avcodec/mpeg12: Don't initialize encoder-only parts of RLTable

ff_mpeg12_init_vlcs() currently initializes index_run, max_level and
max_run of ff_rl_mpeg1/2; yet the only user of these fields is the
MPEG-1/2 encoder which already initializes these tables on its own.
So remove the initializations in ff_mpeg12_init_vlcs(); this also
simplifies making ff_mpeg12_init_vlcs() thread-safe.

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

avcodec/rl: Remove nonstatic storage of RL tables

ff_rl_free() was added in 324e50ee95929a9491b855c5e15451145bd5d1ec,
but never used; nonstatic storage was not used long before that.

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

dnn_interface.h: fix redefining typedefs

Signed-off-by: Guo, Yejun <yejun.guo@intel.com>

dnn/queue: add error check and cleanup

Signed-off-by: Guo, Yejun <yejun.guo@intel.com>

dnn: fix issue when pthread is not supported

Signed-off-by: Guo, Yejun <yejun.guo@intel.com>

dnn: fix redefining typedefs and also refine naming with correct prefix

The prefix for symbols not exported from the library and not
local to one translation unit is ff_ (or FF for types).

Signed-off-by: Guo, Yejun <yejun.guo@intel.com>

hwcontext_drm: make dependency on Linux kernel headers optional

This was introduced in version 4.6. And may not exist all without an
optional package. So to prevent a hard dependency on needing the Linux
kernel headers to compile, make this optional.

Also ignore the status of the ioctl, since it may fail on older kernels
which don't support it. It's okay to ignore as its not fatal and any
serious errors will be caught later by the mmap call.

avfilter/vf_w3fdif: add support for commands

avfilter/vf_xmedian: allow to control eof handling

avfilter/vf_xmedian: add timeline support

vf_dnn_processing.c:  add async support

Signed-off-by: Xie, Lin <lin.xie@intel.com>
Signed-off-by: Wu Zhiwen <zhiwen.wu@intel.com>
Signed-off-by: Guo, Yejun <yejun.guo@intel.com>

dnn_interface: change from 'void *userdata' to 'AVFilterContext *filter_ctx'

'void *' is too flexible, since we can derive info from
AVFilterContext*, so we just unify the interface with this data
structure.

Signed-off-by: Xie, Lin <lin.xie@intel.com>
Signed-off-by: Wu Zhiwen <zhiwen.wu@intel.com>
Signed-off-by: Guo, Yejun <yejun.guo@intel.com>

dnn: add async execution support for openvino backend

Signed-off-by: Xie, Lin <lin.xie@intel.com>
Signed-off-by: Wu Zhiwen <zhiwen.wu@intel.com>
Signed-off-by: Guo, Yejun <yejun.guo@intel.com>

dnn_interface: add interface to support async execution

Signed-off-by: Xie, Lin <lin.xie@intel.com>
Signed-off-by: Wu Zhiwen <zhiwen.wu@intel.com>
Signed-off-by: Guo, Yejun <yejun.guo@intel.com>

dnn_backend_openvino.c: refine code for error handle

Signed-off-by: Xie, Lin <lin.xie@intel.com>
Signed-off-by: Wu Zhiwen <zhiwen.wu@intel.com>
Signed-off-by: Guo, Yejun <yejun.guo@intel.com>

dnn_backend_openvino.c: separate function execute_model_ov

function fill_model_input_ov and infer_completion_callback are
extracted, it will help the async execution for reuse.

Signed-off-by: Xie, Lin <lin.xie@intel.com>
Signed-off-by: Wu Zhiwen <zhiwen.wu@intel.com>
Signed-off-by: Guo, Yejun <yejun.guo@intel.com>

dnn/queue: add queue and safe_queue support

Signed-off-by: Xie, Lin <lin.xie@intel.com>
Signed-off-by: Wu Zhiwen <zhiwen.wu@intel.com>
Signed-off-by: Guo, Yejun <yejun.guo@intel.com>

vf_dnn_processing.c: replace filter_frame with activate func

with this change, dnn_processing can use DNN async interface later.

Signed-off-by: Xie, Lin <lin.xie@intel.com>
Signed-off-by: Wu Zhiwen <zhiwen.wu@intel.com>
Signed-off-by: Guo, Yejun <yejun.guo@intel.com>

avcodec/codec_desc: describe AV_CODEC_ID_MPEG2TS

Signed-off-by: Aman Karmani <aman@tmm1.net>

avformat/rtsp: add satip_raw flag to receive raw mpegts stream

This can be used to receive the raw mpegts stream from a SAT>IP
server, by letting avformat handle the RTSP/RTP/UDP negotiation
and setup, but then simply passing the MP2T stream through
instead of demuxing it further.

