On Tue, 18 Sep 2012, Gerd Hoffmann wrote: > On 09/18/12 16:10, Stefano Stabellini wrote: > > On Tue, 18 Sep 2012, Gerd Hoffmann wrote: > >> When adding DisplayChangeListeners the set_mouse and cursor_define > >> callbacks have been left in DisplayState for some reason. Fix it. > >> > >> Signed-off-by: Gerd Hoffmann <kra...@redhat.com> > > > > This patch is good. > > The one thing I don't like is dpy_cursor_define_supported, because it > > enforces the idea that we cannot register/deregister > > DisplayChangeListeners at run time. Theoretically a new > > DisplayChangeListener that support cursor_define could show up at any > > time. > > Depends on how it is used. > > qxl is fine, it uses dpy_cursor_define_supported only to skip some work > in case nobody cares anyway. > > vmware_vga uses it to signal the capability to the guest, which > obviously doesn't fly with displaysurfaces coming & going ...
yeah.. that is what I am talking about. Maybe adding a comment on top the "offending code" in vmware_vga could suffice for now.
