Hi Daniel, Thank you for the patch. On Mon, May 06, 2019 at 04:46:29PM +0200, Daniel Vetter wrote: > It's mandatory and considered core state since ioctls rely on this > working. > > Thanks to Laurent for pointing out this gap. > > v2: Clarify to "atomic drivers" only. > > Cc: Laurent Pinchart <laurent.pinchart@xxxxxxxxxxxxxxxx> > Cc: Sean Paul <sean@xxxxxxxxxx> > Signed-off-by: Daniel Vetter <daniel.vetter@xxxxxxxxx> > --- > include/drm/drm_connector.h | 4 ++++ > 1 file changed, 4 insertions(+) > > diff --git a/include/drm/drm_connector.h b/include/drm/drm_connector.h > index 02a131202add..f43f40d5888a 100644 > --- a/include/drm/drm_connector.h > +++ b/include/drm/drm_connector.h > @@ -517,6 +517,10 @@ struct drm_connector_state { > * Used by the atomic helpers to select the encoder, through the > * &drm_connector_helper_funcs.atomic_best_encoder or > * &drm_connector_helper_funcs.best_encoder callbacks. How about updating this part as well ? "Used by both the DRM core and the atomic helpers to select the encoder (through the &drm_connector_helper_funcs.atomic_best_encoder), access it and report it to userspace (through the GETCONNECTOR and GETENCODER ioctls). This field shall be set by all atomic drivers, either directly or through atomic helpers." > + * > + * NOTE: Atomic drivers must fill this out (either themselves or through > + * helpers), for otherwise the GETCONNECTOR and GETENCODER IOCTLs will > + * not return correct data to userspace. > */ > struct drm_encoder *best_encoder; > -- Regards, Laurent Pinchart _______________________________________________ dri-devel mailing list dri-devel@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/dri-devel