On Tue, Jun 03, 2014 at 04:49:25PM +0200, Daniel Vetter wrote:
On Tue, Jun 3, 2014 at 4:44 PM, Matt Roper matthew.d.roper@intel.com wrote:
On Tue, Jun 03, 2014 at 03:38:56PM +0200, Daniel Vetter wrote:
The drm core shouldn't depend upon any helpers, and we make sure this doesn't accidentally happen by moving them into the helper-only drm_kms_helper.ko module.
Cc: Matt Roper matthew.d.roper@intel.com Signed-off-by: Daniel Vetter daniel.vetter@ffwll.ch
Are there any KMS drivers that don't 'select DRM_KMS_HELPER' in their Kconfig today? From a quick grep + cscope it looks like radeon and vmwgfx don't select the helper library.
Since drm_crtc_init() is part of the helper library now (since it creates a helper-created primary plane), I think those drivers either need to select the helper in their Kconfig to get a proper build, or they need to be updated to provide their own primary planes and call drm_crtc_init_with_planes(), right?
Radeon definitely uses the crtc helpers, so should have the dependency somewhere. vmwgfx might indeed be broken now a bit. I'll wait for the 0-day kernel test build farm to confirm that ;-) -Daniel
Yeah, you're right; now that I look closer radeon pulls it in from the main drm directory Kconfig and only has ums stuff in its driver directory Kconfig. But I think vmwgfx probably still needs to select the helper library.
If vmwgfx gets a kconfig update to pull in the helper library, then this is Reviewed-by: Matt Roper matthew.d.roper@intel.com
Matt
-- Daniel Vetter Software Engineer, Intel Corporation +41 (0) 79 365 57 48 - http://blog.ffwll.ch