[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: [libvirt] [PATCH 0/2] conf: qemu: support new Hyper-V enlightenments in Qemu-3.1



John Ferlan <jferlan redhat com> writes:

> On 11/14/18 4:04 AM, Andrea Bolognani wrote:
>> On Tue, 2018-11-13 at 18:53 +0100, Vitaly Kuznetsov wrote:
>>> The upcoming Qemu-3.1 release will bring us two new Hyper-V enlightenments:
>>> hv_ipi and hv_evmcs. Support these in libvirt.
>>>
>>> Vitaly Kuznetsov (2):
>>>   conf: qemu: add support for Hyper-V PV IPIs
>>>   conf: qemu: add support for Hyper-V Enlightened VMCS
>> 
>> I have pointed out a few very minor issues with the patches, as
>> well as a couple areas where pre-existing issues could be fixed,
>> although I don't consider the latter a requirement for merging
>> the series ;)
>> 
>> One thing that I would like to see, though, is the change being
>> documented properly in the release notes (docs/news.xml). Please
>> include that and respin.
>> 
>
> One other thing that I've really tried to see done is splitting the
> "conf" and "qemu" patches.
>
> That way one deals with conf, docs, conf, util, xml2xmltest, while the
> other deals with the qemu specific changes qemu, xml2argv.
>
> Sometimes it's "harder" to do that - such as may be the case with
> various switches when a new case is added; however, in that case the
> added case in the conf/docs/etc patch would go with the "default:" or
> "last" case and then be moved with the subsequent qemu patch.
>

I have to admit my overall knowledge of libvirt pretty limited, probably
because of that I'm failing to see benefits of doing such split for
Qemu/KVM-only features (as the whole Hyper-V emulation is
Qemu/KVM-only). How is the first patch going to be tested? Compile-only?
Also, which of these patches should src/cpu/* hunks go to?

Would realy appreciate some guidance here) Thanks!

-- 
Vitaly


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]