[FFmpeg-cvslog] swscale/aarch64: dotprod implementation of rgba32_to_Y

ffmpeg | branch: master | Krzysztof Pyrkosz | Mon Mar 3 22:00:23 2025 +0100| [d765e5f043d981294303fe210d643c5156efeeb3] | committer: Martin Storsjö swscale/aarch64: dotprod implementation of rgba32_to_Y The idea is to split the 16 bit coefficients into lower and upper half, invoke udot for th

[FFmpeg-cvslog] avcodec/aarch64/vvc: Optimize NEON version of vvc_dmvr

ffmpeg | branch: master | Krzysztof Pyrkosz | Mon Mar 3 22:32:55 2025 +0100| [71a91485fa05c1ca478de153d8839794606f8edc] | committer: Martin Storsjö avcodec/aarch64/vvc: Optimize NEON version of vvc_dmvr This patch replaces blocks of instructions performing rounding and widening shifts with on

[FFmpeg-cvslog] avcodec/h26[134]dec: Always report the buffer to be completely consumed

ffmpeg | branch: master | Andreas Rheinhardt | Mon Feb 24 19:30:27 2025 +0100| [e6657d499a6ec5731da5692515b605493a819249] | committer: Andreas Rheinhardt avcodec/h26[134]dec: Always report the buffer to be completely consumed It is pointless to try to report the true number because decode_simp

[FFmpeg-cvslog] avcodec/mpegvideo: Move bitstream_buffer to mpeg4videodec

ffmpeg | branch: master | Andreas Rheinhardt | Tue Feb 25 00:48:08 2025 +0100| [e3ebc1073e06b87ecf110476194c88c4af9b181f] | committer: Andreas Rheinhardt avcodec/mpegvideo: Move bitstream_buffer to mpeg4videodec This is possible by moving the code using it to open a GetBitContext from h263dec.

[FFmpeg-cvslog] avcodec/mpegvideo_dec: Move syncing DivX-stuff to mpeg4videodec.c

ffmpeg | branch: master | Andreas Rheinhardt | Mon Feb 24 17:47:22 2025 +0100| [920217975407bcb9f9ca9ab98a2e2dcb3a98e05a] | committer: Andreas Rheinhardt avcodec/mpegvideo_dec: Move syncing DivX-stuff to mpeg4videodec.c It is only used by MPEG-4 (and is used in h263dec.c and can therefore not

[FFmpeg-cvslog] avcodec/vc1: Add max_b_frames field to VC1Context

ffmpeg | branch: master | Andreas Rheinhardt | Mon Feb 24 19:05:32 2025 +0100| [1adc84136e499302a787537dbc9323c4abb647bf] | committer: Andreas Rheinhardt avcodec/vc1: Add max_b_frames field to VC1Context Don't reuse MpegEncContext.max_b_frames, which is supposed to be encoder-only. Reviewed-b

[FFmpeg-cvslog] avcodec/h263dec: Don't call ff_thread_finish_setup() unnecessarily

ffmpeg | branch: master | Andreas Rheinhardt | Mon Feb 24 22:16:36 2025 +0100| [e4470a8e30de8af6150b4a78ea0637ca7d62b824] | committer: Andreas Rheinhardt avcodec/h263dec: Don't call ff_thread_finish_setup() unnecessarily All hwaccels for MPEG-4/H.263 are run serially even when frame-threading

[FFmpeg-cvslog] avcodec/mpeg4videoenc: Remove dead FF_BUG_MS code

ffmpeg | branch: master | Andreas Rheinhardt | Mon Feb 24 19:09:32 2025 +0100| [78b26225bac58a8c959d8d180a7a376c21b4bbe3] | committer: Andreas Rheinhardt avcodec/mpeg4videoenc: Remove dead FF_BUG_MS code Added in 59fa3f96f48d12e189492ca3670991f91c316d4e, yet avctx->workaround_bugs is never cop

