evie/android_kernel_oneplus_msm8998
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

x86, mce: add machine check exception count in /proc/interrupts

Browse source 
Useful for debugging, but it's also good general policy
to have a counter for all special interrupts there. This makes it easier
to diagnose where a CPU is spending its time.

[ Impact: feature, debugging tool ]

Signed-off-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
This commit is contained in:
Andi Kleen 2009-05-27 21:56:52 +02:00 • committed by H. Peter Anvin
parent 48b1fddbb1
commit 01ca79f141
3 changed files with 17 additions and 0 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

3
arch/x86/include/asm/mce.h
View file

@ -89,6 +89,7 @@ struct mce_log {
extern int mce_disabled; extern int mce_disabled;
#include <asm/atomic.h> #include <asm/atomic.h>
#include <linux/percpu.h>
void mce_setup(struct mce *m); void mce_setup(struct mce *m);
void mce_log(struct mce *m); void mce_log(struct mce *m);
@ -123,6 +124,8 @@ static inline void mce_amd_feature_init(struct cpuinfo_x86 *c) { }
int mce_available(struct cpuinfo_x86 *c); int mce_available(struct cpuinfo_x86 *c);
DECLARE_PER_CPU(unsigned, mce_exception_count);
void mce_log_therm_throt_event(__u64 status); void mce_log_therm_throt_event(__u64 status);
extern atomic_t mce_entry; extern atomic_t mce_entry;

4
arch/x86/kernel/cpu/mcheck/mce.c
View file

@ -57,6 +57,8 @@ int mce_disabled;
atomic_t mce_entry; atomic_t mce_entry;
DEFINE_PER_CPU(unsigned, mce_exception_count);
/* /*
* Tolerant levels: * Tolerant levels:
* 0: always panic on uncorrected errors, log corrected errors * 0: always panic on uncorrected errors, log corrected errors
@ -359,6 +361,8 @@ void do_machine_check(struct pt_regs *regs, long error_code)
atomic_inc(&mce_entry); atomic_inc(&mce_entry);
__get_cpu_var(mce_exception_count)++;
if (notify_die(DIE_NMI, "machine check", regs, error_code, if (notify_die(DIE_NMI, "machine check", regs, error_code,
18, SIGKILL) == NOTIFY_STOP) 18, SIGKILL) == NOTIFY_STOP)
goto out; goto out;

10
arch/x86/kernel/irq.c
View file

@ -12,6 +12,7 @@
#include <asm/io_apic.h> #include <asm/io_apic.h>
#include <asm/irq.h> #include <asm/irq.h>
#include <asm/idle.h> #include <asm/idle.h>
#include <asm/mce.h>
atomic_t irq_err_count; atomic_t irq_err_count;
@ -93,6 +94,12 @@ static int show_other_interrupts(struct seq_file *p, int prec)
seq_printf(p, "%10u ", irq_stats(j)->irq_threshold_count); seq_printf(p, "%10u ", irq_stats(j)->irq_threshold_count);
seq_printf(p, " Threshold APIC interrupts\n"); seq_printf(p, " Threshold APIC interrupts\n");
# endif # endif
#endif
#if defined(CONFIG_X86_MCE) && defined(CONFIG_X86_64)
seq_printf(p, "%*s: ", prec, "MCE");
for_each_online_cpu(j)
seq_printf(p, "%10u ", per_cpu(mce_exception_count, j));
seq_printf(p, " Machine check exceptions\n");
#endif #endif
seq_printf(p, "%*s: %10u\n", prec, "ERR", atomic_read(&irq_err_count)); seq_printf(p, "%*s: %10u\n", prec, "ERR", atomic_read(&irq_err_count));
#if defined(CONFIG_X86_IO_APIC) #if defined(CONFIG_X86_IO_APIC)
@ -161,6 +168,9 @@ u64 arch_irq_stat_cpu(unsigned int cpu)
{ {
u64 sum = irq_stats(cpu)->__nmi_count; u64 sum = irq_stats(cpu)->__nmi_count;
#if defined(CONFIG_X86_MCE) && defined(CONFIG_X86_64)
sum += per_cpu(mce_exception_count, cpu);
#endif
#ifdef CONFIG_X86_LOCAL_APIC #ifdef CONFIG_X86_LOCAL_APIC
sum += irq_stats(cpu)->apic_timer_irqs; sum += irq_stats(cpu)->apic_timer_irqs;
sum += irq_stats(cpu)->irq_spurious_count; sum += irq_stats(cpu)->irq_spurious_count;