
dmce = mce_get_debugfs_dir();
- if (!dmce)
- return -ENOMEM;
- ffake_panic = debugfs_create_file_unsafe("fake_panic", 0444, dmce,
- NULL, &fake_panic_fops);
- if (!ffake_panic)
- return -ENOMEM;
-
- return 0;
+ debugfs_create_file_unsafe("fake_panic", 0444, dmce, NULL,
+ &fake_panic_fops);
}
#else
-static int __init mcheck_debugfs_init(void) { return -EINVAL; }
+static void __init mcheck_debugfs_init(void) { }
#endif
DEFINE_STATIC_KEY_FALSE(mcsafe_key);
@@ -2463,8 +2533,6 @@ EXPORT_SYMBOL_GPL(mcsafe_key);
static int __init mcheck_late_init(void)
{
- pr_info("Using %d MCE banks\n", mca_cfg.banks);
-
if (mca_cfg.recovery)
static_branch_inc(&mcsafe_key);
Copyright © 1996 – 2023 Jason A. Donenfeld. All Rights Reverse Engineered.