[FFmpeg-cvslog] avcodec/mpeg4videodec: Avoid copying packed bitstream data

ffmpeg | branch: master | Andreas Rheinhardt | Tue Feb 25 16:21:07 2025 +0100| [4f2becc2dc4668f837b5ba96c11c3426bc120ac0] | committer: Andreas Rheinhardt avcodec/mpeg4videodec: Avoid copying packed bitstream data This is possible because the packet is reference-counted and because we never com

[FFmpeg-cvslog] avcodec/rv34: Don't update block_index unnecessarily

ffmpeg | branch: master | Andreas Rheinhardt | Sat Jun 8 03:11:48 2024 +0200| [592d75dbebb7f7f620c0c70974f4067dd0a389c2] | committer: Andreas Rheinhardt avcodec/rv34: Don't update block_index unnecessarily It is unused by RV30 and RV40. Signed-off-by: Andreas Rheinhardt > http://git.videol

[FFmpeg-cvslog] avcodec/mpegvideo_enc: Reindent after the previous commit

ffmpeg | branch: master | Andreas Rheinhardt | Tue Feb 25 21:11:33 2025 +0100| [ca56526d7fb4de3718096871f232d980b7e7bfac] | committer: Andreas Rheinhardt avcodec/mpegvideo_enc: Reindent after the previous commit Signed-off-by: Andreas Rheinhardt > http://git.videolan.org/gitweb.cgi/ffmpeg.gi

[FFmpeg-cvslog] avcodec/mpeg12dec: Don't initialize unused parts of ScanTable

ffmpeg | branch: master | Andreas Rheinhardt | Mon Feb 24 20:43:46 2025 +0100| [79d6657e138ac4a0f52ec4eaa4f4a2fcea0a7abb] | committer: Andreas Rheinhardt avcodec/mpeg12dec: Don't initialize unused parts of ScanTable The MPEG-1/2 decoders don't need ScanTable.raster_end (as the coefficients are

[FFmpeg-cvslog] avcodec/msmpeg4dec: Remove redundant check

ffmpeg | branch: master | Andreas Rheinhardt | Tue Feb 25 21:51:20 2025 +0100| [2de9168b563c339d5888aecdd6b923e832cf97ff] | committer: Andreas Rheinhardt avcodec/msmpeg4dec: Remove redundant check Since d50635cd247e17fe16c63219b9ae80d45a8185b1 the dct_unquantize_h263_intra functions have set t

[FFmpeg-cvslog] avcodec/vc1dec: Don't initialize inter_scantable

ffmpeg | branch: master | Andreas Rheinhardt | Tue Feb 25 22:56:21 2025 +0100| [f76195ff656d6bea68feee783160652e2b3e3d60] | committer: Andreas Rheinhardt avcodec/vc1dec: Don't initialize inter_scantable It is unused. Signed-off-by: Andreas Rheinhardt > http://git.videolan.org/gitweb.cgi/ffm

[FFmpeg-cvslog] avcodec/mpegvideo_enc: Don't set qscale_table value prematurely

ffmpeg | branch: master | Andreas Rheinhardt | Sun Jun 30 16:04:13 2024 +0200| [ab768b88e069edf839aab84b73a763b3c82993b4] | committer: Andreas Rheinhardt avcodec/mpegvideo_enc: Don't set qscale_table value prematurely When there are multiple candidates for macroblock type, the encoder tries th

[FFmpeg-cvslog] avcodec/motion_est: Avoid branches for put(_no_rnd) selection

ffmpeg | branch: master | Andreas Rheinhardt | Sun Jun 30 11:31:04 2024 +0200| [d4fd475005a2febbfcd0b76c8ab3584a87430972] | committer: Andreas Rheinhardt avcodec/motion_est: Avoid branches for put(_no_rnd) selection MotionEstContext contains pointers (to function pointers) that have been set o

[FFmpeg-cvslog] tests/fate/vcodec: Test using mpeg2-quantizers for MPEG-4

ffmpeg | branch: master | Andreas Rheinhardt | Tue Jul 2 19:26:20 2024 +0200| [57ade06ffeb8d5485385693cac0f8213d517cacb] | committer: Andreas Rheinhardt tests/fate/vcodec: Test using mpeg2-quantizers for MPEG-4 Signed-off-by: Andreas Rheinhardt > http://git.videolan.org/gitweb.cgi/ffmpeg.gi

