On Thu, Apr 2, 2020 at 12:57 PM Daniel Stone daniel@fooishbar.org wrote:
On Thu, 2 Apr 2020 at 11:40, Laurent Pinchart laurent.pinchart+renesas@ideasonboard.com wrote:
The R-Car DU driver creates a zpos property, ranging from 1 to 7, for all the overlay planes, but leaves the primary plane without a zpos property. The DRM/KMS API doesn't clearly specify if this is acceptable, of it the property is mandatory for all planes when exposed for some of the planes. Nonetheless, weston v8.0 has been reported to have trouble handling this situation.
Yeah. It didn't even occur to me/us that someone would do that, to be honest. We need to have zpos information for all planes that we're using in order for zpos to be at all meaningful, and we can't exactly avoid using the primary plane. Without knowing the primary plane's zpos, we can't know if the overlays are actually overlays or in fact underlays.
Maybe we need to clarify docs that if you expose zpos, then you should expose it on all planes (opting for immutable zpos where it can't be adjusted)? Care to type up that quick doc patch please? -Daniel
The DRM core offers support for immutable zpos properties. Creating an immutable zpos property set to 0 for the primary planes seems to be a good way forward, as it shouldn't introduce any regression, and can fix the issue. Do so.
Perfect. We support immutable properties entirely well, we just need to know about them.
Reported-by: Kuninori Morimoto kuninori.morimoto.gx@renesas.com Signed-off-by: Laurent Pinchart laurent.pinchart+renesas@ideasonboard.com
Reviewed-by: Daniel Stone daniels@collabora.com
Cheers, Daniel _______________________________________________ dri-devel mailing list dri-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/dri-devel