On Mon, Sep 16, 2013 at 7:56 PM, Ville Syrjälä <ville.syrjala@xxxxxxxxxxxxxxx> wrote: > Oh and now that vactive is actually part of the framebuffer as well, we > need to be more careful in the kernel how we adjust the mode. I can't > recall if we have special hardware needs wrt. the vertical timings, but > if we do we should probably review them all to avoid adjustments that > would cause issues with stereo modes. I think we need to either adjust the entire mode before handing it to userspace or if that's not possible we need to reject a modeset if the implicit vactive/vblank in the side-by-side framebuffer doesn't fit. On X even that shouldn't be an issue since the ddx could just duplicate this knowledge, more fun would be wayland or similar generic display servers. I'd opt to solve this when it's a real problem though and not before. -Daniel -- Daniel Vetter Software Engineer, Intel Corporation +41 (0) 79 365 57 48 - http://blog.ffwll.ch _______________________________________________ dri-devel mailing list dri-devel@xxxxxxxxxxxxxxxxxxxxx http://lists.freedesktop.org/mailman/listinfo/dri-devel
