On 04.08.2025 10:29, Penny, Zheng wrote:
> [Public]
> 
>> -----Original Message-----
>> From: Jan Beulich <[email protected]>
>> Sent: Monday, August 4, 2025 3:41 PM
>> To: Penny, Zheng <[email protected]>
>> Cc: Huang, Ray <[email protected]>; Andrew Cooper
>> <[email protected]>; Anthony PERARD <[email protected]>;
>> Orzel, Michal <[email protected]>; Julien Grall <[email protected]>; Roger 
>> Pau
>> MonnĂ© <[email protected]>; Stefano Stabellini <[email protected]>; 
>> xen-
>> [email protected]
>> Subject: Re: [PATCH v1 05/25] xen: introduce CONFIG_DOMCTL
>>
>> On 03.08.2025 11:47, Penny Zheng wrote:
>>> --- a/xen/common/Kconfig
>>> +++ b/xen/common/Kconfig
>>> @@ -627,6 +627,10 @@ config SYSCTL
>>>       This option shall only be disabled on some dom0less systems, or
>>>       PV shim on x86, to reduce Xen footprint.
>>>
>>> +config DOMCTL
>>> +   bool "Enable domctl hypercall"
>>> +   def_bool y
>>> +
>>
>> Just to re-iterate - we don't think we want things to be this fine-grained.
>> (As an aside, nit: "bool" and "def_bool" are partly redundant with one
>> another.)
>>
> 
> Are we suggesting to use one Kconfig, maybe like CONFIG_XENCTL(not a good 
> choice, just popping in my head...), to wrap all scenarios, including 
> sysctl-op, domctl-op, jiqian's platform-op, etc ?

Yes, that's the thought that was circulated, and that I had hoped Stefano
would have conveyed.

> In which case, maybe we still submit commits(or features) serie by serie, 
> more easy to review,  but only when all is completed, we make this Kconfig as 
> an selectable option ?

Likely the best route, but that may then mean stepping back a little on
SYSCTL, before trying to deal with SYSCTL and maybe PLATFORM_OP (albeit I
raised further reservations there).

Jan

Reply via email to