On Tue, Apr 30, 2013 at 9:30 AM, Helga Velroyen <[email protected]> wrote:
> Hi!
>
>
> On Tue, Apr 30, 2013 at 9:26 AM, Guido Trotter <[email protected]> wrote:
>>
>> On Tue, Apr 30, 2013 at 9:18 AM, Helga Velroyen <[email protected]> wrote:
>> > Hi!
>> >
>> >
>> >
>> >
>> > On Tue, Apr 30, 2013 at 9:08 AM, Guido Trotter <[email protected]>
>> > wrote:
>> >>
>> >> On Mon, Apr 29, 2013 at 7:06 PM, Helga Velroyen <[email protected]>
>> >> wrote:
>> >> > This patch extends the man page of the 'gnt-instance' command
>> >> > by describing the newly implemented hypervisor parameter
>> >> > 'vif_type'.
>> >> >
>> >> > Signed-off-by: Helga Velroyen <[email protected]>
>> >>
>> >> LGTM
>> >>
>> >> > ---
>> >> >  man/gnt-instance.rst | 10 ++++++++++
>> >> >  1 file changed, 10 insertions(+)
>> >> >
>> >> > diff --git a/man/gnt-instance.rst b/man/gnt-instance.rst
>> >> > index f65dd47..3a4fbf5 100644
>> >> > --- a/man/gnt-instance.rst
>> >> > +++ b/man/gnt-instance.rst
>> >> > @@ -270,6 +270,16 @@ nic\_type
>> >> >      - e1000 (KVM)
>> >> >      - paravirtual (default for KVM) (HVM & KVM)
>> >> >
>> >> > +vif\_type
>> >> > +    Valid for the Xen HVM hypervisor.
>> >> > +
>> >> > +    This parameter specifies the vif type of the nic configuration
>> >> > +    of the instance. Unsetting the value leads to no type being
>> >> > specified
>> >> > +    in the configuration. The possible options are:
>> >> > +
>> >> > +    - ioemu (default for Xen HVM)
>> >>
>> >> You could even remove "for Xen HVM" here, since the hv parameter is
>> >> only valid for Xen HVM.
>> >
>> >
>> > ok, I'm sending an interdif.
>> >
>> >>
>> >>
>> >> > +    - vif
>> >> > +
>> >> >  disk\_type
>> >> >      Valid for the Xen HVM and KVM hypervisors.
>> >> >
>> >>
>> >> Thanks,
>> >>
>> >> Guido
>> >
>> >
>> > I also realized that it might be good to know that the parameter is only
>> > used if nic_type is not set, I added t his to the description:
>> >
>>
>> Why? Is it invalid to have a "type=" which is not ioemu when specifying a
>> model?
>
>
> According to the specs, "model" is only valid with "type=ioemu". So whenever
> a model is passed, I ignore the vif_type and set it to ioemu to comply the
> specs. But that does not affect the case that caused the bug, the bug is
> about the case when no model and (so far) no type was specified.

Ah, perfect, sounds good.
LGTM on the interdiff.

Thanks,

Guido

Reply via email to