Re: [Xen-devel] [PATCH V2 2/3] xen/vm_event: Support for guest-requested events

["Jan Beulich" <JBeulich@xxxxxxxx>]

>>> On 26.06.15 at 09:17, <rcojocaru@xxxxxxxxxxxxxxx> wrote:
> On 06/26/2015 10:02 AM, Jan Beulich wrote:
>>>>> On 15.06.15 at 11:03, <rcojocaru@xxxxxxxxxxxxxxx> wrote:
>>> --- a/xen/include/public/hvm/hvm_op.h
>>> +++ b/xen/include/public/hvm/hvm_op.h
>>> @@ -389,6 +389,10 @@ 
>>> DEFINE_XEN_GUEST_HANDLE(xen_hvm_evtchn_upcall_vector_t);
>>>  
>>>  #endif /* defined(__i386__) || defined(__x86_64__) */
>>>  
>>> +#if defined(__XEN__) || defined(__XEN_TOOLS__)
>>> +#define HVMOP_request_vm_event 24
>>> +#endif /* defined(__XEN__) || defined(__XEN_TOOLS__) */
>> 
>> Isn't this _specifically_ meant to be usable by a guest?
> 
> Yes, but we still need Xen tools to subscribe to the events, control how
> they are received (in a dom0 or similarly privileged domain) and process
> them. That was my train of thought anyway, maybe I'm missing something
> (or I'm misinterpreting the macros).

The way it's done above you _limit_ the #define's visibility to
Xen and the tool stack, i.e. you namely exclude guests from
seeing (and hence using) it.

Jan


_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel