Re: [Xen-devel] [PATCH v2 4/9] x86/pagewalk: Clean up guest_supports_* predicates

[Andrew Cooper <andrew.cooper3@xxxxxxxxxx>]

On 20/03/17 13:59, Jan Beulich wrote:
>>>> On 20.03.17 at 14:36, <andrew.cooper3@xxxxxxxxxx> wrote:
>> On 20/03/17 08:45, Jan Beulich wrote:
>>> Also I'm still not really happy with the guest_supports_ prefixes
>>> for this and its L2 counterpart: The question here isn't whether the
>>> guest supports it (we can't know whether it does), but whether it
>>> enabled PSE/PAE/LM. Arguably the L3 case is less clear because
>>> of the mentioned lack of an explicit enabled bit, so I can live with
>>> the patch going in unchanged (the L2 side then simply for things
>>> to remain consistent, albeit there's then already the difference of
>>> parameter types).
>> How would you prefer them to be named?
> I think I did (or at least had meant to) suggest guest_uses_...() or
> something similar.

Grammatically, that is still somewhat awkward.

How about guest_can_use_...() ? That logically covers both that the
feature might be missing, or the control register might not be suitably
configured.

~Andrew

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