It seems to be the case that libvdpau-va-gl 0.4.2-1 is reporting only a subset of the decoding capabilities supported by libva-intel-driver 1.8.3-1. Am I correct in this interpretation of the results?
If so, could you provide an outline of what work/code might be necessary to improve libvdpau-va-gl featureset on Kabylake?
VA-API and VDPAU are different, and every codec requires additional translation code. That translation was only implemented for H.264, partially. Partial implementation is still sufficient for most of H.264 video.
It seems to be the case that
libvdpau-va-gl 0.4.2-1
is reporting only a subset of the decoding capabilities supported bylibva-intel-driver 1.8.3-1
. Am I correct in this interpretation of the results?If so, could you provide an outline of what work/code might be necessary to improve
libvdpau-va-gl
featureset on Kabylake?vainfo
:vdpauinfo
: