December 2025 - ffmpeg-cvslog

[ffmpeg] avcodec/jpegls: rename SOF48 to SOF55 (branch master)
by Ramiro Polla 30 Dec '25

30 Dec '25

This is an automated email from the git hooks/post-receive script. Git pushed a commit to branch master in repository ffmpeg. The following commit(s) were added to refs/heads/master by this push: new aa80a7880b avcodec/jpegls: rename SOF48 to SOF55 aa80a7880b is described below commit aa80a7880b9256e2e1610d046004334c5a47129c Author: Ramiro Polla <ramiro.polla(a)gmail.com> AuthorDate: Mon Oct 13 17:24:07 2025 +0200 Commit: Ramiro Polla <ramiro.polla(a)gmail.com> CommitDate: Tue Dec 30 17:30:45 2025 +0000 avcodec/jpegls: rename SOF48 to SOF55 SOF48 (0xf0) was used in a public draft of the standard (FCD 14495). In the final specification it is called SOF55 (0xf7). --- libavcodec/jpeglsenc.c | 2 +- libavcodec/mjpeg.h | 2 +- libavcodec/mjpegdec.c | 8 ++++---- 3 files changed, 6 insertions(+), 6 deletions(-) diff --git a/libavcodec/jpeglsenc.c b/libavcodec/jpeglsenc.c index edc3fa600c..64440de242 100644 --- a/libavcodec/jpeglsenc.c +++ b/libavcodec/jpeglsenc.c @@ -373,7 +373,7 @@ static int encode_picture_ls(AVCodecContext *avctx, AVPacket *pkt, /* write our own JPEG header, can't use mjpeg_picture_header */ put_marker_byteu(&pb, SOI); - put_marker_byteu(&pb, SOF48); + put_marker_byteu(&pb, SOF55); bytestream2_put_be16u(&pb, 8 + comps * 3); // header size depends on components bytestream2_put_byteu(&pb, (avctx->pix_fmt == AV_PIX_FMT_GRAY16) ? 16 : 8); // bpp bytestream2_put_be16u(&pb, avctx->height); diff --git a/libavcodec/mjpeg.h b/libavcodec/mjpeg.h index cd5d0af2ac..e66e39714d 100644 --- a/libavcodec/mjpeg.h +++ b/libavcodec/mjpeg.h @@ -100,7 +100,7 @@ enum JpegMarker { JPG4 = 0xf4, JPG5 = 0xf5, JPG6 = 0xf6, - SOF48 = 0xf7, ///< JPEG-LS + SOF55 = 0xf7, ///< JPEG-LS LSE = 0xf8, ///< JPEG-LS extension parameters JPG9 = 0xf9, JPG10 = 0xfa, diff --git a/libavcodec/mjpegdec.c b/libavcodec/mjpegdec.c index 76c81fb152..c347825c37 100644 --- a/libavcodec/mjpegdec.c +++ b/libavcodec/mjpegdec.c @@ -2431,7 +2431,7 @@ redo_for_pal8: ret = -1; if (!CONFIG_JPEGLS_DECODER && - (start_code == SOF48 || start_code == LSE)) { + (start_code == SOF55 || start_code == LSE)) { av_log(avctx, AV_LOG_ERROR, "JPEG-LS support not enabled.\n"); return AVERROR(ENOSYS); } @@ -2442,7 +2442,7 @@ redo_for_pal8: case SOF1: case SOF2: case SOF3: - case SOF48: + case SOF55: break; default: goto skip; @@ -2496,7 +2496,7 @@ FF_ENABLE_DEPRECATION_WARNINGS if ((ret = ff_mjpeg_decode_sof(s)) < 0) goto fail; break; - case SOF48: + case SOF55: avctx->profile = AV_PROFILE_MJPEG_JPEG_LS; #if FF_API_CODEC_PROPS FF_DISABLE_DEPRECATION_WARNINGS @@ -2591,7 +2591,7 @@ eoi_parser: case SOF1: case SOF2: case SOF3: - case SOF48: + case SOF55: s->got_picture = 0; goto the_end_no_picture; }

1 0

[ffmpeg] lavc/vvc: Prevent OOB write to slice_top_left_ctu_x in PPS CBS (branch master)
by Frank Plowman 30 Dec '25

