

Then I needed to apply patches from second repo to make it work on H3. I copied relevant files which were only in FFmpeg-Cedrus, updated different files and created a new branch cedrus264 in mainline ffmpeg. > OBJS-$(CONFIG_CEDRUS264_ENCODER) += cedrus264.o arm/sunxi/ve.o git FFmpeg/libavcodec/Makefile FFmpeg-Cedrus/libavcodec/Makefile Only in FFmpeg-Cedrus/libavcodec: cedrus264.c-ORGĭiff -r -x '*.pc' -x. Only in FFmpeg-Cedrus/libavcodec: cedrus264.c

Only in FFmpeg-Cedrus/libavcodec/arm/sunxi: ve.h-ORG Only in FFmpeg-Cedrus/libavcodec/arm/sunxi: ve.h Only in FFmpeg-Cedrus/libavcodec/arm/sunxi: ve.c-ORG Only in FFmpeg-Cedrus/libavcodec/arm/sunxi: ve.c

Only in FFmpeg-Cedrus/libavcodec/arm/sunxi: h264enc.h Only in FFmpeg-Cedrus/libavcodec/arm/sunxi: h264enc.c > REGISTER_ENCODER(CEDRUS264, cedrus264) git FFmpeg/libavcodec/allcodecs.c FFmpeg-Cedrus/libavcodec/allcodecs.c Only in FFmpeg-Cedrus: ffmpeg_2-1_bĭiff -r -x '*.pc' -x. vidstabutils: improve documentationĭiff showed only some differences – this is the base version: $ diff -r -x *.pc -x. Until I found this version: HEAD is now at 8495c6086d. avcodec/aacdec_fixed: Fix multiple runtime error: shift exponent 127 is too large for 32-bit type 'int' avcodec/lagarith: Fix runtime error: left shift of 1 by 31 places cannot be represented in type 'int' So I started to traverse back from master by following command: $ git checkout HEAD~1 & git describe -tags -match N The compiled binary from that repo showed “N-65839-g156858b” but commit 156858b doesn’t exist in ffmpeg main tree. For this I needed to find out the version it was forked from. So I needed to find out, what is changed compared to mainline ffmpeg. Next I wanted to port it to recent ffmpeg but the first repo was just commited files without proper git history. In this repo I’ve found another fork which states compatibility with Allwinner H3. There is some information on linux sunxi wiki on this topic ( ), which say that it’s not something straightforward and opensource.įortunately, there is one page which shortly mentions ffmpeg implementation which interested me. One thing I’ve always been interested in is HW encoding. Last days I’ve been experimenting with various features of OPI.
