December 2025 - ffmpeg-cvslog

[ffmpeg] branch master updated (9e974b2c45 -> 3e91d82292)
by ffmpeg-git 24 Dec '25

24 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 9e974b2c45 avformat/hls: fix double space new 2678482c10 fftools/ffmpeg_enc: use av_unreachable instead of av_assert0(0) at the end of a non-void function new 64be23153e fftools/ffmpeg_sched: use av_unreachable instead of av_assert0(0) at the end of a non-void function new 3e91d82292 avcodec/cbs_apv: use av_unreachable instead of av_assert0(0) at the end of a non-void function The 3 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: fftools/ffmpeg_enc.c | 2 +- fftools/ffmpeg_sched.c | 2 +- libavcodec/cbs_apv.c | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-)

2 3

[ffmpeg] avformat/hls: fix double space (branch master)
by Michael Niedermayer 24 Dec '25

24 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 9e974b2c45 avformat/hls: fix double space 9e974b2c45 is described below commit 9e974b2c45453217f6bbc8ca868e24fb69863847 Author: Michael Niedermayer <michael(a)niedermayer.cc> AuthorDate: Wed Dec 24 14:13:56 2025 +0100 Commit: Michael Niedermayer <michael(a)niedermayer.cc> CommitDate: Wed Dec 24 14:14:45 2025 +0100 avformat/hls: fix double space Signed-off-by: Michael Niedermayer <michael(a)niedermayer.cc> --- libavformat/hls.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/libavformat/hls.c b/libavformat/hls.c index bc5494e42d..dabfaae5bc 100644 --- a/libavformat/hls.c +++ b/libavformat/hls.c @@ -977,7 +977,7 @@ static int parse_playlist(HLSContext *c, const char *url, ptr = strchr(ptr, '@'); if (ptr) seg_offset = strtoll(ptr+1, NULL, 10); - if (seg_size < 0 || seg_offset > INT64_MAX - seg_size) { + if (seg_size < 0 || seg_offset > INT64_MAX - seg_size) { ret = AVERROR_INVALIDDATA; goto fail; }

1 0

[ffmpeg] branch master updated (8ac6325143 -> 548b28d5b1)
by ffmpeg-git 24 Dec '25

24 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 8ac6325143 avcodec/adpcm: check block_align for PSXC new d1985442e1 avformat/hls: Check seg size and offset for overflow new 548b28d5b1 avformat/hls: Fix arguments of handle_rendition_args() The 2 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: libavformat/hls.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-)

2 2

[ffmpeg] avcodec/adpcm: check block_align for PSXC (branch master)
by Michael Niedermayer 24 Dec '25

24 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 8ac6325143 avcodec/adpcm: check block_align for PSXC 8ac6325143 is described below commit 8ac6325143958373bd415b75e44d56d62d2cc802 Author: Michael Niedermayer <michael(a)niedermayer.cc> AuthorDate: Mon Dec 22 18:30:52 2025 +0100 Commit: James Almer <jamrial(a)gmail.com> CommitDate: Wed Dec 24 02:51:20 2025 +0000 avcodec/adpcm: check block_align for PSXC Fixes: division by 0 Fixes: 457634826/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_ADPCM_PSXC_DEC_fuzzer-4903668521697280 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/adpcm.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/libavcodec/adpcm.c b/libavcodec/adpcm.c index 72fbb841a4..3da0791ba3 100644 --- a/libavcodec/adpcm.c +++ b/libavcodec/adpcm.c @@ -294,7 +294,7 @@ static av_cold int adpcm_decode_init(AVCodecContext * avctx) break; case AV_CODEC_ID_ADPCM_PSXC: max_channels = 8; - if (avctx->ch_layout.nb_channels <= 0) + if (avctx->ch_layout.nb_channels <= 0 || avctx->block_align <= 0) return AVERROR_INVALIDDATA; break; case AV_CODEC_ID_ADPCM_IMA_DAT4:

1 0

[ffmpeg] avcodec/libjxldec: goto SUCCESS when frame is completed (branch master)
by Jack Lau 24 Dec '25