30 Dec '25

This is an automated email from the git hooks/post-receive script. Git pushed a commit to branch master in repository ffmpeg. The following commit(s) were added to refs/heads/master by this push: new 72a38c12e5 lavc/vvc: Prevent OOB write to slice_top_left_ctu_x in PPS CBS 72a38c12e5 is described below commit 72a38c12e5b84ccb30fba88c39ef2a086013af5b Author: Frank Plowman <post(a)frankplowman.com> AuthorDate: Mon Dec 29 22:14:53 2025 +0000 Commit: Timo Rothenpieler <timo(a)rothenpieler.org> CommitDate: Tue Dec 30 13:39:40 2025 +0000 lavc/vvc: Prevent OOB write to slice_top_left_ctu_x in PPS CBS Prior to the fix, in the case of a tile containing multiple slices (pps_num_exp_slices_in_tile != 0) the number of slices was temporarily allowed to exceed pps_num_slices_in_pic_minus1+1 and therefore VVC_MAX_SLICES. The number of slices was later verified, but while the current slice index was higher than expected it was used to write to a array of size VVC_MAX_SLICES, leading to an OOB write. To rectify this, the patch adds some checks at an earlier stage, to ensure that the slice index i + j at no point exceeds pps_num_slices_in_pic_minus1. Fixes #YWH-PGM40646-30 --- libavcodec/cbs_h266_syntax_template.c | 13 +++++++++++++ 1 file changed, 13 insertions(+) diff --git a/libavcodec/cbs_h266_syntax_template.c b/libavcodec/cbs_h266_syntax_template.c index d9a65a9c19..4f6ae76e27 100644 --- a/libavcodec/cbs_h266_syntax_template.c +++ b/libavcodec/cbs_h266_syntax_template.c @@ -2011,6 +2011,12 @@ static int FUNC(pps) (CodedBitstreamContext *ctx, RWContext *rw, slice_top_left_ctu_y[i] = ctu_y; } else { uint16_t slice_height_in_ctus; + int num_uniform_slices; + + if (i + current->pps_num_exp_slices_in_tile[i] > + current->pps_num_slices_in_pic_minus1 + 1) + return AVERROR_INVALIDDATA; + for (j = 0; j < current->pps_num_exp_slices_in_tile[i]; j++) { ues(pps_exp_slice_height_in_ctus_minus1[i][j], 0, @@ -2031,6 +2037,13 @@ static int FUNC(pps) (CodedBitstreamContext *ctx, RWContext *rw, uniform_slice_height = 1 + (j == 0 ? current->row_height_val[tile_y] - 1: current->pps_exp_slice_height_in_ctus_minus1[i][j-1]); + + num_uniform_slices = (remaining_height_in_ctbs_y + uniform_slice_height - 1) + / uniform_slice_height; + if (i + current->pps_num_exp_slices_in_tile[i] + num_uniform_slices > + current->pps_num_slices_in_pic_minus1 + 1) + return AVERROR_INVALIDDATA; + while (remaining_height_in_ctbs_y > uniform_slice_height) { current->slice_height_in_ctus[i + j] = uniform_slice_height;

1 0

[ffmpeg] fftools/ffmpeg_filter: fix access private API of libavcodec (branch master)
by Zhao Zhili 30 Dec '25

