On Wed, Jun 09, 2021 at 11:25:26AM +0200, Boris Fiuczynski wrote: > On 5/25/21 12:20 PM, Daniel P. Berrangé wrote: > > On Fri, May 14, 2021 at 04:28:57PM -0500, Jonathon Jongsma wrote: > > > This is a short patch series based on an initial patch from Boris Fiuczynski > > > that I massaged a little bit and added a few additional patches. > > > > > > This allows you to define mediated devices in libvirt which are started > > > automatically at boot or parent device plugin. > > > > Sorry I missed review of this series before it was merged, but I think > > this design aproach is seriously wrong. This is exposing autostart as > > an XML property, but in all other parts of libvirt we expose autostart > > as an explicit API. I think we need to make the implemnentation > > consistent with the rest of libvirt. > > Daniel, > just wondering, if libvirt exposes this as an explicit API would the > assumption be that when a user sets up the device to autostart via libvirt > after a system reboot libvirt takes care of everything to bring the device > back up again? > > mdevctl does not take care of device driver binding of alternative device > drivers, e.g. vfio. > > Does that meet libvirt user expectations and be consistent with the rest of > libvirt? Autostart merely means that the equivalent of virNodeDeviceCreate is done at startup. Regards, Daniel -- |: https://berrange.com -o- https://www.flickr.com/photos/dberrange :| |: https://libvirt.org -o- https://fstop138.berrange.com :| |: https://entangle-photo.org -o- https://www.instagram.com/dberrange :|
