Messages in this thread Patch in this message | | | Date | Mon, 1 Nov 2010 11:45:36 -0400 | From | Don Zickus <> | Subject | Re: BUG: using smp_processor_id() in preemptible arch_trigger_all_cpu_backtrace_handler |
| |
On Mon, Nov 01, 2010 at 01:39:56PM +0100, Jan Kiszka wrote: > Hi, > > I was getting this BUG while running into a GPF: > > BUG: using smp_processor_id() in preemptible [00000000] code: qemu-system-x86/2248 > caller is arch_trigger_all_cpu_backtrace_handler+0x1d/0xf7 > Pid: 2248, comm: qemu-system-x86 Not tainted 2.6.36+ #12 > Call Trace: > [<ffffffff81203194>] debug_smp_processor_id+0xd8/0xf4 > [<ffffffff813803ab>] arch_trigger_all_cpu_backtrace_handler+0x1d/0xf7 > [<ffffffff81381e64>] notifier_call_chain+0xa4/0xdb > [<ffffffff81381efe>] __atomic_notifier_call_chain+0x63/0x95 > [<ffffffff81381e9b>] ? __atomic_notifier_call_chain+0x0/0x95 > [<ffffffff81381cfb>] ? sub_preempt_count+0x97/0xaa > [<ffffffff8121de8c>] ? pfn_to_dma_pte+0x73/0x190 > [<ffffffff81381f44>] atomic_notifier_call_chain+0x14/0x16 > [<ffffffff81381f74>] notify_die+0x2e/0x30 > [<ffffffff8137f498>] do_general_protection+0x121/0x142 > [<ffffffff81002df2>] system_call_fastpath+0x16/0x1b
<snip>
> > I guess this is not intended to trigger here, specifically as it showed > up first and may be misinterpreted as the core of the issue.
Heh. Yeah when I migrated the code, I completely forgot the notifier chain could be called from a preemptible context (ie not NMI).
This patch should fix it and I think it is the correct fix. Let me know how it works out.
Cheers, Don
diff --git a/arch/x86/kernel/apic/hw_nmi.c b/arch/x86/kernel/apic/hw_nmi.c index c7c9ae4..1bdd0b5 100644 --- a/arch/x86/kernel/apic/hw_nmi.c +++ b/arch/x86/kernel/apic/hw_nmi.c @@ -63,7 +63,7 @@ arch_trigger_all_cpu_backtrace_handler(struct notifier_block *self, { struct die_args *args = __args; struct pt_regs *regs; - int cpu = smp_processor_id(); + int cpu; switch (cmd) { case DIE_NMI: @@ -74,6 +74,7 @@ arch_trigger_all_cpu_backtrace_handler(struct notifier_block *self, } regs = args->regs; + cpu = smp_processor_id(); if (cpumask_test_cpu(cpu, to_cpumask(backtrace_mask))) { static arch_spinlock_t lock = __ARCH_SPIN_LOCK_UNLOCKED;
| |