September 2024 - ffmpeg-devel

[PATCH 1/3] avformat/iamf_parse: clear padding
by Michael Niedermayer 01 Jan '25

01 Jan '25

Fixes: use of uninitialized value Fixes: 70929/clusterfuzz-testcase-minimized-ffmpeg_dem_IAMF_fuzzer-5931276639469568 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Signed-off-by: Michael Niedermayer <michael(a)niedermayer.cc> --- libavformat/iamf_parse.c | 1 + 1 file changed, 1 insertion(+) diff --git a/libavformat/iamf_parse.c b/libavformat/iamf_parse.c index 296e49157b0..f2b6d4fa518 100644 --- a/libavformat/iamf_parse.c +++ b/libavformat/iamf_parse.c @@ -1076,6 +1076,7 @@ int ff_iamfdec_read_descriptors(IAMFContext *c, AVIOContext *pb, size = avio_read(pb, header, FFMIN(MAX_IAMF_OBU_HEADER_SIZE, max_size)); if (size < 0) return size; + memset(header + size, 0, AV_INPUT_BUFFER_PADDING_SIZE); len = ff_iamf_parse_obu_header(header, size, &obu_size, &start_pos, &type, NULL, NULL); if (len < 0 || obu_size > max_size) { -- 2.45.2

2 5

[RFC 00/13] flvdec/flvenc: add support for enhanced rtmp codecs and multitrack/multichannel
by Timo Rothenpieler 12 Dec '24

12 Dec '24

This is based on the preliminary spec for enhanced rtmp v2: https://veovera.org/docs/enhanced/enhanced-rtmp-v2 The spec is not final, and can still undergo breaking changes, hence this set is purely for comments and review, and not ready to be merged until the final v2 spec is published. There are no samples out in the wild yet, so testing interoperability with other software has not happened yet either. Specially the two other multitrack modes, where multiple tracks are in the same packet, have not been tested at all, since no software can write such files. The set can also be found on GitHub, where ignoring whitespaces makes specially the last patch a lot more readable: https://github.com/BtbN/FFmpeg/tree/enhanced-flv Dennis Sädtler via ffmpeg-devel (2): avformat/flvenc: Implement support for multi-track video avformat/flvdec: Add support for demuxing multi-track FLV Timo Rothenpieler (11): avformat/flvenc: add enhanced audio codecs avformat/flvenc: remove !size check for audio packets avformat/flvdec: add enhanced audio codecs avformat/flvenc: refactor fourcc writing avformat/flvenc: write enhanced rtmp multichannel info for audio with more than two channels avformat/flvdec: parse enhanced rtmp multichannel info avformat/flvenc: add support for writing multi track audio avformat/flvdec: add support for reading multi track audio avformat/rtmpproto: add more enhanced rtmp codecs avformat/flvdec: stop shadowing local variables avformat/flvdec: support all multi-track modes libavformat/flv.h | 21 ++ libavformat/flvdec.c | 654 +++++++++++++++++++++++++++------------- libavformat/flvenc.c | 443 +++++++++++++++++++++------ libavformat/rtmpproto.c | 11 +- 4 files changed, 819 insertions(+), 310 deletions(-) -- 2.43.2

7 28

[PATCH 1/6] avcodec/cbs: sei_3d_reference_displays_info uses length 0 elements
by Michael Niedermayer 11 Dec '24

11 Dec '24

Fixes: 70458/clusterfuzz-testcase-minimized-ffmpeg_BSF_TRACE_HEADERS_fuzzer-5259339779080192 Fixes: Assertion width > 0 && width <= 32 failed at libavcodec/cbs.c:608 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Signed-off-by: Michael Niedermayer <michael(a)niedermayer.cc> --- libavcodec/cbs.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/libavcodec/cbs.c b/libavcodec/cbs.c index b26e39eab4d..dcbc86a5f7d 100644 --- a/libavcodec/cbs.c +++ b/libavcodec/cbs.c @@ -605,7 +605,7 @@ static av_always_inline int cbs_read_unsigned(CodedBitstreamContext *ctx, CBS_TRACE_READ_START(); - av_assert0(width > 0 && width <= 32); + av_assert0(width >= 0 && width <= 32); if (get_bits_left(gbc) < width) { av_log(ctx->log_ctx, AV_LOG_ERROR, "Invalid value at " -- 2.45.2