30 Dec '25

This is an automated email from the git hooks/post-receive script. Git pushed a commit to branch master in repository ffmpeg. The following commit(s) were added to refs/heads/master by this push: new ee2eb6ced8 fftools/ffmpeg_filter: fix access private API of libavcodec ee2eb6ced8 is described below commit ee2eb6ced80f6799f348f56803e939b0f1e0f3b8 Author: Zhao Zhili <quinkblack(a)foxmail.com> AuthorDate: Mon Dec 29 07:25:49 2025 +0000 Commit: Zhao Zhili <quink(a)noreply.code.ffmpeg.org> CommitDate: Tue Dec 30 07:40:15 2025 +0000 fftools/ffmpeg_filter: fix access private API of libavcodec Firstly, mathops.h is a private header of libavcodec and should not be used in fftools. It may reference libavcodec internal symbols, causing link error with dynamic library, e.g., fftools/ffmpeg_filter.c:2687: undefined reference to `ff_rv_zbb_supported' Secondly, mid_pred operates on int types, while current use case is int64_t. --- fftools/ffmpeg_filter.c | 26 ++++++++++++++++++++------ 1 file changed, 20 insertions(+), 6 deletions(-) diff --git a/fftools/ffmpeg_filter.c b/fftools/ffmpeg_filter.c index d5ae59ec03..4387c992da 100644 --- a/fftools/ffmpeg_filter.c +++ b/fftools/ffmpeg_filter.c @@ -40,9 +40,6 @@ #include "libavutil/time.h" #include "libavutil/timestamp.h" -// FIXME private header, used for mid_pred() -#include "libavcodec/mathops.h" - typedef struct FilterGraphPriv { FilterGraph fg; @@ -2489,6 +2486,23 @@ early_exit: return float_pts; } +static int64_t median3(int64_t a, int64_t b, int64_t c) +{ + int64_t max2, min2, m; + + if (a >= b) { + max2 = a; + min2 = b; + } else { + max2 = b; + min2 = a; + } + m = (c >= max2) ? max2 : c; + + return (m >= min2) ? m : min2; +} + + /* Convert frame timestamps to the encoder timebase and decide how many times * should this (and possibly previous) frame be repeated in order to conform to * desired target framerate (if any). @@ -2501,9 +2515,9 @@ static void video_sync_process(OutputFilterPriv *ofp, AVFrame *frame, double delta0, delta, sync_ipts, duration; if (!frame) { - *nb_frames_prev = *nb_frames = mid_pred(fps->frames_prev_hist[0], - fps->frames_prev_hist[1], - fps->frames_prev_hist[2]); + *nb_frames_prev = *nb_frames = median3(fps->frames_prev_hist[0], + fps->frames_prev_hist[1], + fps->frames_prev_hist[2]); if (!*nb_frames && fps->last_dropped) { atomic_fetch_add(&ofilter->nb_frames_drop, 1);

1 0

[ffmpeg] libavformat/file: apply pkt_size option to read mode (branch master)
by caifan3 30 Dec '25

30 Dec '25

This is an automated email from the git hooks/post-receive script. Git pushed a commit to branch master in repository ffmpeg. The following commit(s) were added to refs/heads/master by this push: new 82144fea48 libavformat/file: apply pkt_size option to read mode 82144fea48 is described below commit 82144fea488387539d7008bef5cdbf52ccbc6113 Author: caifan3 <caifan3(a)xiaomi.com> AuthorDate: Mon Dec 22 14:32:00 2025 +0800 Commit: Marton Balint <cus(a)passwd.hu> CommitDate: Tue Dec 30 00:49:13 2025 +0000 libavformat/file: apply pkt_size option to read mode Previously, the pkt_size option only affected write buffering in the file protocol. This change extends its effect to read mode as well. On embedded systems with limited RAM, users can now reduce I/O buffer memory by setting a smaller pkt_size. Signed-off-by: caifan3 <caifan3(a)xiaomi.com> --- doc/protocols.texi | 9 +++++++++ libavformat/file.c | 10 +++++++--- 2 files changed, 16 insertions(+), 3 deletions(-) diff --git a/doc/protocols.texi b/doc/protocols.texi index dee2b845ac..b5330a1160 100644 --- a/doc/protocols.texi +++ b/doc/protocols.texi @@ -352,6 +352,15 @@ means auto (seekable for normal files, non-seekable for named pipes). Many demuxers handle seekable and non-seekable resources differently, overriding this might speed up opening certain files at the cost of losing some features (e.g. accurate seeking). + +@item pkt_size +Set the maximum packet size used for file I/O. + +For writing, this sets the size of each write operation. The default is +262144 bytes. +For reading, if explicitly set, it overrides the default internal buffer size +(32768 bytes) and limits the maximum amount of data read per operation. +Setting a smaller value may reduce memory usage when reading files sequentially. @end table @section ftp diff --git a/libavformat/file.c b/libavformat/file.c index 97f5955f93..23dc7081d3 100644 --- a/libavformat/file.c +++ b/libavformat/file.c @@ -107,7 +107,7 @@ static const AVOption file_options[] = { { "blocksize", "set I/O operation maximum block size", offsetof(FileContext, blocksize), AV_OPT_TYPE_INT, { .i64 = INT_MAX }, 1, INT_MAX, AV_OPT_FLAG_ENCODING_PARAM }, { "follow", "Follow a file as it is being written", offsetof(FileContext, follow), AV_OPT_TYPE_INT, { .i64 = 0 }, 0, 1, AV_OPT_FLAG_DECODING_PARAM }, { "seekable", "Sets if the file is seekable", offsetof(FileContext, seekable), AV_OPT_TYPE_INT, { .i64 = -1 }, -1, 0, AV_OPT_FLAG_DECODING_PARAM | AV_OPT_FLAG_ENCODING_PARAM }, - { "pkt_size", "Maximum packet size", offsetof(FileContext, pkt_size), AV_OPT_TYPE_INT, { .i64 = 262144 }, 1, INT_MAX, AV_OPT_FLAG_DECODING_PARAM | AV_OPT_FLAG_ENCODING_PARAM }, + { "pkt_size", "Maximum packet size", offsetof(FileContext, pkt_size), AV_OPT_TYPE_INT, { .i64 = 0 }, 0, INT_MAX, AV_OPT_FLAG_DECODING_PARAM | AV_OPT_FLAG_ENCODING_PARAM }, { NULL } }; @@ -314,8 +314,12 @@ static int file_open(URLContext *h, const char *filename, int flags) /* Buffer writes more than the default 32k to improve throughput especially * with networked file systems */ - if (!h->is_streamed && flags & AVIO_FLAG_WRITE) - h->min_packet_size = h->max_packet_size = c->pkt_size; + if (!h->is_streamed) { + if (flags & AVIO_FLAG_WRITE) + h->min_packet_size = h->max_packet_size = c->pkt_size ? c->pkt_size : 262144; + else if (flags & AVIO_FLAG_READ && c->pkt_size) + h->max_packet_size = c->pkt_size; + } if (c->seekable >= 0) h->is_streamed = !c->seekable;

