[libvirt] [PATCH] qemu: allow to control host side link status of network device

[Laine Stump]

On 05/18/2017 04:32 PM, Vasiliy Tolstov wrote:
> 2017-05-17 15:28 GMT+03:00 Peter Krempa <pkrempa at redhat.com>:
>>>
>>> Well, *I* think I've given sufficient reasons for having the two link
>>> states controlled separately, but since Dan and Peter had questioned its
>>> usefulness, we should see whether or not I've swayed their opinions :-)
>>
>> I think we should have two elements in this case. I just don't consider
>> setting the host link side to be very useful.
>>
> 
> Sorry, Laine. As i understand you already have alternative patch for
> such feature, can you share it?
> I'm really interesting how much it different from my.

After much searching, I finally found my patch, and learned that it was
incomplete (I'd suspected that) so not of much use. The one thing I
would say is that I had a prerequisite patch that changed the name of
"linkstate" both in the virDomainNetDef and the temp local in the
parsing function to "guestLinkState" so that it was more difficult to
confuse the two.

Aside from that, I looked through your patch and have a few comments the
I will post. (it's not a thorough review, because I have a hard
deadlines of tomorrow AM for a couple of completely unrelated (to
libvirt, or even software development) tasks. (Protip - *never*
volunteer to be the "parent advisor" for your kid's yearbook if you will
be moving during the time that it all must be done. Also never store
anything in heavy glass bottles directly over an electric stovetop that
is itself made of glass).