Summary: | media-video/ffmpeg2theora-0.29-r2 - src/ffmpeg2theora.h:65:5: error: unknown type name 'ReSampleContext' | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | Reto Gantenbein (ganto) <reto.gantenbein> |
Component: | Current packages | Assignee: | Gentoo Media-video project <media-video> |
Status: | RESOLVED FIXED | ||
Severity: | normal | CC: | pacho |
Priority: | Normal | Keywords: | PATCH |
Version: | unspecified | ||
Hardware: | All | ||
OS: | Linux | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- | |
Bug Depends on: | |||
Bug Blocks: | 509294, 574788 | ||
Attachments: |
media-video/ffmpeg2theora-0.29-r2 build.log against media-video/libav-11.2
Patch to fix the "unknown type name 'ReSampleContext'" error patch to fix CODEC_ID* bug |
Description
Reto Gantenbein (ganto)
2015-02-12 22:19:15 UTC
Created attachment 396288 [details]
media-video/ffmpeg2theora-0.29-r2 build.log against media-video/libav-11.2
Bug seems to be fixed in upstream repository. See Debian BTS #739237 (https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=739237) for more information. Attached the patch mentioned in the Debian bug (slightly adjusted to properly apply). The first build error because of the unknown 'ReSampleContext' type seems to be fixed with this. However, the "error: 'CODEC_ID_TEXT' undeclared" (and so on) errors remain. Created attachment 396300 [details, diff]
Patch to fix the "unknown type name 'ReSampleContext'" error
Created attachment 398800 [details, diff]
patch to fix CODEC_ID* bug
this ebuild patch fixes the CODEC_ID* bug
They must have changed the API quite a bit... After adding both "fixes", I get another build error: src/avinfo.c: In function 'json_codec_info': src/avinfo.c:181:5: warning: 'codec_name' is deprecated (declared at /usr/include/libavcodec/avcodec.h:1065) [-Wdeprecated-declarations] } else if (enc->codec_name[0] != '\0') { ^ src/avinfo.c:182:9: warning: 'codec_name' is deprecated (declared at /usr/include/libavcodec/avcodec.h:1065) [-Wdeprecated-declarations] codec_name = enc->codec_name; ^ src/avinfo.c: In function 'json_stream_format': src/avinfo.c:327:28: error: 'AVStream' has no member named 'r_frame_rate' st->r_frame_rate.num, st->r_frame_rate.den); ^ src/avinfo.c:327:50: error: 'AVStream' has no member named 'r_frame_rate' st->r_frame_rate.num, st->r_frame_rate.den); ^ src/avinfo.c: In function 'gen_oshash': src/avinfo.c:448:18: warning: ignoring return value of 'fread', declared with attribute warn_unused_result [-Wunused-result] fread(buffer1, used, 8, file); ^ src/avinfo.c:450:18: warning: ignoring return value of 'fread', declared with attribute warn_unused_result [-Wunused-result] fread(buffer1, 8192, 8, file); ^ src/avinfo.c:452:18: warning: ignoring return value of 'fread', declared with attribute warn_unused_result [-Wunused-result] fread(&buffer1[8192], 8192, 8, file); ^ scons: *** [src/avinfo.o] Error 1 scons: *** [src/ffmpeg2theora.o] Error 1 scons: building terminated because of errors. I think only patching build failures is not really the right solution for this bug. Someone must probably go through the code and adjust it to the new libav API. FWIW -the -r1 enild triggers this (ffmpeg-3 ?) issue: src/ffmpeg2theora.c: In function 'ff2theora_init': src/ffmpeg2theora.c:229:25: error: 'PIX_FMT_YUV420P' undeclared (first use in this function) this->pix_fmt = PIX_FMT_YUV420P; ^ There's now ffmpeg2theora-0.30 in the tree, which seems to work with ffmpeg3. So can we close this bug? (Probably should open a stabilization bug for it as well) (In reply to Hanno Boeck from comment #8) > There's now ffmpeg2theora-0.30 in the tree, which seems to work with > ffmpeg3. So can we close this bug? (Probably should open a stabilization bug > for it as well) yep |