[FFmpeg-cvslog] avcodec/mips/mpegvideo_mmi: Don't check alternate_scan unnecessarily

ffmpeg | branch: master | Andreas Rheinhardt | Mon Jul 1 22:43:33 2024 +0200| [ab38ff9161b9cf52684d3ad1e2ccc60fa01eaa6d] | committer: Andreas Rheinhardt avcodec/mips/mpegvideo_mmi: Don't check alternate_scan unnecessarily Forgotten in 4f7aeffd8c3607aea6464a20eba4cc9106173c52. Reviewed-by: Ra

[FFmpeg-cvslog] avcodec/mpegvideo_dec: Use picture-dimensions in ff_print_debug_info()

ffmpeg | branch: master | Andreas Rheinhardt | Mon Jul 1 01:31:12 2024 +0200| [973c7a0c652dbac5f171e61dcf94f42de094eaa2] | committer: Andreas Rheinhardt avcodec/mpegvideo_dec: Use picture-dimensions in ff_print_debug_info() It will allow to avoid the special case for VC-1 field pictures. Sig

[FFmpeg-cvslog] avcodec/msmpeg4dec: Avoid branch

ffmpeg | branch: master | Andreas Rheinhardt | Mon Jul 1 19:52:34 2024 +0200| [18fc118f6541467d1493e71831df195f91cca1c7] | committer: Andreas Rheinhardt avcodec/msmpeg4dec: Avoid branch A non-LTO-compiler can't optimize this away. Signed-off-by: Andreas Rheinhardt > http://git.videolan.org

[FFmpeg-cvslog] avcodec/msmpeg4dec: Don't process unnecessarily many coefficients

ffmpeg | branch: master | Andreas Rheinhardt | Mon Jul 1 19:18:06 2024 +0200| [4912a2d88ea242baf966eaa7d55dd547d4785d0f] | committer: Andreas Rheinhardt avcodec/msmpeg4dec: Don't process unnecessarily many coefficients This code has been added in f5957f3fe2b8dfda13a9750acc3e3a03cd5d39c9 when

[FFmpeg-cvslog] avcodec/vc1dec: Reenable debug-info output for field pictures

ffmpeg | branch: master | Andreas Rheinhardt | Mon Jul 1 01:43:58 2024 +0200| [a885351ad0a8f569671b086bd027b43acec3e6b1] | committer: Andreas Rheinhardt avcodec/vc1dec: Reenable debug-info output for field pictures Effectively reverts c59b5e3d1e0121ea23b5b326529f5bdca44cf982. This is possible

[FFmpeg-cvslog] avcodec/h261dec: Remove dead check

ffmpeg | branch: master | Andreas Rheinhardt | Mon Jul 1 12:30:29 2024 +0200| [9c16d54a166630e960505389da5f53a6da710c62] | committer: Andreas Rheinhardt avcodec/h261dec: Remove dead check H.261 does not have non-reference frames. Signed-off-by: Andreas Rheinhardt > http://git.videolan.org/

[FFmpeg-cvslog] avcodec/mpeg12dec: Remove write-only assignments

ffmpeg | branch: master | Andreas Rheinhardt | Mon Jul 8 14:55:45 2024 +0200| [db8546dff71475ce34fd7658c3c8da53aa03afbe] | committer: Andreas Rheinhardt avcodec/mpeg12dec: Remove write-only assignments This decoder unquantizes while parsing blocks and does not use dct_unquantize_mpeg1_intra (

[FFmpeg-cvslog] avcodec/mpeg12dec: Really disable scratchpad allocations

ffmpeg | branch: master | Andreas Rheinhardt | Tue Feb 25 16:40:56 2025 +0100| [8d871622542bd596ecf041eceb11716a2b4e47bb] | committer: Andreas Rheinhardt avcodec/mpeg12dec: Really disable scratchpad allocations ba341be09533a077075c71fce5f9dc5b73504234 disabled them for the main thread only, no

