| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
| |
Adding the Nvidia NVEnc H.264 and H.265 encoders.
Based on Initial work by sgothel
--enable-nvenc is the new compile time configure option to enable for builds.
|
|
|
|
| |
#1401.
|
|
|
|
|
| |
This leaves the recovery point heuristic intact and fixes the problem
that was triggering the heuristic to return an early frame
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Changelog: https://github.com/FFmpeg/FFmpeg/blob/e049f7c24fc6aa5fc925f860e2ad940a75cfd84f/Changelog
- avcodec/aacdec_fixed: Fix undefined integer overflow in apply_independent_coupling_fixed()
- avcodec/dirac_dwt_template: Fix undefined behavior in interleave()
- avutil/common: Fix undefined behavior in av_clip_uintp2_c()
- fftools/ffmpeg: Fallback to duration if sample rate is unavailable
- avformat/mov: Only set pkt->duration to non negative values
- avcodec/mpeg4videodec: Clear bits_per_raw_sample if it has originated from a previous instance
- avformat/movenc: fix recognization of cover image streams
- avformat/movenc: properly handle cover image codecs
- avcodec/h264_slice: Fix overflow in recovery_frame computation
- avcodec/h264_ps: Move MAX_LOG2_MAX_FRAME_NUM to header so it can be used in h264_sei
- avcodec/h264_mc_template: Only prefetch motion if the list is used.
- avcodec/xwddec: Use ff_set_dimensions()
- avcodec/wavpack: Fix overflow in adding tail
- avcodec/shorten: Fix multiple integer overflows
- avcodec/shorten: Fix undefined shift in fix_bitshift()
- avcodec/shorten: Fix a negative left shift in shorten_decode_frame()
- avcodec/shorten: Sanity check nmeans
- avcodec/shorten: Check non COMM chunk len before skip in decode_aiff_header()
- avcodec/mjpegdec: Fix integer overflow in ljpeg_decode_rgb_scan()
- avcodec/truemotion2: Fix overflow in tm2_apply_deltas()
- avcodec/opus_silk: Change silk_lsf2lpc() slightly toward silk/NLSF2A.c
- avcodec/amrwbdec: Fix division by 0 in find_hb_gain()
- avcodec/h263dec: Reinitialize idct context if it has not been setup for the active profile
- avcodec/idctdsp: Clear idct/idct_add for studio profile
- avformat/mov: replace a value error by clipping into valid range in mov_read_stsc()
- avformat/bintext: Reduce detection for random .bin files as it more likely is not a multimedia related file
- avformat/mov: Break out early if chunk_count is 0 in mov_build_index()
- avcodec/fic: Avoid some magic numbers related to cursors
- avcodec/mpeg4video: Detect reference studio streams as studio streams
- avcodec/mpeg4videodec: Do not corrupt bits_per_raw_sample
- avcodec/mpeg4videode: Eliminate out of loop VOP startcode reading for studio profile
- avcodec/g2meet: ask for sample with overflowing RGB
- avcodec/idctdsp: Transmit studio_profile to init instead of using AVCodecContext profile
- avcodec/ac3dec: Check that the number of channels with dependant streams is valid
- avcodec/ac3dec: Fix null pointer dereference in ac3_decode_frame()
- avcodec/aacdec_fixed: use 64bit to avoid overflow in rounding in apply_dependent_coupling_fixed()
- oavcodec/aacpsdsp_template: Use unsigned for hs0X to prevent undefined behavior
- avcodec/g723_1dec: Clip bits2 in both directions
- avcodec/mpeg4videoenc: Use 64 bit for times in mpeg4_encode_gop_header()
- avcodec/mlpdec: Only change noise_type if the related fields are valid
- indeo4: Decode all or nothing of a band header.
