Re: [PATCH v2] x86: Fixup IRQs when CPUs go down during shutdown

[Andrew Cooper <andrew.cooper3@citrix.com>]

On 02/12/15 15:30, Jan Beulich wrote:
>>>> On 02.12.15 at 16:09, <ross.lagerwall@citrix.com> wrote:
>> On 12/02/2015 02:02 PM, Jan Beulich wrote:
>>>>>> On 02.12.15 at 14:46, <ross.lagerwall@citrix.com> wrote:
>>>> --- a/xen/arch/x86/smp.c
>>>> +++ b/xen/arch/x86/smp.c
>>>> @@ -286,6 +286,7 @@ void __stop_this_cpu(void)
>>>>
>>>>   static void stop_this_cpu(void *dummy)
>>>>   {
>>>> +    fixup_eoi();
>>>>       __stop_this_cpu();
>>> Is this really needed during shutdown?
>> Possibly not, but I think it's cleaner to do the same as what is used 
>> for CPU down.
> I'm not convinced. Andrew?

Suppose there is an oustanding line interrupt on the pending eoi stack. 
Without this fixup_eoi(), it could stay permanently attached to a cpu
which isn't processing anything further.

With the current use of smp_send_stop(), all cpus will end up in a state
where they can only be recovered with an #INIT, so I suppose it doesn't
actually matter.

Therefore, not performing redundant work is probably the best course of
action.

~Andrew