Kevin Wolf <kw...@redhat.com> writes: > I have to admit that I didn't follow your discussion very closely any > more after a few versions, so just to confirm: You came to the > conclusion that we want to add drive_del to QMP and not only the human > monitor, even though there is no drive_add in QMP?
Fair question. -drive and drive_add conflate host and guest part. They were hacked up to work with -device, and it shows. I'd rather not copy the mess to QMP. Maybe it's best to leave out the QMP part for now, i.e. drop PATCH 2/2. If we can't get blockdev_add, blockdev_del done in time, we reconsider. > The HMP help text doesn't seem to be completely right (sent a comment), > but once that's fixed, I'm going to merge the series based on Markus's ACK.