On Fri, Aug 14, 2015 at 03:09:44PM +0200, Christophe Fergeau wrote: > Hey, > > On Fri, Aug 14, 2015 at 01:54:59PM +0100, Daniel P. Berrange wrote: > > On Fri, Aug 14, 2015 at 02:47:15PM +0200, Christophe Fergeau wrote: > > > Before commit b1ea7b79e1, it was possible to start with -spice > > > disable-ticketing, and then use the "set_password spice" command to > > > enable ticketing with SPICE. Since commit b1ea7b79e1 this is no longer > > > possible as qemu_spice_set_ticket() will return an error unless the > > > 'auth' type is "spice". When ticketing is disabled, 'auth' is "none" so > > > the attempt to set password fails. > > > > > > This commit allows to call qemu_spice_set_ticket() when 'auth' is "none" > > > and changes 'auth' to "spice" when this happens. > > > > IMHO we should not be changing the authentication method as a side > > effect of trying to set the password. > > > > If app has disabled ticketing, it should remain disabled and the > > set password call is right to return an error. > > > > In general I agree with you. However in this case, this used to be > working until ~1 year ago, and this change of behaviour caused a bug in > oVirt (oVirt side is being fixed). This is why I sent this patch. > > The intent of commit b1ea7b seems to be to prevent > qemu_spice_set_passwd() from being called when SASL is used, and does > not mention at all whether preventing going from auth being "none" to > "spice" is intentional. > > If this change of behaviour was an intentional bug fix, and if we are > fine with asking for oVirt changes for this, then I'm ok with dropping > this patch.
Hmm, is oVirt using this via libvirt ? If so, I guess we have to fix it, as that would be a break in current usage. 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 :|