On 13/03/2019 17:42, Julien Grall wrote:
> Hi Andrew,
>
> On 13/03/2019 17:34, Andrew Cooper wrote:
>> On 13/03/2019 15:59, Jan Beulich wrote:
>>>>>> On 13.03.19 at 16:48, <julien.gr...@arm.com> wrote:
>>>> Hi,
>>>>
>>>> On 13/03/2019 15:40, Jan Beulich wrote:
>>>>>>>> On 13.03.19 at 16:24, <julien.gr...@arm.com> wrote:
>>>>>> On 13/03/2019 15:22, Jan Beulich wrote:
>>>>>>>>>> On 18.02.19 at 12:36, <julien.gr...@arm.com> wrote:
>>>>>>>> --- a/xen/include/asm-arm/mm.h
>>>>>>>> +++ b/xen/include/asm-arm/mm.h
>>>>>>>> @@ -321,10 +321,8 @@ struct page_info *get_page_from_gva(struct
>>>>>>>> vcpu *v,
>>>> vaddr_t va,
>>>>>>>>     #define SHARED_M2P_ENTRY         (~0UL - 1UL)
>>>>>>>>     #define SHARED_M2P(_e)           ((_e) == SHARED_M2P_ENTRY)
>>>>>>>>     -/* Xen always owns P2M on ARM */
>>>>>>>> +/* We don't have a M2P on Arm */
>>>>>>>>     #define set_gpfn_from_mfn(mfn, pfn) do { (void) (mfn),
>>>>>>>> (void)(pfn); }
>>>> while (0)
>>>>>>>> -#define mfn_to_gmfn(_d, mfn)  (mfn)
>>>>>>> So is the plan to remove the other macro from Arm then as well?
>>>>>> Do you mean mfn_to_gfn? If so it does not exist on Arm.
>>>>> No, I mean the one in context above - set_gpfn_from_mfn().
>>>> It is used in common code, so we would need to #idef the caller.
>>> Hmm, right, such #ifdef-ary would be undesirable (and two out of
>>> the three common code callers would need it.
>>>
>>>> I think it is better to provide a NOP implementation. Could be
>>>> moved somewhere
>>>> in the common header though. Any opinions?
>>> This would perhaps be better, now that you have HAVE_M2P.
>>
>> Given that "having an M2P" is now an x86-specific concept, I think
>> phasing set_gpfn_from_mfn()'s use out of common code is the way to go.
>
> So you never expect other architecture to use the M2P?

I guess that depends on how likely it is going to be that Xen gains a
new non-HVM-like virtualisation mode on a new architecture.

~Andrew

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel

Reply via email to