- avcodec/ac3dec: Use frame_size if superframe_size is 0
- avformat/mov: Only fail for STCO/STSC contradictions if both exist
- avcodec/dirac_dwt: Fix integer overflow in COMPOSE_DD97iH0 / COMPOSE_DD137iL0
- avcodec/fic: Check available input space for cursor
- avcodec/mpeg4videodec: Check bps (VOL header) before VOP for studio profile
- avcodec/g2meet: Check RGB upper limit
- avcodec/jpeg2000dec: Fix undefined shift in the jpeg2000_decode_packets_po_iteration() CPRL case
- avcodec/jpeg2000dec: Skip init for component in CPRL if nothing is to be done
- avcodec/g2meet: Change order of operations to avoid undefined behavior
- avcodec/flac_parser: Fix infinite loop
- avcodec/mpeg4videodec: Split decode_studio_vol_header() out of decode_studiovisualobject()
- avcodec/mpeg4videodec: Move decode_studiovisualobject() parsing in the branch for visual object parsing
- avcodec/mpeg4video_parser: Avoid litteral 0x1B6, use named constant instead
- avcodec/mpeg4video_parser: Fix incorrect spliting of MPEG-4 studio frames
- avformat/m4vdec: Use the same constant names as libavcodec
- avformat/m4vdec: Fix detection of raw MPEG-4 ES Studio
- avcodec/wavpack: Fix integer overflow in DEC_MED() / INC_MED()
- avcodec/wavpack: Fix integer overflow in wv_unpack_stereo()
- avcodec/error_resilience: Fix integer overflow in filter181()
- avcodec/h263dec: Check slice_ret in mspeg4 slice loop
- avcodec/elsdec: Fix memleaks
- avcodec/vc1_block: simplify ac_val computation
- avcodec/ffv1enc: Check that the crc + version combination is supported
- configure: The eac3_core bitstream filter needs the ac3 parser.
- configure: fix arm inline asm checks
- lavf/libssh: translate a read of 0 to EOF
- ffprobe: fix SEGV when new streams are added
- avformat/mpegts: fix incorrect indentation
- avformat/mpegts: initialize section_buf to fix valgrind test failure
- avformat/mpegts: reindent after last change
- avformat/mpegts: parse sections with multiple tables
- avformat/mpegts: clean up whitespace
- avformat/mpegts: use MAX_SECTION_SIZE instead of hardcoded value
- avformat/mpegts: skip non-PMT tids earlier
- avcodec/mediacodecdec: add workaround for buggy amlogic mpeg2 decoder
- avcodec/mediacodecdec: wait on first frame after input buffers are full
- avcodec/mediacodecdec: restructure mediacodec_receive_frame
- avcodec/mediacodec_wrapper: add helper to fetch SDK_INT
- avcodec/mediacodecdec: refactor pts handling
- avcodec/mediacodecdec: use AV_TIME_BASE_Q
- avcodec/mediacodecdec: clarify delay_flush specific code
- avcodec/videotoolbox: fix decoding of some HEVC videos
- avcodec/hevc: remove videotoolbox hack
- avcodec/videotoolbox: split h264/hevc callbacks
- avcodec/videotoolbox: cleanups
- avcodec/videotoolbox: fix kVTCouldNotFindVideoDecoderErr trying to decode HEVC on iOS
- avcodec/videotoolbox: improve logging of decoder errors
- avcodec/xwddec: fix palette alpha
- avformat/webm_chunk: always use a static buffer for get_chunk_filename
- configure: fix configure check for lilv-0
- avcodec/nvdec_hevc: fix scaling lists
- avcodec/hevcdec: make ff_hevc_frame_nb_refs take a const pointer
- lavf/bluray: translate a read of 0 to EOF
- lavf/dashenc: don't call flush_init_segment before avformat_write_header
- avdevice/decklink_dec: unref packets on avpacket_queue_put error
- avcodec/hnm4video: fix palette alpha
- avcodec/anm: fix palette alpha
- avformat/qtpalette: parse color table according to the QuickTime file format specs
- ffplay: Fix realloc_texture when input texture is NULL.
- hwcontext_vaapi: Fix compilation with libva versions < 1.4.0
- lavf/qsv: clone the frame which may be managed by framework
- lavf: make overlay_qsv work based on framesync
- avformat/segafilm - revert keyframe detection
- avformat/utils: refactor upstream_stream_timings
- avformat/utils: ignore outlier durations on subtitle/data streams as well
|
|
|
|
|
|
|
|
|
| |
The ffmpeg developers added a "heuristic" to the h.264 recovery point
code I wrote to prevent display of corrupt frames. The heuristic
doesn't work very reliably, so I've disabled it.