1 0

[ffmpeg] avcodec/mjpegdec: speed up find_marker() (branch master)
by Ramiro Polla 29 Dec '25

29 Dec '25

This is an automated email from the git hooks/post-receive script. Git pushed a commit to branch master in repository ffmpeg. The following commit(s) were added to refs/heads/master by this push: new 7dda7f3b99 avcodec/mjpegdec: speed up find_marker() 7dda7f3b99 is described below commit 7dda7f3b99a065c5562fb998b655b77f467c5488 Author: Ramiro Polla <ramiro.polla(a)gmail.com> AuthorDate: Tue Sep 23 17:11:40 2025 +0200 Commit: Ramiro Polla <ramiro.polla(a)gmail.com> CommitDate: Mon Dec 29 21:44:07 2025 +0000 avcodec/mjpegdec: speed up find_marker() Minimize number of reads and simplify conditionals. Also use memchr(), as suggested by Andreas Rheinhardt <andreas.rheinhardt(a)outlook.com> --- libavcodec/mjpegdec.c | 14 ++++++++------ 1 file changed, 8 insertions(+), 6 deletions(-) diff --git a/libavcodec/mjpegdec.c b/libavcodec/mjpegdec.c index 9b7afc96e8..76c81fb152 100644 --- a/libavcodec/mjpegdec.c +++ b/libavcodec/mjpegdec.c @@ -2194,17 +2194,19 @@ static int mjpeg_decode_com(MJpegDecodeContext *s) static int find_marker(const uint8_t **pbuf_ptr, const uint8_t *buf_end) { const uint8_t *buf_ptr; - unsigned int v, v2; int val; int skipped = 0; buf_ptr = *pbuf_ptr; - while (buf_end - buf_ptr > 1) { - v = *buf_ptr++; - v2 = *buf_ptr; - if ((v == 0xff) && (v2 >= SOF0) && (v2 <= COM) && buf_ptr < buf_end) { + while ((buf_ptr = memchr(buf_ptr, 0xff, buf_end - buf_ptr))) { + buf_ptr++; + while (buf_ptr < buf_end) { val = *buf_ptr++; - goto found; + if (val != 0xff) { + if ((val >= SOF0) && (val <= COM)) + goto found; + break; + } } skipped++; }

1 0

[ffmpeg] avcodec/mjpegdec: fix some error return codes (branch master)
by Ramiro Polla 29 Dec '25