[FFmpeg-cvslog] avcodec/mpegvideo_enc: Add AV_CODEC_CAP_DR1

ffmpeg | branch: master | Andreas Rheinhardt | Sun Jun 30 17:44:46 2024 +0200| [b3ba961df68437e68aa8cc0a33c5b85574935971] | committer: Andreas Rheinhardt avcodec/mpegvideo_enc: Add AV_CODEC_CAP_DR1 The mpegvideo-based encoders do one uncommon thing with the packet's data given by ff_alloc_pack

[FFmpeg-cvslog] avcodec/mpegvideo_enc: Don't do unnecessary work for AMV

ffmpeg | branch: master | Andreas Rheinhardt | Tue Feb 25 20:56:57 2025 +0100| [772df653566d2d3f69f6e84da8b884ff110253c7] | committer: Andreas Rheinhardt avcodec/mpegvideo_enc: Don't do unnecessary work for AMV Up until now, the initialization of AMV's matrices happened after the initializatio

[FFmpeg-cvslog] tests/fate/vcodec: Test H.261 loop-filter

ffmpeg | branch: master | Andreas Rheinhardt | Sat Jul 6 23:50:47 2024 +0200| [a5d590963c06b7041f41e668e2bfe04ee71fecc5] | committer: Andreas Rheinhardt tests/fate/vcodec: Test H.261 loop-filter Increases coverage. Signed-off-by: Andreas Rheinhardt > http://git.videolan.org/gitweb.cgi/ffmp

[FFmpeg-cvslog] avcodec/h261dec: Inline constant

ffmpeg | branch: master | Andreas Rheinhardt | Sun Jul 7 00:10:37 2024 +0200| [3474475e5853c2c109d90b95a299ffc380dd2b24] | committer: Andreas Rheinhardt avcodec/h261dec: Inline constant The value here has been set in ff_set_qscale() from ff_mpeg1_dc_scale_table, all of whose entries are 8. R

[FFmpeg-cvslog] avcodec/mpegvideo_dec: Don't sync max_b_frames

ffmpeg | branch: master | Andreas Rheinhardt | Mon Feb 24 16:37:19 2025 +0100| [4c0960e968228f059e88dd464b10e8b53ac24b9e] | committer: Andreas Rheinhardt avcodec/mpegvideo_dec: Don't sync max_b_frames Only the VC-1 decoder (which does not use frame-threading) and encoders use it. Signed-off-b

[FFmpeg-cvslog] tests/fate/vcodec: Test alternate_scan

ffmpeg | branch: master | Andreas Rheinhardt | Tue Jul 2 00:45:50 2024 +0200| [c960b42efca6112584ce2b25e9654bb2adb0ff12] | committer: Andreas Rheinhardt tests/fate/vcodec: Test alternate_scan Encoding was untested before this. Notice that the filesize degradation is partially due to mpegvideo

[FFmpeg-cvslog] fftools/ffmpeg_filter: also remove display matrix side data from buffered frames

ffmpeg | branch: release/7.1 | James Almer | Thu Feb 27 23:38:09 2025 -0300| [f023d2c61b3c178cb6d279ff51472ca16f239674] | committer: James Almer fftools/ffmpeg_filter: also remove display matrix side data from buffered frames Some frames may be buffered before a complex filtergraph can be conf

[FFmpeg-cvslog] avformat/rtp_av1: Add necessary headers

ffmpeg | branch: master | Andreas Rheinhardt | Tue Mar 4 15:34:38 2025 +0100| [9e2773034ba177861e32b746b5dbd8d0370661a5] | committer: Andreas Rheinhardt avformat/rtp_av1: Add necessary headers Also use a void* instead of an AVFormatContext as logctx in parse_leb. Signed-off-by: Andreas Rhein

[FFmpeg-cvslog] avcodec: Remove leftover alpha declarations

ffmpeg | branch: master | Andreas Rheinhardt | Sun Mar 2 05:18:43 2025 +0100| [19e907cd725eba0246ca6bfe9526f8bbde0c5729] | committer: Andreas Rheinhardt avcodec: Remove leftover alpha declarations Forgotten in cdd139d760688b14849d02ee1907f68fe692c24e. Signed-off-by: Andreas Rheinhardt > ht

