Re: [Xen-devel] [PATCH v1] x86/altp2m: Add hypercall to create a new view and set sve bits

[Alexandru Stefan ISAILA <aisaila@xxxxxxxxxxxxxxx>]

On 25.10.2019 17:36, Alexandru Stefan ISAILA wrote:
> 
> 
> On 23.10.2019 14:58, Roger Pau Monné wrote:
>> On Wed, Oct 23, 2019 at 09:11:54AM +0000, Alexandru Stefan ISAILA wrote:
>>>
>>>
>>> On 03.09.2019 20:24, Tamas K Lengyel wrote:
>>>> On Tue, Sep 3, 2019 at 9:53 AM Jan Beulich <jbeulich@xxxxxxxx> wrote:
>>>>>
>>>>> On 02.09.2019 10:11, Alexandru Stefan ISAILA wrote:
>>>>>> --- a/xen/include/public/hvm/hvm_op.h
>>>>>> +++ b/xen/include/public/hvm/hvm_op.h
>>>>>> @@ -244,6 +244,7 @@ struct xen_hvm_altp2m_view {
>>>>>>         /* Create view only: default access type
>>>>>>          * NOTE: currently ignored */
>>>>>>         uint16_t hvmmem_default_access; /* xenmem_access_t */
>>>>>> +    uint8_t set_sve; /* bool value */
>>>>>>     };
>>>>>
>>>>> This interface is, given the right configuration, available to
>>>>> guests. Hence you can't simply add a field here. Just consider
>>>>> what happens for an existing caller when there is random data
>>>>> in the field you now assign a meaning.
>>>>
>>>> Perhaps instead of extending the HVMOP it would make more sense to
>>>> just add a xl config option that defines the "default" sve bit for
>>>> altp2m views in the domain?
>>>>
>>>
>>> Adding a xl config option will not work for systems that do use xl.
>>> There is a need that this will work in all cases.
>>
>> I assume that such option would be implemented using a DOMCTL, which
>> can also be used by other toolstacks. I however have no idea whether
>> this is a suitable interface or not for this feature.
>>
> 
> I think that having a HVMOP_altp2m_get_suppress_ve_multi and letting the
HVMOP_altp2m_set_suppress_ve_multi (sorry for the typo)
> caller provide the start gfn and the nr of pages to have the sve bits
> set will provide a good solution for init an dfor further development.
> 
> I will go on this way for version 2 if everyone is ok with this.
> 
> Thanks,
> Alex
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@xxxxxxxxxxxxxxxxxxxx
> https://lists.xenproject.org/mailman/listinfo/xen-devel
> ________________________
> This email was scanned by Bitdefender
> 
_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/mailman/listinfo/xen-devel