24 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 13c91c97d1 avcodec/libjxldec: goto SUCCESS when frame is completed 13c91c97d1 is described below commit 13c91c97d12a28750f572c87cf13934456845df1 Author: Jack Lau <jacklau1222gm(a)gmail.com> AuthorDate: Tue Dec 23 10:53:23 2025 +0800 Commit: James Almer <jamrial(a)gmail.com> CommitDate: Wed Dec 24 02:43:50 2025 +0000 avcodec/libjxldec: goto SUCCESS when frame is completed Fix #21260 The early code didn't reset the JxlDecoder. If the caller re-send the same packet to JxlDecoder after avcodec_flush_buffers() (free ctx->frame). It will directly hit SUCCESS event (skip all EVENTs). In this case, the ctx->frame already flushed by avcodec_flush_buffers, so the received frame will be invalid (has no private_ref) that causes assertion in decode_receive_frame_internal(). MPV meet this issue because it seek the jxl image (second read the same pkt) when it does crop through config `C vf toggle crop=in_w:in_w/2.4`. This patch make sure hit the entire SUCCESS event (contains reset the JxlDecoder) when frame_complete is true. Signed-off-by: Jack Lau <jacklau1222gm(a)gmail.com> --- libavcodec/libjxldec.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/libavcodec/libjxldec.c b/libavcodec/libjxldec.c index a43437fb10..a2f125998e 100644 --- a/libavcodec/libjxldec.c +++ b/libavcodec/libjxldec.c @@ -450,9 +450,8 @@ static int libjxl_receive_frame(AVCodecContext *avctx, AVFrame *frame) av_log(avctx, AV_LOG_ERROR, "Unexpected end of JXL codestream\n"); return AVERROR_INVALIDDATA; } else if (ctx->frame_complete) { - libjxl_finalize_frame(avctx, frame, ctx->frame); ctx->jret = JXL_DEC_SUCCESS; - return 0; + goto success; } return AVERROR_EOF; } @@ -591,6 +590,7 @@ static int libjxl_receive_frame(AVCodecContext *avctx, AVFrame *frame) * but it will also be fired when the next image of * an image2pipe sequence is loaded up */ +success: libjxl_finalize_frame(avctx, frame, ctx->frame); JxlDecoderReset(ctx->decoder); libjxl_init_jxl_decoder(avctx);

1 0

[ffmpeg] avcodec/jpeg2000dec: Make M_b check broader (branch master)
by Michael Niedermayer 23 Dec '25

23 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 d60c1d72c1 avcodec/jpeg2000dec: Make M_b check broader d60c1d72c1 is described below commit d60c1d72c1299e683c1d2f86cb4332b4d88348a5 Author: Michael Niedermayer <michael(a)niedermayer.cc> AuthorDate: Mon Dec 22 23:17:24 2025 +0100 Commit: michaelni <michael(a)niedermayer.cc> CommitDate: Tue Dec 23 20:11:29 2025 +0000 avcodec/jpeg2000dec: Make M_b check broader Fixes: shift exponent -1 is negative Fixes: 429330004/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_JPEG2000_DEC_fuzzer-4733213845291008 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/jpeg2000dec.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/libavcodec/jpeg2000dec.c b/libavcodec/jpeg2000dec.c index de1a73b92b..276c0a6e43 100644 --- a/libavcodec/jpeg2000dec.c +++ b/libavcodec/jpeg2000dec.c @@ -2270,8 +2270,8 @@ static inline int tile_codeblocks(const Jpeg2000DecoderContext *s, Jpeg2000Tile band->coord[1][0] == band->coord[1][1]) continue; - if ((codsty->cblk_style & JPEG2000_CTSY_HTJ2K_F) && M_b >= 31) { - avpriv_request_sample(s->avctx, "JPEG2000_CTSY_HTJ2K_F and M_b >= 31"); + if (M_b >= 31) { + avpriv_request_sample(s->avctx, "M_b >= 31"); return AVERROR_PATCHWELCOME; }

1 0

[ffmpeg] vulkan: set .decode_extension to 0 for compute-based decoders (branch master)
by Lynne 23 Dec '25

