avcodec/mxpegdec: Fix memleaks upon init failure

Reviewed-by: Anton Khirnov <anton@khirnov.net>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/mjpegdec: Fix memleak upon init failure

This affected all decoders that used ff_mjpeg_decode_init() as init
function; and it also affected decoders that open jpeg decoders via
ff_codec_open2_recursive() as well as MxPEG.

Reviewed-by: Anton Khirnov <anton@khirnov.net>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/smc: Mark decoder as init-threadsafe

Reviewed-by: Anton Khirnov <anton@khirnov.net>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/sipr: Mark decoder as init-threadsafe

Reviewed-by: Anton Khirnov <anton@khirnov.net>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/shorten: Mark decoder as init-threadsafe

Reviewed-by: Anton Khirnov <anton@khirnov.net>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/sgirledec: Mark decoder as init-threadsafe

Reviewed-by: Anton Khirnov <anton@khirnov.net>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/sgienc: Combine av_log() statements

Reviewed-by: Anton Khirnov <anton@khirnov.net>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/sgienc: Mark encoder as init-threadsafe

Reviewed-by: Anton Khirnov <anton@khirnov.net>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/sgidec: Mark decoder as init-threadsafe

Reviewed-by: Anton Khirnov <anton@khirnov.net>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/tmv: Mark decoder as init-threadsafe

Reviewed-by: Anton Khirnov <anton@khirnov.net>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/ulti: Mark decoder as init-threadsafe

Reviewed-by: Anton Khirnov <anton@khirnov.net>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/ws-snd1: Mark decoder as init-threadsafe

Reviewed-by: Anton Khirnov <anton@khirnov.net>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/xxan: Mark decoder as init-threadsafe

Reviewed-by: Anton Khirnov <anton@khirnov.net>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/xxan: Cleanup generically on init failure

Reviewed-by: Anton Khirnov <anton@khirnov.net>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/xsubenc: Mark decoder as init-threadsafe

Reviewed-by: Anton Khirnov <anton@khirnov.net>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/xsubdec: Mark decoder as init-threadsafe

Reviewed-by: Anton Khirnov <anton@khirnov.net>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/xl: Mark decoder as init-threadsafe

Reviewed-by: Anton Khirnov <anton@khirnov.net>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/xfacedec: Mark decoder as init-threadsafe

Reviewed-by: Anton Khirnov <anton@khirnov.net>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/xan: Mark decoder as init-threadsafe

Reviewed-by: Anton Khirnov <anton@khirnov.net>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/xan: Cleanup generically on init failure

Reviewed-by: Anton Khirnov <anton@khirnov.net>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/yuv4dec: Mark decoder as init-threadsafe

Reviewed-by: Anton Khirnov <anton@khirnov.net>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/y41penc: Mark encoder as init-threadsafe

Reviewed-by: Anton Khirnov <anton@khirnov.net>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/y41pdec: Mark decoder as init-threadsafe

Reviewed-by: Anton Khirnov <anton@khirnov.net>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/yop: Mark decoder as init-threadsafe

Reviewed-by: Anton Khirnov <anton@khirnov.net>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/cljrdec: Mark decoder as init-threadsafe

Reviewed-by: Anton Khirnov <anton@khirnov.net>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/loco: Mark decoder as init-threadsafe

Reviewed-by: Anton Khirnov <anton@khirnov.net>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/pcm: Mark decoders as init-threadsafe

Reviewed-by: Anton Khirnov <anton@khirnov.net>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/pcm: Make encoders init-threadsafe

Reviewed-by: Anton Khirnov <anton@khirnov.net>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/adpcmenc: Mark encoders as init-threadsafe

Reviewed-by: Anton Khirnov <anton@khirnov.net>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/adpcm: Mark decoders as init-threadsafe

They don't modify any global state

Reviewed-by: Anton Khirnov <anton@khirnov.net>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/cdtoons: Mark decoder as init-threadsafe

Reviewed-by: Anton Khirnov <anton@khirnov.net>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/a64multienc: Mark encoders as init-threadsafe

Reviewed-by: Anton Khirnov <anton@khirnov.net>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avfilter/af_stereotools: add support for commands

avfilter/af_stereotools: add timeline support

avfilter/af_stereotools: add two more modes

fftools/cmdutils: also print warnings when using -sinks and -sources

Signed-off-by: Marton Balint <cus@passwd.hu>

avdevice/decklink: warn about too old decklink API version

Signed-off-by: Marton Balint <cus@passwd.hu>

avdevice/decklink_enc: simplify usage of buffercount type

Also remove some leftover declaration of CreateDeckLinkIteratorInstance().

