On 2013年04月29日 22:26, Jiri Denemark wrote:
On Mon, Apr 22, 2013 at 17:38:23 +0800, Li Zhang wrote:
> On 2013年04月22日 17:10, Jiri Denemark wrote:
>> If you want to add powernv feature just because you need to distinguish
>> if the host supports KVM or not, there's much better way... In guest
>> section of capabilities XML, KVM support is indicated by <domain
>> type='kvm'> element. And that's what existing apps already use to
detect
>> KVM presence/absence.
> As my understanding, there is still some difference from 'kvm' capability.
> 'powernv' is only considered as one CPU feature of PPC64.
> If other PPC platforms support KVM in the future, this feature can be
> used to identify whether migration can be executed . :)
Well, as I already said, unless there is a way to explicitly
enable/disable powernv feature in guest CPU, I don't see any reason for
exposing the feature to users/apps. We definitely don't want apps/users
to detect support for migration (or anything else) by checking host CPU
features. If migration is not supported, then any migration API can fail
with VIR_ERR_OPERATION_UNSUPPORTED. That said, powernv feature can be
used internally by libvirt to check if some configurations/operations
are supported but it doesn't have to be exposed to users/apps. The CPU
driver stuff is there for configuring *guest* CPU.
I see, got it. This feature can be removed.
So it is not necessary to add <feature> in CPU driver for ppc64.
Thanks.
Jirka