29 Dec '25

This is an automated email from the git hooks/post-receive script. Git pushed a commit to branch master in repository ffmpeg. The following commit(s) were added to refs/heads/master by this push: new 254124f869 avcodec/mjpegdec: fix some error return codes 254124f869 is described below commit 254124f8699ab9225e0ad6119e06980e752ac836 Author: Ramiro Polla <ramiro.polla(a)gmail.com> AuthorDate: Fri Sep 12 12:04:23 2025 +0200 Commit: Ramiro Polla <ramiro.polla(a)gmail.com> CommitDate: Mon Dec 29 17:22:34 2025 +0100 avcodec/mjpegdec: fix some error return codes --- libavcodec/mjpegdec.c | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) diff --git a/libavcodec/mjpegdec.c b/libavcodec/mjpegdec.c index 05150e982c..9b7afc96e8 100644 --- a/libavcodec/mjpegdec.c +++ b/libavcodec/mjpegdec.c @@ -210,7 +210,7 @@ int ff_mjpeg_decode_dqt(MJpegDecodeContext *s) } index = get_bits(&s->gb, 4); if (index >= 4) - return -1; + return AVERROR_INVALIDDATA; av_log(s->avctx, AV_LOG_DEBUG, "index=%d\n", index); /* read quant table */ for (i = 0; i < 64; i++) { @@ -326,7 +326,7 @@ int ff_mjpeg_decode_sof(MJpegDecodeContext *s) if(s->lossless && s->avctx->lowres){ av_log(s->avctx, AV_LOG_ERROR, "lowres is not possible with lossless jpeg\n"); - return -1; + return AVERROR(ENOSYS); } height = get_bits(&s->gb, 16); @@ -349,7 +349,7 @@ int ff_mjpeg_decode_sof(MJpegDecodeContext *s) nb_components = get_bits(&s->gb, 8); if (nb_components <= 0 || nb_components > MAX_COMPONENTS) - return -1; + return AVERROR_INVALIDDATA; if (s->interlaced && (s->bottom_field == !s->interlace_polarity)) { if (nb_components != s->nb_components) { av_log(s->avctx, AV_LOG_ERROR, @@ -757,8 +757,9 @@ int ff_mjpeg_decode_sof(MJpegDecodeContext *s) } av_frame_unref(s->picture_ptr); - if (ff_get_buffer(s->avctx, s->picture_ptr, AV_GET_BUFFER_FLAG_REF) < 0) - return -1; + ret = ff_get_buffer(s->avctx, s->picture_ptr, AV_GET_BUFFER_FLAG_REF); + if (ret < 0) + return ret; s->picture_ptr->pict_type = AV_PICTURE_TYPE_I; s->picture_ptr->flags |= AV_FRAME_FLAG_KEY; s->got_picture = 1; @@ -1677,7 +1678,7 @@ int ff_mjpeg_decode_sos(MJpegDecodeContext *s, const uint8_t *mb_bitmask, if (!s->got_picture) { av_log(s->avctx, AV_LOG_WARNING, "Can not process SOS before SOF, skipping\n"); - return -1; + return AVERROR_INVALIDDATA; } /* XXX: verify len field validity */

1 0

[ffmpeg] branch master updated (d28250eada -> 6223b7fa0b)
by ffmpeg-git 29 Dec '25

