On 15/12/16 11:52, Anton Khirnov wrote: > Partially based on a patch by Timo Rothenpieler <[email protected]>. > Additional scaling list handling fix by Jun Zhao <[email protected]>. > --- > Changelog | 2 +- > configure | 3 + > libavcodec/Makefile | 1 + > libavcodec/allcodecs.c | 1 + > libavcodec/hevcdec.c | 6 +- > libavcodec/vaapi_decode.c | 1 + > libavcodec/vaapi_hevc.c | 439 > ++++++++++++++++++++++++++++++++++++++++++++++ > 7 files changed, 451 insertions(+), 2 deletions(-) > create mode 100644 libavcodec/vaapi_hevc.c
Tested with: * i965 on Skylake (8 bit working; fate passes except for known-problematic streams). * i965 on Kaby Lake (8 and 10 bit working; fate passes as Skylake). * mesa/gallium on Polaris (8 and 10 bit mostly working, output looks ok (10 bit dithered for 8 bit output); a lot of fate fails, though). So, LGTM (with <https://lists.libav.org/pipermail/libav-devel/2016-December/081321.html> and <https://lists.libav.org/pipermail/libav-devel/2016-December/081425.html> applied to fix the segfaults). Thanks, - Mark _______________________________________________ libav-devel mailing list [email protected] https://lists.libav.org/mailman/listinfo/libav-devel
