Re: [Qemu-devel] Should we auto-generate IDs? (was: [PATCH] qdev-monitor.c: Add device id generation)

[Daniel P. Berrange]

On Wed, Aug 26, 2015 at 02:01:41PM -0400, Programmingkid wrote:
> > If a user is talking to the QEMU monitor directly there are plenty of ways
> > to go wrong, of which forgetting to provide an ID is a really minor one.
> 
> What other problems did you have in mind?
> 
> > That's why it is generally left to higher level mgmt layers to talk to
> > QEMU and deal with all the issues in this area. IOW if users are talking
> > to the monitor directly, IMHO they've already lost.
> 
> I'm not following you. What do you mean by higher level mgmt layers?

Using QEMU via libvirt, or a similar management layer and not try
to talk to the monitor and/or CLI args which are complex to get
right and not really designed for user friendliness in general.

> Let me put it this way, if a user were to add a usb device to QEMU, say
> a usb-mouse, but forgot to give it an ID. How do you expect that user to
> remove the device from QEMU?

object_del should be made to accept the QOM object path, eg the first
anonymous device appears with a path  /machine/peripheral-anon/device[0]
so you could just do 'object_del /machine/peripheral-anon/device[0]'

If people really want pretty short IDs, then they can remember to
specify them upfront, or use a higher level app that avoids this
kind of problem.

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 :|