From 29b70081f7cb094513d5189e82d3478b50777a28 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Sun, 30 Oct 2005 14:59:40 -0800 Subject: [PATCH] i386 nmi_watchdog: Merge check_nmi_watchdog fixes from x86_64 The per cpu nmi watchdog timer is based on an event counter. idle cpus don't generate events so the NMI watchdog doesn't fire and the test to see if the watchdog is working fails. - Add nmi_cpu_busy so idle cpus don't mess up the test. - kmalloc prev_nmi_count to keep kernel stack usage bounded. - Improve the error message on failure so there is enough information to debug problems. Signed-off-by: Eric W. Biederman Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/i386/kernel/nmi.c | 39 +++++++++++++++++++++++++++++++++++++-- 1 file changed, 37 insertions(+), 2 deletions(-) (limited to 'arch') diff --git a/arch/i386/kernel/nmi.c b/arch/i386/kernel/nmi.c index 72515b8a1b1..d661703ac1c 100644 --- a/arch/i386/kernel/nmi.c +++ b/arch/i386/kernel/nmi.c @@ -100,16 +100,44 @@ int nmi_active; (P4_CCCR_OVF_PMI0|P4_CCCR_THRESHOLD(15)|P4_CCCR_COMPLEMENT| \ P4_CCCR_COMPARE|P4_CCCR_REQUIRED|P4_CCCR_ESCR_SELECT(4)|P4_CCCR_ENABLE) +#ifdef CONFIG_SMP +/* The performance counters used by NMI_LOCAL_APIC don't trigger when + * the CPU is idle. To make sure the NMI watchdog really ticks on all + * CPUs during the test make them busy. + */ +static __init void nmi_cpu_busy(void *data) +{ + volatile int *endflag = data; + local_irq_enable(); + /* Intentionally don't use cpu_relax here. This is + to make sure that the performance counter really ticks, + even if there is a simulator or similar that catches the + pause instruction. On a real HT machine this is fine because + all other CPUs are busy with "useless" delay loops and don't + care if they get somewhat less cycles. */ + while (*endflag == 0) + barrier(); +} +#endif + static int __init check_nmi_watchdog(void) { - unsigned int prev_nmi_count[NR_CPUS]; + volatile int endflag = 0; + unsigned int *prev_nmi_count; int cpu; if (nmi_watchdog == NMI_NONE) return 0; + prev_nmi_count = kmalloc(NR_CPUS * sizeof(int), GFP_KERNEL); + if (!prev_nmi_count) + return -1; + printk(KERN_INFO "Testing NMI watchdog ... "); + if (nmi_watchdog == NMI_LOCAL_APIC) + smp_call_function(nmi_cpu_busy, (void *)&endflag, 0, 0); + for (cpu = 0; cpu < NR_CPUS; cpu++) prev_nmi_count[cpu] = per_cpu(irq_stat, cpu).__nmi_count; local_irq_enable(); @@ -123,12 +151,18 @@ static int __init check_nmi_watchdog(void) continue; #endif if (nmi_count(cpu) - prev_nmi_count[cpu] <= 5) { - printk("CPU#%d: NMI appears to be stuck!\n", cpu); + endflag = 1; + printk("CPU#%d: NMI appears to be stuck (%d->%d)!\n", + cpu, + prev_nmi_count[cpu], + nmi_count(cpu)); nmi_active = 0; lapic_nmi_owner &= ~LAPIC_NMI_WATCHDOG; + kfree(prev_nmi_count); return -1; } } + endflag = 1; printk("OK.\n"); /* now that we know it works we can reduce NMI frequency to @@ -136,6 +170,7 @@ static int __init check_nmi_watchdog(void) if (nmi_watchdog == NMI_LOCAL_APIC) nmi_hz = 1; + kfree(prev_nmi_count); return 0; } /* This needs to happen later in boot so counters are working */ -- cgit v1.2.3