On Tue, Jan 13, 2015 at 08:07:15PM +0100, Christian Borntraeger wrote: > Am 13.01.2015 um 17:04 schrieb Stefan Hajnoczi: > [...] > > I'm really starting to get worried that you are going to break things. > > This DASD hack is a layering violation but okay, go ahead if you want. > > But now you are also thinking about breaking live migration. > > > > The proper thing to do is to introduce libvirt XML syntax for DASD. > > That way the geometry can be handled as part of the machine > > configuration. Then live migration and storage management tools can do > > the right thing. > > > > I've said this should be done in libvirt repeatedly but you keep wanting > > to hack QEMU instead of doing this cleanly :(. > > > > If you have plans to expand on this hack, please scrap this series and > > introduce libvirt XML syntax instead. > > We had no plans to expand on this band-aid. I just tried to come up with > ideas beyond this hack to make this more acceptible to you (because I though > that > you want something on top). Seems that I misunderstood you, so if you prefer > to just keep this hack and not do anything further in that direction, this is > totally fine with me. > > So lets just fix up the small nits and go ahead, ok?
Yes, I'm fine with this series (modulo the review comments which require a v6). Stefan
pgpQDfp1dN8wP.pgp
Description: PGP signature