When I get a chance I'll see if I can come up with a better way to
handle what the heuristic was added for.
|
|
|
|
| |
This reverts commit a589396d51a11a8697e2909e75e097db5bed22a1.
|
| |
|
| |
|
|
|
|
| |
Still required, only HandBrake will not build it anymore.
|
|
|
|
| |
Closes #549.
|
| |
|
| |
|
|
|
| |
configure.ac is modified by the patch from #1380
|
|
|
|
| |
Seems like libvpx should have a generic target for future/unknown macOS versions.
|
|
|
|
| |
The avx-512 patch should avoid breakage with wonky compilers and older systems such as macOS <= 10.11. The pthread detection patch can be applied more generally but really is a safety check targeting win32.
|
| |
|
|
|
|
| |
Fixes -mavx which fixes -march=native as used by Homebrew on macOS. Fixes #1351.
|
| |
|
| |
|
|
|
|
| |
It is the default.
|
|
|
|
|
|
|
|
| |
The generated vpx.pc file results in undefined symbols during ffmpeg's
configure tests. libpthread is not getting added to the link line as
required. The vpx.pc file needs to include -lpthread in the Libs instead
of Libs.private when generating only a static library.
Fixes https://github.com/HandBrake/HandBrake/issues/1383
|
|
|
|
| |
Practically, there is no difference. But this is how QSV was done and is easy to read.
|
|
|
|
| |
#1144.
|
|
|
|
| |
Issue should be fixed by 7cddc6fe8779e1e098b48ecdfddf6b2c15050578.
|
|
|
|
|
|
| |
The generated opus.pc file results in undefined symbols during ffmpeg's
configure tests. libm is not getting added to the link line as
required. The opus.pc file needs to include -lm in the Libs instead of
Libs.private when generating *only* a static library.
|
|
|
|
|
| |
There's no telling which of these may be needed internally for ffmpeg
encoders, decoders, and filters.
|
|
|
|
|
|
|
|
|
|
|
|
| |
Fixes FFmpeg configure error on macOS:
```
./configure: line 588: eval: --: invalid option
eval: usage: eval [arg ...]
make[1]: *** [contrib/ffmpeg/.stamp.ffmpeg.configure] Error 2
```
...caused by attempting to `eval --arch=x86_64=yes`.
|
|
|
|
| |
Fixes #1372.
|
|
|
|
| |
Addresses another suggestion here https://github.com/HandBrake/HandBrake/issues/1370#issuecomment-392958720
|
|
|
|
|
|
| |
Addresses suggestions here https://github.com/HandBrake/HandBrake/issues/1370#issuecomment-392958720
-fomit-frame-pointer may still be required on 32-bit Intel; however, we no longer support 32-bit builds.
|
|
|
|
|
|
| |
This reverts commit 83dfde11d85a3ee1a7583049c75f1044bd62c94b.
Apparently, I was correct the first time around and these need to be enabled, just not explicitly.
|
|
|
|
|
| |
lzma may be used by the ffmpeg tiff decoder which can be used in
matroska files.
|
|
|
|
|
| |
It's not needed and can cause build errors depending on which mingw tool
chain is installed.
|
|
|
|
| |
Should have been more thorough in 74dbfebb28d9e65d2f43161bb2b3db8f94005a39.
|
| |
|
| |
|
|
|
|
| |
Closes #974 and closes #1078.
|
|
|
|
| |
Per https://github.com/HandBrake/HandBrake/pull/1078#issuecomment-384461808.
|
| |
|
|
|
|
| |
Use https and swap url order.
|
|
|
|
| |
Fixes the build. Hardware acceleration should really be tested before we enable it, even in the nightly.
|
| |
|
|
|
|
| |
See http://ffmpeg.org/pipermail/ffmpeg-devel/2017-November/219642.html
|
|
|
|
|
|
|
|
| |
Patch 2/2 for https://github.com/HandBrake/HandBrake/issues/974
moving to FFMPEG 3.4.1 from LIBAV 12.2.
All patches have been moved to subfolder 'old' since they do not apply cleanly anymore.
Work has to be performed to validate whether patches are still required.
|
|
|
|
|
|
|
|
| |
Patch 1/2 for https://github.com/HandBrake/HandBrake/issues/974
moving to FFMPEG 3.4.1 from LIBAV 12.2.
The re-ordering of to be linked modules was required to solve
statically linked dependencies. See libhb/module.defs etc.
|
|
|
|
|
| |
FreeBSD's diff doesn't support --version option.
Use hash command to check if diff is available or not.
|
| |
|
| |
|
|
|
|
| |
Build contains debug symbols without this.
|
|
|
|
| |
This fixes corrupted output issue on macos
|