29 Dec '25

This is an automated email from the git hooks/post-receive script. Git pushed a change to branch master in repository ffmpeg. from d28250eada avcodec/adpcm: Fix step_index decoding by skipping padding byte new a6e5fa3fbb avutil/iamf: add an AVOption for AVIAMFLayer.demixing_matrix new 03a981445e avformat/iamf_parse: fix setting denominator in AVIAMFLayer.demixing_matrix new e3498c8051 avformat/iamf_parse: set AVIAMFLayer.nb_demixing_matrix new 21ff60d2cf avformat/iamf_writer: fix writting some ambisonics fields in Audio Elements new a0fc454871 avformat/iamf_writer: check that stream count is consistent for ambisonic Audio Elements new 8cd4d4cb0b avformat/iamf_writer: add support for Projection mode ambisonic Audio elements new 129744ac72 avformat/iamf_writer: reindent after the previous change new 9c5a37388b fftools/ffprobe: print demixing_matrix in Projection mode ambisonic IAMF Audio Element groups new 6223b7fa0b tests/fate/iamf: add tests for Projection mode ambisonic Audio Elements The 9 revisions listed above as "new" are entirely new to this repository and will be described in separate emails. The revisions listed as "add" were already present in the repository and have only been added to this reference. Summary of changes: Changelog | 1 + doc/APIchanges | 3 + fftools/ffprobe.c | 12 +++- libavformat/iamf_parse.c | 12 ++-- libavformat/iamf_writer.c | 33 +++++++--- libavformat/version.h | 2 +- libavutil/iamf.c | 4 ++ libavutil/iamf.h | 11 ++-- libavutil/version.h | 2 +- tests/fate/iamf.mak | 20 ++++++ ...amf-ambisonic_1 => iamf-ambisonic_1-projection} | 7 +- ...f-5_1-copy => iamf-ambisonic_1-projection-copy} | 74 ++++++---------------- ...-5_1-copy => iamf-ambisonic_1-projection-demux} | 74 ++++++---------------- .../audio_element-ambisonic_1-projection | 2 + 14 files changed, 120 insertions(+), 137 deletions(-) copy tests/ref/fate/{iamf-ambisonic_1 => iamf-ambisonic_1-projection} (96%) copy tests/ref/fate/{iamf-5_1-copy => iamf-ambisonic_1-projection-copy} (83%) copy tests/ref/fate/{iamf-5_1-copy => iamf-ambisonic_1-projection-demux} (83%) create mode 100644 tests/streamgroups/audio_element-ambisonic_1-projection

2 9

[ffmpeg] avcodec/adpcm: Fix step_index decoding by skipping padding byte (branch master)
by tangsha 29 Dec '25

29 Dec '25

This is an automated email from the git hooks/post-receive script. Git pushed a commit to branch master in repository ffmpeg. The following commit(s) were added to refs/heads/master by this push: new d28250eada avcodec/adpcm: Fix step_index decoding by skipping padding byte d28250eada is described below commit d28250eada1b4c257d6810b83129e18e5346e4ed Author: tangsha <tangsha(a)kylinos.cn> AuthorDate: Mon Dec 1 16:15:35 2025 +0800 Commit: Marton Balint <cus(a)passwd.hu> CommitDate: Mon Dec 29 14:18:35 2025 +0000 avcodec/adpcm: Fix step_index decoding by skipping padding byte The bug was caused by incorrect reading of the step_index field: the original code read a 16-bit value (including a padding byte) instead of the correct 8-bit step_index as defined by the ADPCM format. This led to distorted audio or incorrect step calculations when decoding specific ADPCM-encoded files. Fix by: 1. Reading step_index as an 8-bit unsigned byte via bytestream2_get_byteu() 2. Skipping the subsequent 8-bit padding byte with bytestream2_skip() --- libavcodec/adpcm.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/libavcodec/adpcm.c b/libavcodec/adpcm.c index 3da0791ba3..31340af677 100644 --- a/libavcodec/adpcm.c +++ b/libavcodec/adpcm.c @@ -1512,7 +1512,8 @@ static int adpcm_decode_frame(AVCodecContext *avctx, AVFrame *frame, ADPCMChannelStatus *cs = &c->status[i]; cs->predictor = samples_p[i][0] = sign_extend(bytestream2_get_le16u(&gb), 16); - cs->step_index = sign_extend(bytestream2_get_le16u(&gb), 16); + cs->step_index = bytestream2_get_byteu(&gb); + bytestream2_skipu(&gb, 1); if (cs->step_index > 88u){ av_log(avctx, AV_LOG_ERROR, "ERROR: step_index[%d] = %i\n", i, cs->step_index);

1 0

[ffmpeg] avcodec/mjpegdec: only test the size bound in sequential mjpeg (branch master)
by Michael Niedermayer 29 Dec '25