3 12

[PATCH] avformat/mpegts: correctly skip TP_extra_header in m2ts
by llyyr 07 Dec '24

07 Dec '24

instead of just resyncing and skipping a bunch of TS packets, leading to a loss of frames. Before this, a stray byte with the value of 0x47 in TP_extra_header would throw off the detection of where TS packets start. A typical file that could cause issues would look like this: 00000300: 238f 4780 4750 1110 0000 01e0 0000 84c0 ^^ ^^ The first four bytes here are TP_extra_header and the actual TS packet starts at offset 0x304 FFmpeg would try to read a packet at 0x300 but since nothing skips the 4 byte TP_extra_header, find that the first byte is not 0x47 and immediately go into mpegts_resync, and incorrectly detect the stray 0x47 in the TP_extra_header at 0x302 as the new sync byte. Fix this by correctly skipping the first 4 bytes if the source packet is 192 bytes. Signed-off-by: llyyr <llyyr.public(a)gmail.com> --- libavformat/mpegts.c | 12 +++++++++++- 1 file changed, 11 insertions(+), 1 deletion(-) diff --git a/libavformat/mpegts.c b/libavformat/mpegts.c index c66a1ea6ede0..0d80ad80f1aa 100644 --- a/libavformat/mpegts.c +++ b/libavformat/mpegts.c @@ -2944,6 +2944,12 @@ static int read_packet(AVFormatContext *s, uint8_t *buf, int raw_packet_size, AVIOContext *pb = s->pb; int len; + // 192 bytes source packet that start with a 4 bytes TP_extra_header + // followed by 188 bytes of TS packet. The sync byte is at offset 4, so skip + // the first 4 bytes otherwise we'll end up syncing to the wrong packet. + if (raw_packet_size == TS_DVHS_PACKET_SIZE) + avio_skip(pb, 4); + for (;;) { len = ffio_read_indirect(pb, buf, TS_PACKET_SIZE, data); if (len != TS_PACKET_SIZE) @@ -2966,7 +2972,11 @@ static int read_packet(AVFormatContext *s, uint8_t *buf, int raw_packet_size, static void finished_reading_packet(AVFormatContext *s, int raw_packet_size) { AVIOContext *pb = s->pb; - int skip = raw_packet_size - TS_PACKET_SIZE; + int skip; + if (raw_packet_size == TS_DVHS_PACKET_SIZE) + skip = raw_packet_size - TS_DVHS_PACKET_SIZE; + else + skip = raw_packet_size - TS_PACKET_SIZE; if (skip > 0) avio_skip(pb, skip); } base-commit: e9197db4f7227a341b781c227a0a8f3a99033b5e -- 2.45.1

2 6

[PATCH] lavc/libx265: flag as experimental
by Anton Khirnov 03 Dec '24

03 Dec '24

This encoder leaks and overreads, as can be seen e.g. by running an encode under valgrind with default encoder parameters. This was known upstream since at least 2019 (e.g. bitbucket issue #482) but never fixed until now. Since upstream does not seem to practice basic code hygiene, make sure people do not use this encoder without knowing what they are getting into. --- libavcodec/libx265.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/libavcodec/libx265.c b/libavcodec/libx265.c index 420d0953af..d4511251a5 100644 --- a/libavcodec/libx265.c +++ b/libavcodec/libx265.c @@ -895,7 +895,8 @@ FFCodec ff_libx265_encoder = { .p.id = AV_CODEC_ID_HEVC, .p.capabilities = AV_CODEC_CAP_DR1 | AV_CODEC_CAP_DELAY | AV_CODEC_CAP_OTHER_THREADS | - AV_CODEC_CAP_ENCODER_REORDERED_OPAQUE, + AV_CODEC_CAP_ENCODER_REORDERED_OPAQUE | + AV_CODEC_CAP_EXPERIMENTAL, .p.priv_class = &class, .p.wrapper_name = "libx265", .init = libx265_encode_init, -- 2.39.1

8 15

[PATCH 01/10] tools/target_dec_fuzzer: Add threshold for SRGC
by Michael Niedermayer 02 Dec '24

