On Thu, Feb 21 2019, Brad Smith <[email protected]> wrote: > On 2/21/2019 2:34 PM, Jeremie Courreges-Anglas wrote: > >> On Thu, Feb 21 2019, Brad Smith <[email protected]> wrote: >>> Update API usage to deal with current vs new FFmpeg API. >> Builds fine on amd64. One nit, >> >> [...] >> >>> Index: patches/patch-modules_codec_avcodec_avcommon_compat_h >>> =================================================================== >>> RCS file: patches/patch-modules_codec_avcodec_avcommon_compat_h >>> diff -N patches/patch-modules_codec_avcodec_avcommon_compat_h >>> --- /dev/null 1 Jan 1970 00:00:00 -0000 >>> +++ patches/patch-modules_codec_avcodec_avcommon_compat_h 21 Feb 2019 >>> 17:04:13 -0000 >>> @@ -0,0 +1,19 @@ >>> +$OpenBSD$ >>> + >>> +Update for newer FFmpeg API. >>> + >>> +Index: modules/codec/avcodec/avcommon_compat.h >>> +--- modules/codec/avcodec/avcommon_compat.h.orig >>> ++++ modules/codec/avcodec/avcommon_compat.h >>> +@@ -36,6 +36,11 @@ >>> + ( (LIBAVCODEC_VERSION_MICRO < 100 && LIBAVCODEC_VERSION_INT >= >>> AV_VERSION_INT( a, b, c ) ) || \ >>> + (LIBAVCODEC_VERSION_MICRO >= 100 && LIBAVCODEC_VERSION_INT >= >>> AV_VERSION_INT( a, d, e ) ) ) >>> + >>> ++#ifndef FF_MAX_B_FRAMES >>> ++# define FF_MAX_B_FRAMES 16 // FIXME: remove this >>> ++#endif >> Should this be committed too? > > Yes, that symbol goes away with newer FFmpeg.
ack. Lightly tested with a few movies, diff committed. -- jca | PGP : 0x1524E7EE / 5135 92C1 AD36 5293 2BDF DDCC 0DFA 74AE 1524 E7EE
