Yukimasa Sugizaki

I would send a gentle ping to developers.

It should be noted that the example is for VC6, and the upstream VC4 KMS driver does not support QPU execution due to the lack of memory protection. See doe300/VC4CL#51.

It seems that [`vc_dispmanx_resource_write_data_handle`](https://github.com/raspberrypi/userland/blob/master/interface/vmcs_host/vc_vchi_dispmanx.c#L486) and [`vc_dispmanx_resource_write_data`](https://github.com/raspberrypi/userland/blob/master/interface/vmcs_host/vc_vchi_dispmanx.c#L390) are both doing bulk transmitting, so they aren't able to do zero-copy from MMAL to dispmanx? (Maybe I'm wrong here.)

Excuse me for wandering from the subject, but I found that since MMAL's `vc.ril.video_render` supports RGBA and zero-copy rendering, `video_render` can be used as a zero-copy rendering API instead of...

Yes. I want VPU to do alpha blending if possible.

Would you please tell me where the patch is and I'll be able to test it.

Thanks. By the way, what is your final goal of implementing V4L2 multimedia driver though there already is MMAL which works well?

Oh, that's great! Please continue rocking!

Yeah, I ended up with the same result, and the kernel sometimes crashes when executed the app multiple times

Thank you for your interest! Currently, I've been focusing on VC6 because VC6 is simply faster than VC4. In addition, we, [Idein Inc.](https://idein.jp/), have a private ML library that supports...