Re: [PATCH] x86: fix CMCI injection

[Egger Christoph <chegger@amazon.de>]

On 25.03.13 15:30, Jan Beulich wrote:
>>>> On 25.03.13 at 14:41, Egger Christoph <chegger@amazon.de> wrote:
>> On 26.02.13 07:04, Keir Fraser wrote:
>>> On 25/02/2013 16:49, "Jan Beulich" <JBeulich@suse.com> wrote:
>>>> --- a/xen/arch/x86/cpu/mcheck/mce.c
>>>> +++ b/xen/arch/x86/cpu/mcheck/mce.c
>>>> @@ -34,6 +34,7 @@ bool_t __read_mostly mce_broadcast = 0;
>>>>    bool_t is_mc_panic;
>>>>    unsigned int __read_mostly nr_mce_banks;
>>>>    unsigned int __read_mostly firstbank;
>>>> +uint8_t __read_mostly cmci_apic_vector;
>> ...
>>>> @@ -1479,11 +1474,15 @@ long do_mca(XEN_GUEST_HANDLE_PARAM(xen_m
>>>>                on_selected_cpus(cpumap, x86_mc_mceinject, NULL, 1);
>>>>                break;
>>>>            case XEN_MC_INJECT_TYPE_CMCI:
>>>> -            if ( !cmci_support )
>>>> +            if ( !cmci_apic_vector )
>>
>> cmci_apic_vector is uninitialized here when the platform does not
>> support CMCI.
>
> How that? It's zero in that case (just like any other global variable
> without explicit initializer), which is exactly what the conditional
> tests for.

umm... right. Sorry for the noise.

Christoph