On 23/05/17 16:07, Laurent Pinchart wrote: > Hi Tomi, > > Thank you for the patch. > > On Wednesday 17 May 2017 10:56:40 Tomi Valkeinen wrote: >> At the moment the dispc driver uses a custom enum for rotation. Change >> it to use the DRM's DRM_ROTATE_*. >> >> Note that mirroring is at the moment handled as a separate boolean in >> the dispc driver, so we only use the DRM_ROTATE_* values. >> >> Note, DSS HW uses clockwise rotation, DRM counter-clockwise. > > I've tried to review this patch by checking how the driver converts from DRM > rotation to DSS rotation, but unless I'm mistaken the only entry point to the > DSS where rotation is passed is through the .ovl_setup() operation, and info- >> rotation doesn't seem to ever be set to a non-zero value. Am I missing > something or is the rotation code in DSS actually not needed ? > > It's hard to review if the driver does the right thing by checking how input > values are handled before and after the patch when the only input value ever > set is 0 :-) However, I see no issue in the patch itself, so With tiler, in many cases (RGB) the DSS doesn't have to care about the angle. But with YUV422 (in the following patches) we do need that information in the DSS. DMA and VRFB rotation used the rotation value, if I recall right. Also, omapdrm was broken and never passed the angle (fixed in later patch). So, for this patch, please ignore the all the oddness with rotation code, just look at changing the custom enum to drm's =). Tomi
Attachment:
signature.asc
Description: OpenPGP digital signature
_______________________________________________ dri-devel mailing list dri-devel@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/dri-devel
