> > I think we should probably report an error if xres / yres > > are set at the global level and also set against any individual > > output, so the two approaches are mutually exclusive.
> ACK - fixed for the next patch. Doh, I ran into an error. Seems like if I don't set xres/yres, then it gets defaulted to 1280x800. https://gitlab.com/qemu-project/qemu/-/blob/master/include/hw/virtio/virtio-gpu.h#L175-176 So...could we just get away with documenting that if you set xres/yres and outputs[0].xres/yres, then the later is going to overwrite the former? I'm not sure how to detect if xres/yres has been set globally to 1280x800, or if we are just getting the defaults. What do you think? On Wed, Aug 27, 2025 at 12:44 PM Daniel P. Berrangé <berra...@redhat.com> wrote: > On Wed, Aug 27, 2025 at 11:48:34AM -0400, Andrew Keesler wrote: > > Sending again (replying all this time). > > > > > IIUC from the current code, xres/yres are only set against the > > > first head. The 2nd and later heads are left undefined by the > > > virtio-gpu-base code at least - I'm unclear if something else > > > in QEMU will fill in defaults later, or if they set to 0x0. > > > > That is correct - xres/yres are only set against the first head in the > > current code. The wording in my commit message (and cover letter) was > > misleading. I will fix that to say that "If no > > virtio_gpu_base_conf.outputs are provided, > > virtio_gpu_base_conf.xres/virtio_gpu_base_conf.yres will still be > > respected _for the first head_ for backwards compatibility". > > > > The only way I could get QEMU to fill in xres/yres for a 2nd and later > > head was to trigger virtio_gpu_ops.ui_info via a VNC client. > > > > > Is it relevant to set xres/yres on outputs, even when they are > > > not (yet) enabled ? Perhaps we should have an explicit 'enabled: bool' > > > property in the 'outputs' struct ? > > > > Maybe you might want to set xres/yres on an output, but not enable it > > yet? I don't have any concrete examples of when you might want to do > > that, maybe you have some examples? > > No, I don't have an example. > > > I feel like I could see a user setting xres/yres on an output, > > forgetting to set enabled on that output, and then being confused why > > their head is blank. Because of this, my vote would be to default to > > enabling an output when it has configuration. I am easily swayed by > > your guidance, though. > > Ok, lets just document that setting xres/yres will cause the > output to be automatically enabled. We can always retrofit > an 'enabled' property at a later date if someone arrives with > a use case for setting res without enabling the output. > > With 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 :| > >