23 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 e5eef271a5 vulkan: set .decode_extension to 0 for compute-based decoders e5eef271a5 is described below commit e5eef271a5f6e20de9eff7660680bf42e5ef4f89 Author: Lynne <dev(a)lynne.ee> AuthorDate: Tue Dec 23 13:52:17 2025 +0100 Commit: Lynne <dev(a)lynne.ee> CommitDate: Tue Dec 23 15:42:24 2025 +0100 vulkan: set .decode_extension to 0 for compute-based decoders They require no special video decode extension. The push descriptor extension was a placeholder for when the check was performed inside the init function. --- libavcodec/vulkan_dpx.c | 1 - libavcodec/vulkan_ffv1.c | 1 - libavcodec/vulkan_prores_raw.c | 1 - 3 files changed, 3 deletions(-) diff --git a/libavcodec/vulkan_dpx.c b/libavcodec/vulkan_dpx.c index eaf131c383..faad28d624 100644 --- a/libavcodec/vulkan_dpx.c +++ b/libavcodec/vulkan_dpx.c @@ -31,7 +31,6 @@ extern const char *ff_source_dpx_copy_comp; const FFVulkanDecodeDescriptor ff_vk_dec_dpx_desc = { .codec_id = AV_CODEC_ID_DPX, - .decode_extension = FF_VK_EXT_PUSH_DESCRIPTOR, .queue_flags = VK_QUEUE_COMPUTE_BIT, }; diff --git a/libavcodec/vulkan_ffv1.c b/libavcodec/vulkan_ffv1.c index 93f2ccae65..168871d5d9 100644 --- a/libavcodec/vulkan_ffv1.c +++ b/libavcodec/vulkan_ffv1.c @@ -38,7 +38,6 @@ extern const char *ff_source_ffv1_dec_comp; const FFVulkanDecodeDescriptor ff_vk_dec_ffv1_desc = { .codec_id = AV_CODEC_ID_FFV1, - .decode_extension = FF_VK_EXT_PUSH_DESCRIPTOR, .queue_flags = VK_QUEUE_COMPUTE_BIT, }; diff --git a/libavcodec/vulkan_prores_raw.c b/libavcodec/vulkan_prores_raw.c index 0a844156fb..8ccaf1c07f 100644 --- a/libavcodec/vulkan_prores_raw.c +++ b/libavcodec/vulkan_prores_raw.c @@ -32,7 +32,6 @@ extern const char *ff_source_prores_raw_idct_comp; const FFVulkanDecodeDescriptor ff_vk_dec_prores_raw_desc = { .codec_id = AV_CODEC_ID_PRORES_RAW, - .decode_extension = FF_VK_EXT_PUSH_DESCRIPTOR, .queue_flags = VK_QUEUE_COMPUTE_BIT, };

1 0

[ffmpeg] avcodec/gifenc: avoid redundant local color table when global palette is used (branch master)
by Damitha Gunawardena 23 Dec '25