02 Dec '24

Fixes: Timeout Fixes: 71234/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_SRGC_fuzzer-5098445864501248 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Signed-off-by: Michael Niedermayer <michael(a)niedermayer.cc> --- tools/target_dec_fuzzer.c | 1 + 1 file changed, 1 insertion(+) diff --git a/tools/target_dec_fuzzer.c b/tools/target_dec_fuzzer.c index 81ee4ef1a4e..432e9488168 100644 --- a/tools/target_dec_fuzzer.c +++ b/tools/target_dec_fuzzer.c @@ -296,6 +296,7 @@ int LLVMFuzzerTestOneInput(const uint8_t *data, size_t size) { case AV_CODEC_ID_SMACKAUDIO: maxsamples /= 4096; break; case AV_CODEC_ID_SMACKVIDEO: maxpixels /= 64; break; case AV_CODEC_ID_SNOW: maxpixels /= 128; break; + case AV_CODEC_ID_SRGC: maxpixels /= 128; break; case AV_CODEC_ID_TARGA: maxpixels /= 128; break; case AV_CODEC_ID_TAK: maxsamples /= 1024; break; case AV_CODEC_ID_TGV: maxpixels /= 32; break; -- 2.46.1

4 28

[PATCH 1/2] MAINTAINERS: aacdec_usac seems not actively maintained
by Michael Niedermayer 02 Dec '24

02 Dec '24

My ping on a patch on 21 Jul has no reaction Signed-off-by: Michael Niedermayer <michael(a)niedermayer.cc> --- MAINTAINERS | 1 + 1 file changed, 1 insertion(+) diff --git a/MAINTAINERS b/MAINTAINERS index a513bec72a9..d73536370a0 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -146,6 +146,7 @@ Codecs: 4xm.c [2] Michael Niedermayer 8bps.c Roberto Togni 8svx.c Jaikrishnan Menon + aacdec_usac.c [0] aacenc*, aaccoder.c Rostislav Pehlivanov adpcm.c Zane van Iperen alacenc.c Jaikrishnan Menon -- 2.46.0

3 7

[PATCH 1/6] avutil/frame: add av_frame_remove_side_data_changed
by Niklas Haas 29 Nov '24