Signed-off-by: Marton Balint <cus@passwd.hu>

avdevice/decklink_dec: add support for 50/60 fps timecode

Signed-off-by: Marton Balint <cus@passwd.hu>

avutil/timecode: add av_timecode_init_from_components

Signed-off-by: Marton Balint <cus@passwd.hu>

avutil/timecode: allow drop frame timecodes for multiples of 30000/1001 fps

Signed-off-by: Marton Balint <cus@passwd.hu>

avutil/film_grain_params: fix doxy for ar_coeff_* fields

Signed-off-by: James Almer <jamrial@gmail.com>

avcodec/pthread_slice: Don't use static variable, fix race

ff_slice_thread_init() uses a static variable to hold a function
pointer, although the value of said pointer needn't be saved between
different runs of this function at all.

The reason for this being so is probably that said pointer points to
a static function (if used); but storage class specifiers like "static"
are not part of the type of an object and so including it in the pointer
declaration is wrong (anyway, "static" means different things in both
contexts: for the function declaration it affects linkage, for the
variable storage duration).

Using a static variable here can lead to races, e.g. when initializing
VP9 (for which said function pointer was added) and H.264 with slice
threading. The latter has the FF_CODEC_CAP_INIT_THREADSAFE flag set and
is therefore unaffected by the lock guarding initializations of
decoders.

Reviewed-by: Anton Khirnov <anton@khirnov.net>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/mpeg12enc: Disallow using MPEG-2 intra VLC table for mpeg1video

Using MPEG-2 intra VLC tables is spec-incompliant for MPEG-1 and given
that an MPEG-1 bitstream can't signal whether MPEG-2 intra VLC tables
have been used the output is broken. Therefore this option is removed
immediately without any deprecation period.

Reviewed-by: James Almer <jamrial@gmail.com>
Reviewed-by: Marton Balint <cus@passwd.hu>
Reviewed-by: Anton Khirnov <anton@khirnov.net>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avfilter/af_acrossover: add per output band gain

avutil/avsscanf: fix possible overreads when dealing with %c or %s

avfilter/af_acrossover: remove pointless allocation of static size

avfilter/af_acrossover: document roll-off of each filter order

avfilter/af_acrossover: move coefficients and state to simple arrays

avfilter/af_anequalizer: add timeline and slice support

avcodec/cbs_h2645: remove dead code in cbs_h2645_replace_ps()

Missed in 0720e15c0330f0f101c02d5c2c6548e964079f6d. PS units are now guaranteed
to always be reference counter.

Signed-off-by: James Almer <jamrial@gmail.com>

avcodec/cbs_av1: ensure Sequence Header unit content is always reference counted

Prevents a NULL pointer dereference.

Signed-off-by: James Almer <jamrial@gmail.com>

avcodec/vaapi_av1: read film grain param values from AV1Frames

Signed-off-by: James Almer <jamrial@gmail.com>

avcodec/dxva2_av1: read film grain param values from AV1Frames

Signed-off-by: James Almer <jamrial@gmail.com>

avcodec/nvdec_av1: read film grain param values from AV1Frames

Signed-off-by: James Almer <jamrial@gmail.com>

avcodec/av1dec: infer and store film grain param values in AV1Frame

They are not always coded in the bistream for each frame. In some cases, the
values need to be taken from a reference frame.

See section 6.8.20 from the AV1 spec.

Signed-off-by: James Almer <jamrial@gmail.com>

avcodec/cbs_av1: split film grain param fields into their own struct

Cosmetic change in preparation for the following patches.

Signed-off-by: James Almer <jamrial@gmail.com>

Revert "avcodec/nvdec_av1: fix setting film grain parameters for frames with update_grain == 0"

This reverts commit f9eec6298387fe72cd8035ff45276cfc3da784a8.

This does not effectively cover all cases. The values for some frames need
to be inferred by the decoder.

Signed-off-by: James Almer <jamrial@gmail.com>

avfilter/af_agate: add support for commands and timeline

avfilter/af_acrossover: remove unneeded emms_c() calls

avfilter/af_acrossover: unroll biquad_process loop

Makes code significantly faster for higher orders.

avfilter/af_acrossover: use fdsp for inverter

avfilter/af_acrossover: always align nb_samples for fdsp call

avfilter/af_asupercut: add float sample format support

avcodec/hevc_cabac: Limit value in coeff_abs_level_remaining_decode() tighter

The max depth is 16bps, the max allowed coefficient depth is depth+6
Fixes: signed integer overflow: 1074266112 + 1073725439 cannot be represented in type 'int'
Fixes: 26493/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_HEVC_fuzzer-5657763331702784
Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

avformat/cafdec: Check the return code from av_add_index_entry()

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

avformat/cafdec: Check for EOF in index read loop

Fixes: OOM
Fixes: 27398/clusterfuzz-testcase-minimized-ffmpeg_dem_CAF_fuzzer-541296033975500
Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

avformat/cafdec: Check that bytes_per_packet and frames_per_packet are non negative

These fields are not signed in the spec (1.0) so they cannot be negative
Changing bytes_per_packet to unsigned would not solve this as it is exported
as block_align which is signed

Fixes: Infinite loop
Fixes: 26492/clusterfuzz-testcase-minimized-ffmpeg_dem_CAF_fuzzer-5632087614554112
Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

avformat/mpegts: Limit iterations of get_packet_size()

This is a pathological case where the fuzzer provides only 2 bytes per iteration.