23 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 bb769d9d90 avcodec/gifenc: avoid redundant local color table when global palette is used bb769d9d90 is described below commit bb769d9d901315087f2a7c06b251f27f14b8266f Author: Damitha Gunawardena <damitha.gunawardena(a)gmail.com> AuthorDate: Wed Nov 12 15:12:07 2025 +1100 Commit: Ramiro Polla <ramiro.polla(a)gmail.com> CommitDate: Tue Dec 23 14:34:13 2025 +0000 avcodec/gifenc: avoid redundant local color table when global palette is used Addressing the fact that first frame has a LCT when global pallette is used for pal8. issue: #20896 --- libavcodec/gif.c | 2 ++ tests/ref/fate/gifenc-pal8 | 2 +- 2 files changed, 3 insertions(+), 1 deletion(-) diff --git a/libavcodec/gif.c b/libavcodec/gif.c index c11cad8d38..132ce07e6c 100644 --- a/libavcodec/gif.c +++ b/libavcodec/gif.c @@ -492,6 +492,8 @@ static int gif_encode_frame(AVCodecContext *avctx, AVPacket *pkt, memcpy(s->palette, palette, AVPALETTE_SIZE); s->transparent_index = get_palette_transparency_index(palette); s->palette_loaded = 1; + if (s->use_global_palette) + palette = NULL; } else if (!memcmp(s->palette, palette, AVPALETTE_SIZE)) { palette = NULL; } diff --git a/tests/ref/fate/gifenc-pal8 b/tests/ref/fate/gifenc-pal8 index 1a074b8fd8..d4a42b03b2 100644 --- a/tests/ref/fate/gifenc-pal8 +++ b/tests/ref/fate/gifenc-pal8 @@ -3,7 +3,7 @@ #codec_id 0: gif #dimensions 0: 217x217 #sar 0: 0/1 -0, 0, 0, 1, 2109, 0x39642b3d +0, 0, 0, 1, 1341, 0xe4e2af18 0, 1, 1, 1, 236, 0x332769fd, F=0x0 0, 2, 2, 1, 186, 0x770d5061, F=0x0 0, 3, 3, 1, 208, 0x55784c8f, F=0x0

1 0

[ffmpeg] branch master updated (d41b8d95a7 -> 26a7fc2ac4)
by ffmpeg-git 23 Dec '25

23 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 d41b8d95a7 swscale/output: Use 64bit in addition in yuv2gbrp16_full_X_c() for RGB + Y new 6460d3bc81 avfilter/vf_waveform: don't cast const away new 26a7fc2ac4 avfilter/vf_waveform: Avoid relocations The 2 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: libavfilter/vf_waveform.c | 100 +++++++++++++++++++++++----------------------- 1 file changed, 50 insertions(+), 50 deletions(-)

2 2

[ffmpeg] swscale/output: Use 64bit in addition in yuv2gbrp16_full_X_c() for RGB + Y (branch master)
by Michael Niedermayer 23 Dec '25

23 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 d41b8d95a7 swscale/output: Use 64bit in addition in yuv2gbrp16_full_X_c() for RGB + Y d41b8d95a7 is described below commit d41b8d95a7234cf0b34fcb84e307385fb3ef30a1 Author: Michael Niedermayer <michael(a)niedermayer.cc> AuthorDate: Mon Dec 22 23:44:54 2025 +0100 Commit: michaelni <michael(a)niedermayer.cc> CommitDate: Tue Dec 23 09:14:44 2025 +0000 swscale/output: Use 64bit in addition in yuv2gbrp16_full_X_c() for RGB + Y Fixes: signed integer overflow: -1159988356 + -1082982400 cannot be represented in type 'int' Fixes: 461519938/clusterfuzz-testcase-minimized-ffmpeg_SWS_fuzzer-4777382021234688 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Signed-off-by: Michael Niedermayer <michael(a)niedermayer.cc> --- libswscale/output.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/libswscale/output.c b/libswscale/output.c index 4b3aab3109..a3b8da54bf 100644 --- a/libswscale/output.c +++ b/libswscale/output.c @@ -2513,9 +2513,9 @@ yuv2gbrp16_full_X_c(SwsInternal *c, const int16_t *lumFilter, G = V * c->yuv2rgb_v2g_coeff + U * c->yuv2rgb_u2g_coeff; B = U * c->yuv2rgb_u2b_coeff; - dest16[2][i] = av_clip_uintp2(((Y + R) >> 14) + (1<<15), 16); - dest16[0][i] = av_clip_uintp2(((Y + G) >> 14) + (1<<15), 16); - dest16[1][i] = av_clip_uintp2(((Y + B) >> 14) + (1<<15), 16); + dest16[2][i] = av_clip_uintp2(((Y + (int64_t)R) >> 14) + (1<<15), 16); + dest16[0][i] = av_clip_uintp2(((Y + (int64_t)G) >> 14) + (1<<15), 16); + dest16[1][i] = av_clip_uintp2(((Y + (int64_t)B) >> 14) + (1<<15), 16); if (hasAlpha) dest16[3][i] = av_clip_uintp2(A, 30) >> 14;

1 0