[FFmpeg-cvslog] avcodec/mpeg12dec: Use proper logcontext

ffmpeg | branch: master | Andreas Rheinhardt | Tue Mar 4 23:06:38 2025 +0100| [883120aaf04a42252ff95d0f48934006feb8a94c] | committer: Andreas Rheinhardt avcodec/mpeg12dec: Use proper logcontext When used with the "mpegvideo" or "mpeg1video" decoder (which do not have a private class) this wou

[FFmpeg-cvslog] avformat/mpegenc, mpegts. mxfenc: Mark (de)muxers declarations as internal

ffmpeg | branch: master | Andreas Rheinhardt | Sun Mar 2 15:56:31 2025 +0100| [56cfbe3c9fcc45b95a5ef0e103721aec348e32de] | committer: Andreas Rheinhardt avformat/mpegenc, mpegts. mxfenc: Mark (de)muxers declarations as internal Otherwise compilers might emit code that presumes there to be a G

[FFmpeg-cvslog] avutil/attributes_internal: Add EXTERN macro for extern+hidden

ffmpeg | branch: master | Andreas Rheinhardt | Sun Mar 2 15:27:48 2025 +0100| [57f184221df261e041fcacc29fc1d4da196436ef] | committer: Andreas Rheinhardt avutil/attributes_internal: Add EXTERN macro for extern+hidden This is inspired by the equivalent dav1d attribute introduced by Henrik Gramn

[FFmpeg-cvslog] aarch64/hevcdsp_idct_neon: Optimize idct dc

ffmpeg | branch: master | Zhao Zhili | Thu Feb 20 00:50:23 2025 +0800| [5977bff569b1e1feecf681059472f4106c1573cb] | committer: Zhao Zhili aarch64/hevcdsp_idct_neon: Optimize idct dc clang does better than the assembly code before the patch, especially for small size: hevc_idct_4x4_dc_8_c:

[FFmpeg-cvslog] avformat/movenc: Add AVS3 support

ffmpeg | branch: master | Zhao Zhili | Tue Feb 18 19:24:48 2025 +0800| [a053516e644e07d1e625901539671c3682c7af49] | committer: Zhao Zhili avformat/movenc: Add AVS3 support 'avs3' is registered at mp4ra.org. The Avs3ConfigurationBox 'av3c' inside 'avs3' hasn't been registered yet, but is specif

[FFmpeg-cvslog] avformat/isom_tags: Add tag for AVS3

ffmpeg | branch: master | Zhao Zhili | Mon Feb 24 21:51:46 2025 +0800| [2d7966aee158682d3112c0f293254a96ef26950c] | committer: Zhao Zhili avformat/isom_tags: Add tag for AVS3 Signed-off-by: Zhao Zhili > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=2d7966aee158682d3112c0f293254a9

[FFmpeg-cvslog] avcodec/libuavs3d: process extradata

ffmpeg | branch: master | Zhao Zhili | Mon Feb 24 21:51:51 2025 +0800| [5a32496962853beb302ebce307821c7c960fa54f] | committer: Zhao Zhili avcodec/libuavs3d: process extradata Signed-off-by: Zhao Zhili > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=5a32496962853beb302ebce307821c7

[FFmpeg-cvslog] aarch64/hevcdsp_idct_neon: Add implementation for idct dc 12

ffmpeg | branch: master | Zhao Zhili | Thu Feb 20 00:50:31 2025 +0800| [3e9777dc7546425651bc265a49187d6bbb42bdf4] | committer: Zhao Zhili aarch64/hevcdsp_idct_neon: Add implementation for idct dc 12 Reduce binary size at the same time. The performance compared to clang -O3 is the same. Review

[FFmpeg-cvslog] avcodec/exr: do not output 32bit floats when a file stores 16bit floats

ffmpeg | branch: master | Michael Niedermayer | Thu Jan 9 23:08:23 2025 +0100| [0e917389fe73c932049635d947bba076f1709589] | committer: Michael Niedermayer avcodec/exr: do not output 32bit floats when a file stores 16bit floats exr should not do a internal pixel format convert Sponsored-by: S