On Thu, Feb 03, 2011 at 09:16:09PM -0700, Eric Blake wrote:
Not sure this makes sense to do; posting it for comments,
but I'm okay with ditching the idea (especially since
existing RHEL 6.0 guests have already been using XML that
targets virtio, which means changing to vdagent would
cause the need for a naming transition).
This patch doesn't fit the modelling.
index 0e82394..b99f3ce 100644
--- a/tests/qemuxml2argvdata/qemuxml2argv-channel-spicevmc.xml
+++ b/tests/qemuxml2argvdata/qemuxml2argv-channel-spicevmc.xml
@@ -26,7 +26,7 @@
<channel name='main' mode='secure'/>
</graphics>
<channel type='spicevmc'>
- <target type='virtio' name='com.redhat.spice.0'/>
+ <target type='vdagent'/>
<address type='virtio-serial' controller='1' bus='0'
port='3'/>
</channel>
In <target> 'type' is refering to the guest hardware type,
which is always 'virtio'. 'vdgent' refers to a service running
over the device, not the device type. The '@name' attribute
is used to identify services, in this particular case 'com.redhat.com.spice.0'
refers to the vdagent service type.
Regards,
Daniel
--
|:
http://berrange.com -o-
http://www.flickr.com/photos/dberrange/ :|
|:
http://libvirt.org -o-
http://virt-manager.org :|
|:
http://autobuild.org -o-
http://search.cpan.org/~danberr/ :|
|:
http://entangle-photo.org -o-
http://live.gnome.org/gtk-vnc :|