On Mon, Dec 01, 2008 at 12:23:12PM +0100, Gerd v. Egidy wrote:
Hi Daniel,
> > This patch alone not, but this patch + the one in my first mail
> > (see
> >
https://www.redhat.com/archives/libvir-list/2008-November/msg00457.html)
> > together make it work for me. The first patch fixes the autostart order,
> > the second one adds the necessary conn structure.
>
> Oh yes, I totally missed the patch in your first mail. The first patch
> is definitely correct and I'll apply that shortly.
I just added some documentation (same as in virInitialize) to make sure this
bug does not get introduced again. New version attached.
> While your second
> patch is also functionally OK, I'm not entirely happy with creating a
> connection object deep inside the QEMU driver code.
Yeah, that was exactly my thought too.
> So I'm going to
> think about whether there's a better way todo that bit.
I looked through the rest of the qemu-initialization and it looks like this is
the only point where the conn-object is needed. So a solution would be to
directly access the network driver functions. But on the other hand these
functions seem all to imply that there is a valid conn available. So we would
have to change that, at least for the lookup-by-name case.
I've committed this patch now. I'll send a suggestion for the second patch
shortly...
Daniel
--
|: Red Hat, Engineering, London -o-
http://people.redhat.com/berrange/ :|
|:
http://libvirt.org -o-
http://virt-manager.org -o-
http://ovirt.org :|
|:
http://autobuild.org -o-
http://search.cpan.org/~danberr/ :|
|: GnuPG: 7D3B9505 -o- F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 :|