Re: [PATCH v15 07/19] pvh: Disable unneeded features of HVM containers

[George Dunlap <george.dunlap@eu.citrix.com>]

On 11/12/2013 03:03 PM, Jan Beulich wrote:
>>>> On 12.11.13 at 15:56, George Dunlap <george.dunlap@eu.citrix.com> wrote:
>> On 11/12/2013 01:51 PM, Jan Beulich wrote:
>>>>>> On 11.11.13 at 15:57, George Dunlap <george.dunlap@eu.citrix.com> wrote:
>>>> @@ -1414,6 +1442,7 @@ int hvm_hap_nested_page_fault(paddr_t gpa,
>>>>        /* For the benefit of 32-bit WinXP (& older Windows) on AMD CPUs,
>>>>         * a fast path for LAPIC accesses, skipping the p2m lookup. */
>>>>        if ( !nestedhvm_vcpu_in_guestmode(v)
>>>> +         && is_hvm_vcpu(v)
>>>
>>> This change looks unnecessary - nestedhvm_vcpu_enter_guestmode()
>>> (the only way by which nestedhvm_vcpu_in_guestmode() could return
>>> other than zero) must obviously never be used on a PVH guest.
>>
>> You've got the polarity of that check wrong -- PVH guests are never in
>> guest mode, therefore !nestedhvm_vcpu_in_guestmode() is always true.
>
> Oh, indeed. In which case I'd slightly prefer !is_pvh_vcpu() here,
> but I guess it's not really significant anyway.

The plan I think is for is_pvh to go away, and instead to have a flag 
indicating whether the guest has / is using a LAPIC.

  -George