On Mon, Jul 4, 2016 at 5:43 PM, Michael S. Tsirkin <m...@redhat.com> wrote:
> I think it's ok to add a new message if it actually brings some
> benefit, but I'm not sure why it makes sense to do it just in case.

The main benefit today would be to have a single code path to handle
disconnection, not dozens. The experimental branch I have (I have not
touched it for a while, it's unfinished etc), does also check that the
backend support a stop state, changing the link status state for ex.
Anyway, it's not part of this series, but I believe it would be way
easier to support this method in the long run than any time backend
disconnect...

-- 
Marc-André Lureau

Reply via email to