For example, this command would demux/remux the mpegts stream:

    SATIP_URL='satip://192.168.1.99:554/?src=1&freq=12188&pol=h&ro=0.35&msys=dvbs&mtype=qpsk&plts=off&sr=27500&fec=34&pids=0,17,18,167,136,47,71'
    ffmpeg -i $SATIP_URL -map 0 -c copy -f mpegts -y remux.ts

Whereas this command will simply write out the raw stream, with
the original PAT/PMT/PIDs intact:

    ffmpeg -rtsp_flags satip_raw -i $SATIP_URL -map 0 -c copy -f data -y raw.ts

Signed-off-by: Aman Karmani <aman@tmm1.net>

avformat/rtsp: add support for satip://

The SAT>IP protocol[1] is similar to RTSP. However SAT>IP servers
are assumed to speak only MP2T, so DESCRIBE is not used in the same
way. When no streams are active, DESCRIBE will return 404 according
to the spec (see section 3.5.7). When streams are active, DESCRIBE
will return a list of all current streams along with information
about their signal strengths.

Previously, attemping to use ffmpeg with a rtsp:// url that points
to a SAT>IP server would work with some devices, but fail due to 404
response on others. Further, if the SAT>IP server was already
streaming, ffmpeg would incorrectly consume the DESCRIBE SDP response
and join an existing tuner instead of requesting a new session with
the URL provided by the user. These issues have been noted by many
users across the internet[2][3][4].

This commit adds proper spec-compliant support for SAT>IP, including:

- support for the satip:// psuedo-protocol[5]
- avoiding the use of DESCRIBE
- parsing and consuming the com.ses.streamID response header
- using "Transport: RTP/AVP;unicast" because the optional "/UDP"
  suffix confuses some servers

This patch has been validated against multiple SAT>IP vendor devices:

- Telestar Digibit R2
  (https://telestar.de/en/produkt/digibit-r1-2/)
- Kathrein EXIP 418
  (https://www.kathrein-ds.com/en/produkte/sat-zf-verteiltechnik/sat-ip/227/exip-418)
- Kathrein EXIP 4124
  (https://www.kathrein-ds.com/en/products/sat-if-signal-distribution/sat-ip/226/exip-4124)
- Megasat MEG-8000
  (https://www.megasat.tv/produkt/sat-ip-server-3/)
- Megasat Twin
  (https://www.megasat.tv/en/produkt/sat-ip-server-twin/)
- Triax TSS 400
  (https://www.conrad.com/p/triax-tss-400-mkii-sat-ip-server-595256)

[1] https://www.satip.info/sites/satip/files/resource/satip_specification_version_1_2_2.pdf
[2] https://stackoverflow.com/questions/61194344/does-ffmpeg-violate-the-satip-specification-describe-syntax
[3] https://github.com/kodi-pvr/pvr.iptvsimple/issues/196
[4] https://forum.kodi.tv/showthread.php?tid=359072&pid=2995884#pid2995884
[5] https://www.satip.info/resources/channel-lists/

avcodec/tiff: Disallow striped and tiled tiffs except for DNG

strips + tiles is not allowed in TIFF
DNG uses a separate codepath

Regression since da5b3d002862d1e105002a6dc1567e6551860896.

Fixes: NULL pointer dereference
Fixes: poc1
Fixes: Ticket8960
Found-by: 1vanChen of NSFOCUS Security Team
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

avformat/mpegts: Fix argument type for av_log

Reviewed-by: Marton Balint <cus@passwd.hu>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

avformat/mxfdec: Do not clear arrays in mxf_read_index_entry_array()

They are written too immediately, so it should not be needed.

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

lavu: use address-of operator checking clock_gettime

When targeting a recent enough macOS/iOS version that has clock_gettime
it won't be a weak symbol, in which case clang warns for this check
as it's always true:

  warning: address of function 'clock_gettime' will always
  evaluate to 'true'

This warning is silenced by using the address-of operator to make
the intent explicit.

avformat/hlsproto: update comment of file description

reference link to rfc8216

Signed-off-by: liuqi05 <liuqi05@kuaishou.com>

avformat/hls: update comment of file description

reference link to rfc8216

Signed-off-by: liuqi05 <liuqi05@kuaishou.com>

doc/muxers: correct description of hls_enc_key and hls_enc_iv

because hls_enc_key and hls_enc_iv get 16byte char
for example:
-hls_enc_key 0123456789abcdef -hls_enc_iv abcdefghijklmnop

Reviewed-by: Gyan Doshi <ffmpeg@gyani.pro>
Signed-off-by: Steven Liu <liuqi05@kuaishou.com>

avfilter/vf_decimate: always properly handle EOF

avfilter/vf_decimate: stop dropping last frames with ppsrc

avfilter/vf_decimate: fix broken logic with ppsrc handling

Make output bit exact with VIVTC VDecimate filter.

avfilter/vf_framerate: fix infinite loop with 1-frame input

Fixes infinite loop in:
ffmpeg -f lavfi -i testsrc=d=0.04 -vf framerate=50 -f null none

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

doc/protocols: document cache option

Add entry for read_ahead_limit

avfilter/vf_xmedian: add support for commands

avfilter/af_vibrato: make sure table size is at least 1

Fixes overreads.

avfilter/af_tremolo: make sure table size is at least 1

avcodec/cfhd: properly handle midstream size changes for transform type 2

libswscale: avoid UB nullptr-with-offset.

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

avformat/swfdec: Allocate output buffer after reading input

Fixes: Timeout (>10sec -> 0.26sec)
Fixes: 27419/clusterfuzz-testcase-minimized-ffmpeg_dem_SWF_fuzzer-5678307361947648
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: clip sample rate

Fixes: 1.21126e+111 is outside the range of representable values of type 'int'
Fixes: 27398/clusterfuzz-testcase-minimized-ffmpeg_dem_CAF_fuzzer-5412960339755008
Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

tools/target_dec_fuzzer: Adjust maxpixels for G2M

Fixes: Timeout (50sec -> 3sec)
Fixes: 27383/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_G2M_fuzzer-5196953666977792
Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

lavf/img2dec: Increase score for very large jpeg images.

Avoids a conflict with the raw mjpeg demuxer.

lavd/decklink_dec: Do not set codec_tags.

Only set the pix_fmt for rawvideo.

Fixes ticket #9005.

Reviewed-by: Marton Balint

ffmpeg: delay first stats

Wait for all output files to be initialized before printing first stats.

Avoids breaking output file dump report.

avformat/hlsenc: Add dependency on avc.o

a2b1dd0ce301450a47c972745a6b33c4c273aa5d added a dependency
on ff_nal_unit_extract_rbsp.

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

ffmpeg: don't delay printing initial stats

The first stats is printed after the initial stats_period has elapsed. With a large period,
it may appear that ffmpeg has frozen at startup.

The initial stats is now printed after the first transcode_step.

ffmpeg: add option stats_period

At present, progress stats are updated at a hardcoded interval of
half a second. For long processes, this can lead to bloated
logs and progress reports.

Users can now set a custom period using option -stats_period
Default is kept at 0.5 seconds.

avfilter/vf_decimate: fix overreads when using ppsrc

Derive input parameters from correct inlink when using ppsrc.

Previously both input frames would use dimensions of first inlink,
causing crash if first inlink w/h was smaller than second one.

dnn: add NV12 pixel format support

Signed-off-by: Ting Fu <ting.fu@intel.com>
Signed-off-by: Guo, Yejun <yejun.guo@intel.com>

avcodec/ffv1dec: Fix off by 1 error with quant tables

Fixes: assertion failure
Fixes: 28447/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_FFV1_fuzzer-5369575948550144
Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Reviewed-by: Paul B Mahol <onemda@gmail.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

avformat/mpegts: simplify nb_packets code

Reviewed-by: Marton Balint <cus@passwd.hu>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

avformat/mpegts: Increase pcr_incr width to 64bit

Fixes: division by zero
Fixes: 26459/clusterfuzz-testcase-minimized-ffmpeg_dem_MPEGTSRAW_fuzzer-5666350112178176
Fixes: 28154/clusterfuzz-testcase-minimized-ffmpeg_dem_MPEGTSRAW_fuzzer-5195728439476224
Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Reviewed-by: Marton Balint <cus@passwd.hu>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

avfilter/vf_decimate: switch to activate

avfilter/af_crystalizer: implement inverse filtering

avfilter/af_crystalizer: use proper type for variable

avfilter/af_crystalizer: improve filter description

avformat/libamqp: parse vhost in uri

Reviewed-by: Marton Balint <cus@passwd.hu>
Signed-off-by: Andriy Gelman <andriy.gelman@gmail.com>

avfilter: add shufflepixels video filter

doc/codecs: add entries for export_side_data

Document addition of venc_params and film_grain.

avcodec/fastaudio: Fix invalid shift exponent

Fixes: shift exponent 32 is too large for 32-bit type 'unsigned int'
Fixes: 25434/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_FASTAUDIO_fuzzer-6252363168612352
Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

avfilter/vf_maskedthreshold: add support for commands

avfilter/vf_maskedminmax: add support for commands

avfilter/vf_maskedclamp: add support for commands

avfilter/vf_bilateral: add support for commands

doc/filters: document that cas filter supports commands

tools/target_dem_fuzzer: remove unused label

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

avcodec/utils: Check bitrate for overflow in get_bit_rate()

