fate: Add fate-prores-gray

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

avcodec/proresdec2: Add support for grayscale videos

Signed-off-by: Håvard Espeland <espeland@mixedrealities.no>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

avfilter/avfiltergraph: Clear graph pointers in ff_filter_graph_remove_filter()

When a filter is no longer part of a graph, its pointers should be cleared
so no stale pointers remain.

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

Merge commit '44d16df413878588659dd8901bba016b5a869fd1'

* commit '44d16df413878588659dd8901bba016b5a869fd1':
  h264_parser: eliminate H264SliceContext usage

Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>

ffserver: fixed deallocation bug in build_feed_streams

Signed-off-by: Gregor Riepl <onitake@gmail.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

avformat/utils: Fix use of uninitialized variable

Fixes CID1361961

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

avcodec/dca: move EXSS sampling frequency arrays to dca.c

Avoids unwanted parser dependency on dcadata.

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

avcodec/vble: add frame threading support

Signed-off-by: Paul B Mahol <onemda@gmail.com>

doc/developer.texi: Add a code of conduct

See: [FFmpeg-devel] [Vote] Code of Conduct

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

avfilter/src_movie: add various commands

add seek command
add get_duration command

Update to codecpar by commiter

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

lavc/mediacodec: add missing MediaCodec.Get{Input,Output}Buffer() checks

lavc/mediacodec: factorize static fields initialization

avcodec/Makefile: add missing dependencies to dca parser

They are needed starting with the previous commit

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

avcodec/dca_parser: set duration for core-less streams

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

avcodec/dca_exss: prepare to be called from parser context

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

avcodec/nvenc: Require Maxwell for lossless

avcodec/nvenc: CUDA frames support

Signed-off-by: Timo Rothenpieler <timo@rothenpieler.org>

avcodec/nvenc: replace custom FIFOs with AVFifos

Signed-off-by: Timo Rothenpieler <timo@rothenpieler.org>

avcodec/nvenc: combine input and output surface structures

There is no point in separate structures as they have 1:1 relationship,
they are always used together and they have same lifetime.

Signed-off-by: Timo Rothenpieler <timo@rothenpieler.org>

avcodec/nvenc: split large functions into smaller ones

Functions names and scopes are from libav. This commit basically moves
code around without changing it; it shouldn't change any functionality
except some small leak fixes on error paths.

Signed-off-by: Timo Rothenpieler <timo@rothenpieler.org>

avformat/udp: Close the socket after destroying the thread using the socket

for send() this cannot work and even for recv() it feels very wrong and hackish

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

avformat/udp: redesign threaded udp tx code

This fixes partially completed send()
Avoids holding the mutex during send()
fixes race conditions in error handling
removes copied non thread specific blocking code
Fixes deadlocks on closure
Fixes data loss on closure

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

avformat/udp: Add a delay between packets for streaming to clients with short buffer

This commit enables sending UDP packets in a background thread with specified delay.
When sending packets without a delay some devices with small RX buffer
( MAG200 STB, for example) will drop tail packets in bursts causing
decoding errors.

To use it specify "fifo_size" with "packet_gap" .

The output url will looks like udp://xxx:yyy?fifo_size=<output fifo
size>&packet_gap=<delay in usecs>

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

libavcodec/dv_tablegen: fixed wrong printf format

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

libavfilter/vf_hwdownload: show error when ff_formats_ref fails

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

avfilter/af_hdcd: Replace complexer expression by litteral constant

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

avformat/mov: Avoid "Unintended sign extension"

Fixes CID1361957

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

avcodec/mips/idctdsp_mmi: Version 2 of the optimizations for loongson mmi

    1. no longer use the register names directly and optimized code format
    2. to be compatible with O32, specify type of address variable with mips_reg and handle the address variable with PTR_ operator

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

avcodec/mips/pixblockdsp_mmi: Version 2 of the optimizations for loongson mmi

    1. no longer use the register names directly and optimized code format
    2. to be compatible with O32, specify type of address variable with mips_reg and handle the address variable with PTR_ operator

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

avcodec/mips/blockdsp_mmi: Version 2 of the optimizations for loongson mmi

    1. no longer use the register names directly and optimized code format
    2. to be compatible with O32, specify type of address variable with mips_reg and handle the address variable with PTR_ operator

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

avcodec/dca_exss: always update downmix flags

Fixes potential parsing failure when total number of channels indicated
by EXSS asset descriptor decreases mid-stream and stale embedded downmix
flags are referenced.

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

avfilter/vf_blend: fix incorrect Y variable when threading is used

Signed-off-by: Paul B Mahol <onemda@gmail.com>

libavutil/opencl: fixed uninitialized var warning

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

libavcodec/cfhd: fixed wrong printf format

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

libavfilter/af_hdcd: fixed negative signed value shift

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

fate: add aecho test

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

avcodec/truemotion2rt: also export key frame info

Signed-off-by: Paul B Mahol <onemda@gmail.com>

avcodec/iff: mention RGB8/RGBN decoder

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

avcodec/mvcdec: simplify, no need to use reget buffer

Signed-off-by: Paul B Mahol <onemda@gmail.com>

avformat/matroskaenc: reindent after last commit

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

avformat/matroskaenc: wrap par->format use as sample format under codec_type == AVMEDIA_TYPE_AUDI

Fixes CID1361946

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

avformat/avienc: Fix memleaks on errors

Fixes CID1361951

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

avformat/avienc: Rename reshuffles return variable to ensure it is not mixed up

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

avfilter/af_loudnorm: fix crash when ebur128 initialization was not successfull/complete

Signed-off-by: Paul B Mahol <onemda@gmail.com>

avformat: add MTAF demuxer

Signed-off-by: Paul B Mahol <onemda@gmail.com>

avcodec: add adpcm MTAF decoder

Signed-off-by: Paul B Mahol <onemda@gmail.com>

avfilter/avf_ahistogram: raise minimum acmax to 1

If acmax can be 0 then it could result in a division by 0
Fixes CID1351345

Reviewed-by: Paul B Mahol <onemda@gmail.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

avcodec/aic: add frame threading support

Signed-off-by: Paul B Mahol <onemda@gmail.com>

avcodec/mips: loongson optimize h264pred with mmi v3

1. no longer use the register names directly and optimized code format
2. to be compatible with O32, specify type of address variable with mips_reg and handle the address variable with PTR_ operator
3. ff_pred16x16_plane_ functions only support N64 ABI now

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

avcodec/mips: loongson optimize hpeldsp with mmi v1

1.the codes are compatible with O32 ABI
2.use uld and mtc1 to workaround cpu 3A2000 gslwlc1 bug (gslwlc1 instruction extension bug in O32 ABI)

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

avcodec/mips/mpegvideo_mmi: Version 2 of the optimizations for loongson mmi

1. no longer use the register names directly and optimized code format
2. to be compatible with O32, specify type of address variable with mips_reg and handle the address variable with PTR_ operator

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

avcodec/sgirledec: simplify, no need to use reget buffer

Signed-off-by: Paul B Mahol <onemda@gmail.com>

avcodec/mips/h264chroma_mmi: Version 2 of the optimizations for loongson mmi

1. no longer use the register names directly and optimized code format
2. to be compatible with O32, specify type of address variable with mips_reg and handle the address variable with PTR_ operator
3. use uld and mtc1 to workaround cpu 3A2000 gslwlc1 bug (gslwlc1 instruction extension bug in O32 ABI)

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

avcodec/dca: use LUT for LBR frequency ranges

Values for unsupported frequencies > 48000 Hz are still included (parser
will make use of them).

Also convert sampling frequencies array to unsigned.

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

avcodec/dca: simplify 'residual ok' flag tracking

Move this from separate structure field to a packet flag.

Behavior should be equivalent, except that residual flag is now properly
cleared when packet has no core frame at all.

Also print a message when forcing recovery mode due to invalid residual
to make debugging easier.

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

avcodec/dca_parser: improve frame end search

Parse core frame size directly when searching for frame end instead of
using value extracted from previous frame.

Account for unused bits when calculating sync word distance for 14-bit
streams to avoid alias sync detection.

Parse EXSS frame size and skip over EXSS frame to avoid alias sync
detection.

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

avcodec/dca: don't set initial sample_fmt

Valid sample_fmt will be set by dcadec_decode_frame() based on stream type.

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

avcodec/dca: remove useless debug message

Most DTS-in-WAV streams trigger this, making debug output hard to read.

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

avutil/eval-test: Check av_expr_parse_and_eval() for failure and also check it in the fate test

Fixes CID1361940

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

avcodec/dca: remove Rice code length limit

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

doc/filters: fix order of tinterlace filter drop modes

Signed-off-by: Paul B Mahol <onemda@gmail.com>

avcodec/utils: Move avctx->codec check before its use

Fixes CID1361954

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

avcodec/adpcm: pick correct step_index for IMA AMV

Fixes #5538

Signed-off-by: Paul B Mahol <onemda@gmail.com>

avcodec/options-test: Check avcodec_open2() return value

Fixes: CID1361937
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

lavf/mpegts: add ID3 entry to the REGD_types array

This allows to recognize ID3 packets from their corresponding descriptor
tag.

lavf/mpegtsenc: set metadata stream type and write descriptor for ID3 packets

This allow to remux data packets which are then recognized as ID3 packets.

avcodec/diracdec: Fix potential integer overflow

Fixes CID1361948

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

avformat/iff: parse DPAN chunk

Signed-off-by: Paul B Mahol <onemda@gmail.com>

avcodec/iff: add support for ANBR files

Signed-off-by: Paul B Mahol <onemda@gmail.com>

avformat/dtshddec: move data_start outside of DTSHDDemuxContext

It's only used in dtshd_read_header()