29 Nov '24

From: Niklas Haas <git(a)haasn.dev> Many filters modify certain aspects of frame data, e.g. through resizing (vf_*scale* family), color volume mapping (vf_lut*, vf_tonemap*), or possibly others. When this happens, we should strip all frame side data that will no longer be correct/relevant after the operation. For example, changing the image size should invalidate AV_FRAME_DATA_PANSCAN because the crop window (given in pixels) no longer corresponds to the actual image size. For another example, tone-mapping filters (e.g. from HDR to SDR) should strip all of the dynamic HDR related metadata. Since there are a lot of similar with basically similar operations, it make sense to consolidate this stripping logic into a common helper function. I decided to put it into libavutil as it may be useful for API users as well, who often have their own internal processing and filtering. --- doc/APIchanges | 3 +++ libavutil/frame.c | 31 +++++++++++++++++++++++++++++++ libavutil/frame.h | 14 ++++++++++++++ libavutil/version.h | 2 +- 4 files changed, 49 insertions(+), 1 deletion(-) diff --git a/doc/APIchanges b/doc/APIchanges index 0566fcdcc5..26af725528 100644 --- a/doc/APIchanges +++ b/doc/APIchanges @@ -2,6 +2,9 @@ The last version increases of all libraries were on 2024-03-07 API changes, most recent first: +2024-04-xx - xxxxxxxxxx - lavu 59.17.100 - frame.h + Add av_frame_remove_side_data_changed(). + 2024-04-24 - 8616cfe0890 - lavu 59.16.100 - opt.h Add AV_OPT_SERIALIZE_SEARCH_CHILDREN. diff --git a/libavutil/frame.c b/libavutil/frame.c index 0775e2abd9..d5482c258e 100644 --- a/libavutil/frame.c +++ b/libavutil/frame.c @@ -1015,6 +1015,37 @@ void av_frame_remove_side_data(AVFrame *frame, enum AVFrameSideDataType type) remove_side_data(&frame->side_data, &frame->nb_side_data, type); } +static const struct { + enum AVFrameSideDataType type; + int aspect; +} side_data_aspects[] = { + { AV_FRAME_DATA_PANSCAN, AV_FRAME_CHANGED_SIZE }, + { AV_FRAME_DATA_MOTION_VECTORS, AV_FRAME_CHANGED_SIZE }, + { AV_FRAME_DATA_MASTERING_DISPLAY_METADATA, AV_FRAME_CHANGED_COLOR_VOLUME }, + + { AV_FRAME_DATA_SPHERICAL, AV_FRAME_CHANGED_SIZE }, + { AV_FRAME_DATA_CONTENT_LIGHT_LEVEL, AV_FRAME_CHANGED_COLOR_VOLUME }, + { AV_FRAME_DATA_ICC_PROFILE, AV_FRAME_CHANGED_COLOR_VOLUME }, + { AV_FRAME_DATA_DYNAMIC_HDR_PLUS, AV_FRAME_CHANGED_COLOR_VOLUME }, + { AV_FRAME_DATA_REGIONS_OF_INTEREST, AV_FRAME_CHANGED_SIZE }, + { AV_FRAME_DATA_DETECTION_BBOXES, AV_FRAME_CHANGED_SIZE }, + { AV_FRAME_DATA_DOVI_RPU_BUFFER, AV_FRAME_CHANGED_COLOR_VOLUME }, + { AV_FRAME_DATA_DOVI_METADATA, AV_FRAME_CHANGED_COLOR_VOLUME }, + { AV_FRAME_DATA_DYNAMIC_HDR_VIVID, AV_FRAME_CHANGED_COLOR_VOLUME }, + { AV_FRAME_DATA_VIDEO_HINT, AV_FRAME_CHANGED_SIZE }, +}; + +void av_frame_remove_side_data_changed(AVFrame *frame, int changed_aspects) +{ + if (!changed_aspects) + return; + + for (int i = 0; i < FF_ARRAY_ELEMS(side_data_aspects); i++) { + if (changed_aspects & side_data_aspects[i].aspect) + av_frame_remove_side_data(frame, side_data_aspects[i].type); + } +} + const AVSideDataDescriptor *av_frame_side_data_desc(enum AVFrameSideDataType type) { unsigned t = type; diff --git a/libavutil/frame.h b/libavutil/frame.h index 60bb966f8b..7e07ecf91f 100644 --- a/libavutil/frame.h +++ b/libavutil/frame.h @@ -983,6 +983,20 @@ AVFrameSideData *av_frame_get_side_data(const AVFrame *frame, */ void av_frame_remove_side_data(AVFrame *frame, enum AVFrameSideDataType type); +/** + * Flags for stripping side data based on changed aspects. + */ +enum { + /* Video only */ + AV_FRAME_CHANGED_SIZE = 1 << 0, ///< changed dimensions / crop + AV_FRAME_CHANGED_COLOR_VOLUME = 1 << 1, ///< changed color volume +}; + +/** + * Remove all relevant side data after a corresponding change to the frame + * data, based on the given bitset of frame aspects. + */ +void av_frame_remove_side_data_changed(AVFrame *frame, int changed_aspects); /** * Flags for frame cropping. diff --git a/libavutil/version.h b/libavutil/version.h index ea289c406f..3b5a2e7aaa 100644 --- a/libavutil/version.h +++ b/libavutil/version.h @@ -79,7 +79,7 @@ */ #define LIBAVUTIL_VERSION_MAJOR 59 -#define LIBAVUTIL_VERSION_MINOR 16 +#define LIBAVUTIL_VERSION_MINOR 17 #define LIBAVUTIL_VERSION_MICRO 100 #define LIBAVUTIL_VERSION_INT AV_VERSION_INT(LIBAVUTIL_VERSION_MAJOR, \ -- 2.44.0

3 11

[PATCH] tests/fate/filter-video: don't convert owdenoise test to mpeg range
by Ramiro Polla 28 Nov '24