Fixes: Timeout (>30 -> 0.9sec)
Fixes: 26488/clusterfuzz-testcase-minimized-ffmpeg_dem_MPEGTS_fuzzer-5911031077142528
Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

avcodec/notchlc: Check available space for luma block code and alpha code

Fixes: Timeout (too long -> 2sec)
Fixes: 25439/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_NOTCHLC_fuzzer-5688211127664640
Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

avfilter/af_asupercut: add order option

avfilter/af_acrossover: split in correct spot

Previously split was made in wrong (half of original) position.

avfilter/af_acrossover: reduce memory usage

Split biquad coefficients from biquad state.

avcodec/pafvideo: Cleanup generically on init failure

Reviewed-by: Paul B Mahol <onemda@gmail.com>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/pnmenc: Mark encoders as init-threadsafe

Reviewed-by: Michael Niedermayer <michael@niedermayer.cc>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/photocd: Mark decoder as init-threadsafe

Reviewed-by: Paul B Mahol <onemda@gmail.com>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/pcm-dvdenc: Mark encoder as init-threadsafe

Reviewed-by: Paul B Mahol <onemda@gmail.com>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/pcm-dvd: Mark decoder as init-threadsafe

Reviewed-by: Paul B Mahol <onemda@gmail.com>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/pafvideo: Mark decoder as init-threadsafe

Reviewed-by: Paul B Mahol <onemda@gmail.com>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/pafaudio: Mark decoder as init-threadsafe

Reviewed-by: Paul B Mahol <onemda@gmail.com>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/cdxl: Mark decoder as init-threadsafe

Reviewed-by: Paul B Mahol <onemda@gmail.com>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/svq1dec: Make decoder init-threadsafe

Reviewed-by: Michael Niedermayer <michael@niedermayer.cc>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avcodec/a64multienc: Fix memleak upon init failure

Reviewed-by: Paul B Mahol <onemda@gmail.com>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>

avfilter/af_asubboost: make wet option apply to final output

Also changes some default values for options after this change.
This makes distinction between feedback and wet option.
Before they would produce same output if values were swapped.

avfilter/avf_showfreqs: add group delay data mode

avformat/rtspdec: fix potential mem leak in listen mode

Currently a repeating setup request (with the same stream id) will
simply overwrite rtp_handle/transport_priv without freeing the
resources first. This is fixed by closing the previous setup request.

Reviewed-by: Martin Storsjö <martin@martin.st>
Signed-off-by: Andriy Gelman <andriy.gelman@gmail.com>

avformat/rtspdec: show method request in log

Reviewed-by: Martin Storsjö <martin@martin.st>
Signed-off-by: Andriy Gelman <andriy.gelman@gmail.com>

avfilter/af_acrossover: add option to adjust input gain

avfilter/af_acrossover: really fix single-pole allpass coefficients

Now LR2, LR6, LR10, LR14 and LR18 have really flat sum.

doc/filters: mention when adenorm filter should be placed in graph

avfilter/af_acrossover: remove unecessary allpass calls

avcodec/utils: Require a non zero channels unless AV_CODEC_CAP_CHANNEL_CONF is set

This avoids per codec checks for channels not being 0

Fixes: division by 0
Fixes: 25419/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_FASTAUDIO_fuzzer-5632544761184256
Fixes: 25433/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_FASTAUDIO_fuzzer-6215671900536832
Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Suggested-by: Paul B Mahol <onemda@gmail.com>
See: [FFmpeg-devel] [PATCH 1/3] avcodec/fastaudio: Check channel
Reviewed-by: Paul B Mahol <onemda@gmail.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

avcodec: Set AV_CODEC_CAP_CHANNEL_CONF in more decoders

Suggested-by: Paul B Mahol <onemda@gmail.com>
See: [FFmpeg-devel] [PATCH 1/3] avcodec/fastaudio: Check channels
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

avfilter/af_aemphasis: add timeline/slice and commands support

avfilter/af_aemphasis: switch to planar sample format

doc/filters: fix several errors/typos

avfilter/af_acrossover: add support for float sample format

avcodec: deprecate thread_safe_callbacks

They add considerable complexity to frame-threading implementation,
which includes an unavoidably leaking error path, while the advantages
of this option to the users are highly dubious.

It should be always possible and desirable for the callers to make their
get_buffer2() implementation thread-safe, so deprecate this option.

avfilter/af_acrossover: fix single pole allpass coefficients

doc/filters: document adenorm commands support