x86, NMI: fix watchdog failure message
authorAristeu Rozanski <arozansk@redhat.com>
Fri, 15 Aug 2008 12:36:14 +0000 (08:36 -0400)
committerIngo Molnar <mingo@elte.hu>
Fri, 15 Aug 2008 13:34:59 +0000 (15:34 +0200)
> it just won't work at boot time - the second logic unit will be stuck:
>
> Booting processor 1/2 APIC 0x1
> Initializing CPU#1
> Calibrating delay using timer specific routine.. 5586.12 BogoMIPS (lpj=2793063)
> CPU: Trace cache: 12K uops, L1 D cache: 16K
> CPU: L2 cache: 1024K
> CPU: Physical Processor ID: 0
> CPU: Processor Core ID: 1
> CPU1: Thermal monitoring enabled (TM1)
>               Intel(R) Pentium(R) D CPU 2.80GHz stepping 04
> Brought up 2 CPUs
> testing NMI watchdog ... <4>WARNING: CPU#1: NMI appears to be stuck (0->0)!

while at it... - fix that newline

Signed-off-by: Aristeu Rozanski <aris@redhat.com>
Cc: jvillalo@redhat.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
arch/x86/kernel/nmi.c

index ac6d51222e7d3562abb1e7b5e8bd4a16a1c1c1f4..919473ad4a293e8bfccdc6a8eacb1dad39bc9622 100644 (file)
@@ -142,11 +142,15 @@ int __init check_nmi_watchdog(void)
                if (!per_cpu(wd_enabled, cpu))
                        continue;
                if (get_nmi_count(cpu) - prev_nmi_count[cpu] <= 5) {
+                       printk("\n");
                        printk(KERN_WARNING "WARNING: CPU#%d: NMI "
                                "appears to be stuck (%d->%d)!\n",
                                cpu,
                                prev_nmi_count[cpu],
                                get_nmi_count(cpu));
+                       printk(KERN_WARNING "Please report this to "
+                              "linux-kernel@vger.kernel.org and attach "
+                              "the output of 'dmesg' command.\n");
                        per_cpu(wd_enabled, cpu) = 0;
                        atomic_dec(&nmi_active);
                }