I somehow manage to screw up applying Laurent's patch in 
eca93e28c256:
"drm: Check in setcrtc if the primary plane supports the fb pixel
format". It was a conflict with
commit 
3461b30b3e171e16498f3d7bc59ab703aec475c8
Author: Daniel Vetter <daniel.vetter@ffwll.ch>
Date:   Thu Mar 5 10:32:44 2015 +0100
    drm/plane-helper: unexport drm_primary_helper_create_plane
and I just didn't check that the solution from wiggle made sense.
Cc: Dan Carpenter <dan.carpenter@oracle.com>
Cc: laurent.pinchart@ideasonboard.com
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Acked-by: laurent.pinchart@ideasonboard.com
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
 
        if (primary == NULL) {
                DRM_DEBUG_KMS("Failed to allocate primary plane\n");
                return NULL;
-               /*
-                * Remove the format_default field from drm_plane when dropping
-                * this helper.
-                */
-               primary->format_default = true;
        }
 
+       /*
+        * Remove the format_default field from drm_plane when dropping
+        * this helper.
+        */
+       primary->format_default = true;
+
        /* possible_crtc's will be filled in later by crtc_init */
        ret = drm_universal_plane_init(dev, primary, 0,
                                       &drm_primary_helper_funcs,