Fixes: signed integer overflow: 617890810133996544 * 16 cannot be represented in type 'long'
Fixes: 26565/clusterfuzz-testcase-minimized-ffmpeg_dem_MV_fuzzer-5092054700654592
Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

avformat/mov: Check if hoov is at the end

Fixes: Timeout, probably infinite loop
Fixes: 26559/clusterfuzz-testcase-minimized-ffmpeg_dem_MOV_fuzzer-5391165484171264
Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

avcodec/hevc_ps: Change scaling_list_dc_coef to 8 bit

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

avcodec/hevc_ps: check scaling_list_dc_coef

Fixes: signed integer overflow: 2147483640 + 8 cannot be represented in type 'int'
Fixes: 28449/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_HEVC_fuzzer-5686013259284480
Reviewed-by: James Almer <jamrial@gmail.com>
Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

avfilter/vf_fillborders: add fade mode

fate/image: add missing ffprobe dependency to fate-dpx-probe

And use the existing probeframes helper while at it.

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

fate/image: update fate-dpx-probe reference file

Regression since 20b09b20a942d4aad38f9fa1324b713168d3db9a

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

avformat/rtsp: set AV_OPT_FLAG_DEPRECATED on deprecated options

Reviewed-by: Zhao Zhili <zhilizhao@tencent.com>
Signed-off-by: Andriy Gelman <andriy.gelman@gmail.com>

avfilter/vf_fillborders: add more fill modes

avcodec/dpx: fix use of uninitialised value

Found by Valgrind.

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

avcodec/hevc_sei: return the correct error code on User Data Registered parsing failure

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

avcodec/hevc_sei: print a log message when a unsupported ITU-T T35 SEI messages is parsed

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

avcodec/hevc_sei: refactor parsing User Data Registered ITU-T T35 SEI messages

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

avcodec/h264_sei: print a log message when a unsupported ITU-T T35 SEI messages is parsed

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

avcodec/h264_sei: refactor parsing User Data Registered ITU-T T35 SEI messages

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

avfilter/af_asoftclip: add two more useful options for finer filtering

avfilter/af_biquads: fix error in ro calculation

avcodec/dpx: Fix B&W film scans from Lasergraphics Inc

Signed-off-by: Harry Mallon <harry.mallon@codex.online>

avcodec/dpx: Read color information from DPX header

Signed-off-by: Harry Mallon <harry.mallon@codex.online>

avcodec/dpx: Report color_range from DPX header

Signed-off-by: Harry Mallon <harry.mallon@codex.online>

avcodec/dpx: Read SMPTE timecode from DPX

Signed-off-by: Harry Mallon <harry.mallon@codex.online>

avcodec/dpx: Read alternative frame rate from television header

Signed-off-by: Harry Mallon <harry.mallon@codex.online>

fate: Add dpx-probe test

Signed-off-by: Harry Mallon <harry.mallon@codex.online>

avformat/url: Change () position in ff_make_absolute_url()

No testcase
Reviewed-by: Nicolas George <george@nsup.org>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

avfilter/af_biquads: add one-pole shelf filters by adding pole option

avcodec/av1dec: support exporting Film Grain params as frame side data

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

avfilter/vf_limiter: add support for commands

avcodec/cuviddec: unref output frame on failure

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

avcodec/cuviddec: check for av_buffer_ref() failure

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

avcodec/libdav1d: stop setting AVFrame->best_effort_timestamp

It's now set by the generic decode code.

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

avcodec/mjpegdec: stop setting AVFrame->best_effort_timestamp

It's now set by the generic decode code.

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

avcodec/decode: set best_effort_timestamp on output frames for all decoders

Fixes a decoding regression introduced by e9a2a87773, and as a side effect also
fixes bogus values set to certain audio frames that had some samples discarded,
where the offsets added to pts, pkt_dts and pkt_duration were not reflected in
best_effort_timestamp.

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

avfilter/af_afreqshift: add fltp sample format support

fate: fix fate-filter-hqx on big-endian arches

One of the inputs to the fate test has an rgba pixel format which needs
to be converted to rgb32 (argb on big-endian) for the hqx filter. Because auto
scaling in the fate test is disabled, this needs a separate scale
filter.

Reviewed-by: Michael Niedermayer <michael@niedermayer.cc>
Signed-off-by: Andriy Gelman <andriy.gelman@gmail.com>