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

avcodec/exr: Fix potential integer overflow

Fixes CID1361949

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

avcodec/h264: Check init_get_bits8() for failure

Fixes CID1361935

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

ffmpeg_filter: Fix AVBufferSrcParameters leak

Fixed: CID1361950

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

avformat/avidec: Detect index with too short entries

Fixes Ticket5498

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

avcodec/m101: Ask for samples with too small extradata_size

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

avfilter: add loudnorm

Signed-off-by: Kyle Swanson <k@ylo.ph>

avcodec/m101: Check BPS value

Fixes null pointer dereference
Fixes Ticket5520

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

avcodec/iff: add forgotten checks for end of input buffer

Signed-off-by: Paul B Mahol <onemda@gmail.com>

avcodec/iff: scala anim32 & anim16 decompression

Signed-off-by: Paul B Mahol <onemda@gmail.com>

avformat/avio: Fix null pointer dereference in case of memleak

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

avformat/avio: fix memory leak in url_find_protocol

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

avformat/utils: Check negative bps before shifting in ff_get_pcm_codec_id()

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

Merge commit '71d3305c2711d4f6ec8b92db09ff64cf4e19a58e'

* commit '71d3305c2711d4f6ec8b92db09ff64cf4e19a58e':
  h264_parse: make sure the ref count is zeroed on all failure paths

Merged-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>

Merge commit 'a6e27f7add2698fdd89911632b570c3d0c3f2aaa'

* commit 'a6e27f7add2698fdd89911632b570c3d0c3f2aaa':
  h264: factor out parsing the reference count into a separate file

Merged-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>

Merge commit '56b17a33f231859cbccbd741b4763617cb4ecf03'

* commit '56b17a33f231859cbccbd741b4763617cb4ecf03':
  h264: stop testing whether the reference count changes in ff_set_ref_count()

Merged-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>

Merge commit 'e9f884416c76558098fc1f63825b293956e8483c'

* commit 'e9f884416c76558098fc1f63825b293956e8483c':
  h264: move reading direct_spatial_mv_pred out of ff_set_ref_count()

Merged-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>

Merge commit '8d0cc8ca97678f4ca87948ebabcbaab5a4f4c1f6'

* commit '8d0cc8ca97678f4ca87948ebabcbaab5a4f4c1f6':
  h264_parser: switch to h2645_parse for NAL unescaping

Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>

h2645_parse: support badly muxed mp4 streams

Some streams contain an additional AnnexB NAL inside the mp4/nalff NALU.
This commonly occurs in interlaced streams where both fields are packed
into the same MP4 NAL with an AnnexB startcode in between.

Port handling of this format from the previous h264 nal handling.

Fixes trac #5529

avcodec/iff: fix artifacts with uncompressed ham ANIMs

Signed-off-by: Paul B Mahol <onemda@gmail.com>

avcodec/mips/h264dsp_mmi: Version 2 of the optimizations for loongson mmi

1. no longer use the register names directly and optimized code format
2. to be compatible with O32, specify type of address variable with mips_reg and handle the address variable with PTR_ operator
3. optimize some unaligned loads and stores
4. use uld and mtc1 to workaround cpu 3A2000 gslwlc1 bug (gslwlc1 instruction extension bug in O32 ABI)

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

avformat/dtshddec: implement AUPR_HDR chunk parsing

Use it to get stream duration, sample rate, channel count and initial padding

Reviewed-by: Paul B Mahol <onemda@gmail.com>
Signed-off-by: James Almer <jamrial@gmail.com>

avformat/dtshddec: parse chunks stored after audio data

Reviewed-by: Paul B Mahol <onemda@gmail.com>
Signed-off-by: James Almer <jamrial@gmail.com>

avcodec/iff: add support for ANIMs with keyframes and concatenated ANIMs

Signed-off-by: Paul B Mahol <onemda@gmail.com>

avcodec/iff: check if there is enough bytes left

Signed-off-by: Paul B Mahol <onemda@gmail.com>

avformat/utils: Do not compute the bitrate from duration == 0

Fixes division by 0 in fate-acodec-ra144

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

ffmpeg: Check that r_frame_rate is set before attempting to use it

Avoids unexpected occurance and dependency on NaN behavior and divisions by 0

Testcase: fate-lavf-fate-avi_cram

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

swresample/resample: Fix division by 0 with tap_count=1

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

swresample/rematrix: Use clipping s16 rematrixing if overflows are possible

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

avformat/adxdec: set bit_rate, fixes duration calculation

Signed-off-by: Paul B Mahol <onemda@gmail.com>

swresample/rematrix: Use error diffusion to avoid error in the DC component of the matrix

This fixes the sum of the integer coefficients ending up summing to a value
larger than the value representing unity.

This issue occurs with qN0.dts when converting to stereo

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