05:08 <hunderteins> hello

05:12 <hunderteins> to whoever had his hands in the hw accel video playback: Thank you very much.

08:03 <stdint> hunderteins, you are welcome

08:17 <Ke> hunderteins: what does this mean, can I get xvideo without blobs somehow?

08:17 <Ke> or just decode

09:11 <hunderteins> Ke: no sorry .. with blobs . very efficient and easy on the cpu

16:01 <Myy> Meow

16:04 <Myy> phh, Do you have some experience with the DMA MMAP interface ? :3

16:05 <phh> uh nope

16:05 <phh> I've seen few implementations, and it always seemed easy, but that's all i can tell :P

16:06 <Myy> Ah too bad. I'm having some bugs with the interface. While I seem to pass the rights arguments, the mapped memory is not the expected one. Basically I want to map a currently DMA-ready empty buffer, but when I read the mmap'ed buffer I got my logs instead... :3

17:50 <Myy> Ok, it seems that using dma_common_mmap was a bad idea. Using dma_mmap_attrs and dma_mmap_coherent, I got my buffer mmapped correctly.

17:51 <Myy> Err, dma_mmap_attrs or dma_mmap_coherent (which is an alias for mmap_attrs with the last argument set to 0)

21:09 <Myy> stdint, ayaka, Does the VPU base register (0) contains anything that could help diagnosing decoding issues ?

21:09 <Myy> On RK3288 systems

21:20 <Myy> phh, Hey, do you know any software that's able to decode a single H264 raw frame with no forward/backward references ?

21:34 <Myy> Well, it seems that the status if provided in register 1 and this register returns no error, so maybe I'm just decoding a pitch black frame