[Spice-devel] [PATCH spice-server 1/2] gl: fix client mouse mode

[Frediano Ziglio]

> 
> Hi,
> 
> On Tue, Jun 13, 2017 at 05:18:41AM -0400, Marc-André Lureau wrote:
> > > > ----- Original Message -----
> > > > > Since 2.8, QEMU now longer creates QXL primary surfaces when
> > > > > using GL.  This change broke client-side mouse mode, because
> > > > > Spice server relies on primary surface conditions.
> > > > > 
> > > > > When GL is enabled, use GL scanout informations.
> > > > > Mouse mode is always client when GL surfaces are used.
> > > > > 
> > > > > This patch and most of the message are based on a patch from
> > > > > Marc-André Lureau, just moving responsibility from reds to RedQxl.
> > > > > 
> > > > > Signed-off-by: Frediano Ziglio <fziglio at redhat.com>
> > > > > ---
> > > > >  server/red-qxl.c | 7 +++++++
> > > > >  1 file changed, 7 insertions(+)
> > > > > 
> > > > > diff --git a/server/red-qxl.c b/server/red-qxl.c
> > > > > index 53f3338..896bb76 100644
> > > > > --- a/server/red-qxl.c
> > > > > +++ b/server/red-qxl.c
> > > > > @@ -883,6 +883,13 @@ void spice_qxl_gl_scanout(QXLInstance *qxl,
> > > > >      /* FIXME: find a way to coallesce all pending SCANOUTs */
> > > > >      dispatcher_send_message(qxl_state->dispatcher,
> > > > >                              RED_WORKER_MESSAGE_GL_SCANOUT,
> > > > >                              &payload);
> > > > > +
> > > > > +    qxl_state->x_res = width;
> > > > > +    qxl_state->y_res = height;
> > > > > +    qxl_state->use_hardware_cursor = TRUE;
> > > > > +    qxl_state->primary_active = TRUE;
> > > >
> > > > Are those values restored when switching back to QXL-2D?
> > > >
> > > >
> > >
> > > What do you mean? Suspend/resume, reboot or other cases?
> >
> > Yes, with older qemu version <2.8, or depending on how the server is
> > used, you may switch freely from/to GL/2D etc.
> >
> > With qemu, this happens before virgl driver is enabled, so across
> > reboots.
> >
> > I am worried about mixing the state of GL/QXL with 2D/QXL.
> 
> Atm, here, on wayland, both patches (1/2 of this series and the previous
> patch from Marc-André) only fix the client mouse mode when the guest
> user is signed in. On reboot the mouse will disappear and it'll keep that
> way when guest user login.
> 
> I recall testing this in the past and it was working just fine and also
> another use give a positive feedback on both solutions.
> 
> Note sure if that matters but libvirt/qemu here is from
> fedora-virt-preview repo.
> 
> To be honest, that could simply be a bug in the client/wayland as IIRC
> teuf was working in something around this... but I find interesting that
> both patches works 100% when the user is logged in.
> 

Did you try with X11 ? Is it only an issue with Wayland?
Did the server send new cursors or not?
What happens if you close the client and open again?

Frediano