29 Dec '25

This is an automated email from the git hooks/post-receive script. Git pushed a commit to branch master in repository ffmpeg. The following commit(s) were added to refs/heads/master by this push: new ecd2919174 avcodec/mjpegdec: only test the size bound in sequential mjpeg ecd2919174 is described below commit ecd29191747b4546cf2b5fbe84402e18163a3430 Author: Michael Niedermayer <michael(a)niedermayer.cc> AuthorDate: Fri Dec 19 19:45:56 2025 +0100 Commit: michaelni <michael(a)niedermayer.cc> CommitDate: Mon Dec 29 13:57:33 2025 +0000 avcodec/mjpegdec: only test the size bound in sequential mjpeg The original fix was intended only for sequential mjpeg, but it was also used for progressive which broke. This commit fixes this regression Fixes: issue21225 The testcase 6381/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_JPEGLS_fuzzer-5665032743419904 still exits within 240ms Signed-off-by: Michael Niedermayer <michael(a)niedermayer.cc> --- libavcodec/mjpegdec.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/libavcodec/mjpegdec.c b/libavcodec/mjpegdec.c index 4d7cdfde12..05150e982c 100644 --- a/libavcodec/mjpegdec.c +++ b/libavcodec/mjpegdec.c @@ -340,9 +340,11 @@ int ff_mjpeg_decode_sof(MJpegDecodeContext *s) if (av_image_check_size(width, height, 0, s->avctx) < 0) return AVERROR_INVALIDDATA; - // A valid frame requires at least 1 bit for DC + 1 bit for AC for each 8x8 block. - if (s->buf_size && (width + 7) / 8 * ((height + 7) / 8) > s->buf_size * 4LL) - return AVERROR_INVALIDDATA; + if (!s->progressive && !s->ls) { + // A valid frame requires at least 1 bit for DC + 1 bit for AC for each 8x8 block. + if (s->buf_size && (width + 7) / 8 * ((height + 7) / 8) > s->buf_size * 4LL) + return AVERROR_INVALIDDATA; + } nb_components = get_bits(&s->gb, 8); if (nb_components <= 0 ||

1 0

[ffmpeg] libavformat: Fix the docs for av_mime_codec_str and ff_make_codec_str (branch master)
by Martin Storsjö 29 Dec '25

29 Dec '25

This is an automated email from the git hooks/post-receive script. Git pushed a commit to branch master in repository ffmpeg. The following commit(s) were added to refs/heads/master by this push: new 78bfcf003b libavformat: Fix the docs for av_mime_codec_str and ff_make_codec_str 78bfcf003b is described below commit 78bfcf003b21ed5cf002f0a6b66a0babfc6cfc06 Author: Martin Storsjö <martin(a)martin.st> AuthorDate: Fri Dec 19 20:07:19 2025 +0200 Commit: Martin Storsjö <martin(a)martin.st> CommitDate: Mon Dec 29 12:05:14 2025 +0000 libavformat: Fix the docs for av_mime_codec_str and ff_make_codec_str The doxygen comments were missed as these functions were updated during review; they don't take separate pointer/length parameters but use an AVBPrint struct now instead. Also clarify that ff_make_codec_str doesn't log if the logctx parameter is NULL. --- libavformat/avformat.h | 3 +-- libavformat/internal.h | 6 +++--- 2 files changed, 4 insertions(+), 5 deletions(-) diff --git a/libavformat/avformat.h b/libavformat/avformat.h index cee845b308..bd34132e00 100644 --- a/libavformat/avformat.h +++ b/libavformat/avformat.h @@ -2898,8 +2898,7 @@ struct AVBPrint; * @param frame_rate an AVRational for the frame rate, for deciding the * right profile for video codecs. Pass an invalid * AVRational (1/0) to indicate that it is unknown. - * @param str the output string buffer - * @param size the size of the string pointed to by str + * @param out the AVBPrint to write the output to * @return <0 on error */ int av_mime_codec_str(const AVCodecParameters *par, diff --git a/libavformat/internal.h b/libavformat/internal.h index d32a3376cc..245f6eeb86 100644 --- a/libavformat/internal.h +++ b/libavformat/internal.h @@ -664,12 +664,12 @@ int ff_parse_opts_from_query_string(void *obj, const char *str, int allow_unkown /** * Make a RFC 4281/6381 like string describing a codec. * - * @param logctx a context for potential log messages + * @param logctx a context for potential log messages; if NULL, nothing is + * logged * @param par pointer to an AVCodecParameters struct describing the codec * @param frame_rate an optional pointer to AVRational for the frame rate, * for deciding the right profile for video codecs - * @param str the output string buffer - * @param size the size of the string pointed to by str + * @param out the AVBPrint to write the output to * @return <0 on error */ int ff_make_codec_str(void *logctx, const AVCodecParameters *par,

1 0