28 Nov '24

--- The new owdenoise-scenwin-jpeg.raw reference file is available at https://0x0.st/XguT.raw tests/fate/filter-video.mak | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/tests/fate/filter-video.mak b/tests/fate/filter-video.mak index 5b8a294afd..0f351ea404 100644 --- a/tests/fate/filter-video.mak +++ b/tests/fate/filter-video.mak @@ -2,8 +2,8 @@ VIDEO_FILTER = $(call ALLYES, $(1:%=%_FILTER) $(2) FILE_PROTOCOL IMAGE2_DEMUXER PGMYUV_DECODER RAWVIDEO_ENCODER NUT_MUXER MD5_PROTOCOL) FATE_FILTER_SAMPLES-$(call FILTERDEMDECENCMUX, PERMS OWDENOISE TRIM SCALE, SMJPEG, MJPEG, RAWVIDEO, RAWVIDEO, PIPE_PROTOCOL) += fate-filter-owdenoise-sample -fate-filter-owdenoise-sample: CMD = ffmpeg -auto_conversion_filters -idct simple -i $(TARGET_SAMPLES)/smjpeg/scenwin.mjpg -vf "trim=duration=0.5,perms=random,owdenoise=10:20:20:enable=not(between(t\,0.2\,1.2))" -an -f rawvideo -color_range mpeg - -fate-filter-owdenoise-sample: REF = $(SAMPLES)/filter-reference/owdenoise-scenwin.raw +fate-filter-owdenoise-sample: CMD = ffmpeg -auto_conversion_filters -idct simple -i $(TARGET_SAMPLES)/smjpeg/scenwin.mjpg -vf "trim=duration=0.5,perms=random,owdenoise=10:20:20:enable=not(between(t\,0.2\,1.2))" -an -f rawvideo - +fate-filter-owdenoise-sample: REF = $(SAMPLES)/filter-reference/owdenoise-scenwin-jpeg.raw fate-filter-owdenoise-sample: CMP_TARGET = 1 fate-filter-owdenoise-sample: FUZZ = 3539 fate-filter-owdenoise-sample: CMP = oneoff -- 2.30.2

3 6

[PATCH v13 0/8] [WIP] webp: add support for animated WebP decoding
by Thilo Borgmann 27 Nov '24

27 Nov '24

Marked WIP because we'd want to introduce private bsf's first; review welcome before that though v13 Switched to new ProgressFrame API Support for pixel format changes in animations Propagate MIN_DELAY in case of invalid delay in animations Added more tests for animations with pixel format changes v12 VP8 decoder decoupled again The whole animated sequence goes into one packet The (currently public) bitstream filter splits animations up into non-conformant packets Now with XMP metadata support (as string, like MOV) Patch 5/8 is still there for making changes in lavc/webp reviewable but shall be stashed when pushing. -Thilo Josef Zlomek (2): libavcodec/webp: add support for animated WebP libavformat/webp: add WebP demuxer Thilo Borgmann via ffmpeg-devel (6): avcodec/webp: remove unused definitions avcodec/webp: separate VP8 decoding avcodec/bsf: Add awebp2webp bitstream filter avcodec/webp: make init_canvas_frame static fate: add test for animated WebP avcodec/webp: export XMP metadata Changelog | 2 + configure | 1 + doc/demuxers.texi | 28 + libavcodec/bitstream_filters.c | 1 + libavcodec/bsf/Makefile | 1 + libavcodec/bsf/awebp2webp.c | 353 +++++++ libavcodec/codec_desc.c | 3 +- libavcodec/webp.c | 983 ++++++++++++++++-- libavformat/Makefile | 1 + libavformat/allformats.c | 1 + libavformat/webpdec.c | 384 +++++++ tests/fate/image.mak | 9 + tests/ref/fate/exif-image-webp | 4 +- tests/ref/fate/webp-anim | 22 + tests/ref/fate/webp-chfmt1 | 23 + tests/ref/fate/webp-chfmt2 | 106 ++ tests/ref/fate/webp-rgb-lena-lossless | 2 +- tests/ref/fate/webp-rgb-lena-lossless-rgb24 | 2 +- tests/ref/fate/webp-rgb-lossless | 2 +- .../fate/webp-rgb-lossless-palette-predictor | 2 +- tests/ref/fate/webp-rgb-lossy-q80 | 2 +- tests/ref/fate/webp-rgba-lossless | 2 +- tests/ref/fate/webp-rgba-lossy-q80 | 2 +- 23 files changed, 1855 insertions(+), 81 deletions(-) create mode 100644 libavcodec/bsf/awebp2webp.c create mode 100644 libavformat/webpdec.c create mode 100644 tests/ref/fate/webp-anim create mode 100644 tests/ref/fate/webp-chfmt1 create mode 100644 tests/ref/fate/webp-chfmt2 -- 2.39.3 (Apple Git-146)

2 12