x86: avoid theoretical spurious NMI backtraces with CONFIG_CPUMASK_OFFSTACK=y
authorRusty Russell <rusty@rustcorp.com.au>
Tue, 21 Apr 2009 06:33:41 +0000 (16:03 +0930)
committerIngo Molnar <mingo@elte.hu>
Tue, 21 Apr 2009 08:09:50 +0000 (10:09 +0200)
In theory (though not shown in practice) alloc_cpumask_var() doesn't zero
memory, so CPUs might print an "NMI backtrace for cpu %d" once on boot.

(Bug introduced in fcef8576d8a64fc603e719c97d423f9f6d4e0e8b).

[ Impact: avoid theoretical syslog noise in rare configs ]

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Cc: Steven Rostedt <rostedt@goodmis.org>
LKML-Reference: <alpine.DEB.2.00.0904202113520.10097@gandalf.stny.rr.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
arch/x86/kernel/apic/nmi.c

index 2ba52f35a88c324ddb848711a1d3d9c11df57f92..ce4fbfa315a16ac1f4e45f34281121577f5477a6 100644 (file)
@@ -138,7 +138,7 @@ int __init check_nmi_watchdog(void)
        if (!prev_nmi_count)
                goto error;
 
-       alloc_cpumask_var(&backtrace_mask, GFP_KERNEL);
+       alloc_cpumask_var(&backtrace_mask, GFP_KERNEL|__GFP_ZERO);
        printk(KERN_INFO "Testing NMI watchdog ... ");
 
 #ifdef CONFIG_SMP