Welcome to mirror list, hosted at ThFree Co, Russian Federation.

github.com/FFmpeg/FFmpeg.git - Unnamed repository; edit this file 'description' to name the repository.
summaryrefslogtreecommitdiff
path: root/libavcodec/vp9.c
AgeCommit message (Collapse)Author
2016-10-24vp9: change order of operations in adapt_prob().Ronald S. Bultje
This is intended to workaround bug "665 Integer Divide Instruction May Cause Unpredictable Behavior" on some early AMD CPUs, which causes a div-by-zero in this codepath, such as reported in Mozilla bug #1293996. Note that this isn't guaranteed to fix the bug, since a compiler is free to reorder instructions that don't depend on each other. However, it appears to fix the bug in Firefox, and a similar patch was applied to libvpx also (see Chrome bug #599899).
2016-10-07Merge commit '32c8359093d1ff4f45ed19518b449b3ac3769d27'Hendrik Leppkes
* commit '32c8359093d1ff4f45ed19518b449b3ac3769d27': lavc: export the timestamps when decoding in AVFrame.pts Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>
2016-06-05vp9: Return the correct size when decoding a superframeDenis Charmet
According to avcodec.h, avcodec_decode_video2 should return the number of bytes used if a frame was decoded. The current implementation returns size - used size of all the subframes. This fixes the VLC's bug https://trac.videolan.org/vlc/ticket/16836. The superframe is always fully consumed. Signed-off-by: Diego Biurrun <diego@biurrun.de>
2016-04-17vp9: ignore frame_context_index field in key-/intraonly-frames.Ronald S. Bultje
Reproduces a bug to remain consistent with libvpx' behaviour.
2016-03-12vp9: fix a few signed integer left-shifts.Ronald S. Bultje
Fixes trac tickets 5127, 5129, 5130.
2016-02-13vp9: only call ff_get_format on stream format changes.Ronald S. Bultje
In practice, this means we don't call it N times for N-threaded decoding. Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2016-01-01lavc: add vp9 profiles to AVCodecDescriptorHendrik Leppkes
2015-12-22vaapi: Add VP9 hwaccell supportTimo Rothenpieler
Signed-off-by: Timo Rothenpieler <timo@rothenpieler.org>
2015-12-07avcodec: implement vp9 dxva2 hwaccelHendrik Leppkes
2015-12-07vp9: add hwaccel hooksHendrik Leppkes
2015-12-07vp9: fix pixel format changes with threadingHendrik Leppkes
2015-12-02vp9: always keep s->bytesperpixel and ctx->pix_fmt in sync.Ronald S. Bultje
Fixes mozilla bug 1229128.
2015-10-24vp9: uses ff_set_dimensions (which sets coded_width/height).Ronald S. Bultje
Fixes ticket 4935.
2015-10-16vp9: use AVFrame.buf[0] to check if a frame is validHendrik Leppkes
AVFrame.data[0] is not guaranteed to be set with a HWAccel
2015-10-15vp9: expose reference frames in VP9SharedContextHendrik Leppkes
2015-10-15vp9: split header into separate struct and expose in vp9.hRonald S. Bultje
This allows hwaccels to access the bitstream header information.
2015-10-04avcodec: use HAVE_THREADS header guards to silence -Wunused-functionGanesh Ajjanagadde
When compiled with --disable-pthreads, e.g http://fate.ffmpeg.org/report.cgi?time=20150917015044&slot=alpha-debian-qemu-gcc-4.7, a bunch of -Wunused-functions are reported due to missing header guards around threading related functions. This patch should silence such warnings. Signed-off-by: Ganesh Ajjanagadde <gajjanagadde@gmail.com>
2015-09-15vp9: switch min_tile_cols location so it shifts up instead of down.Ronald S. Bultje
This fixes cases where the shifted number is 64, but we shifted non- zero numbers away in the shift. The change makes behaviour consistent with libvpx.
2015-09-12vp9: Fix emu[] edge overflow conditions for >8bpp/non-420.Ronald S. Bultje
2015-09-12vp9: re-initialize internal buffers on bpp change also.Ronald S. Bultje
2015-09-10vp9: fix overflow in 8x8 topleft 32x32 idct ssse3 version.Ronald S. Bultje
Also disable the mmx/iwht optimization when the bitexact flag is set. With synthetically coded coefficients (i.e. these that lead to a residual well outside the [-255,255] range), our optimizations will overflow. It doesn't make sense to fix the overflows, since they can only occur on synthetic input, not on real fwht-generated input. Thus, add a bitexact flag that disables this optimization.
2015-09-10vp9: permanently delete segmentation refmap on any size change.Ronald S. Bultje
2015-09-05vp9: check return value of ff_thread_ref_frame().Ronald S. Bultje
Fixes CID 1322309.
2015-09-05vp9: fix edge copy for 10/12bpp frames.Ronald S. Bultje
2015-09-05vp9: do unscaled MC in scaled path if size of this reference matches.Ronald S. Bultje
This can happen if we do bidirectional MC, where one reference has the same size as the current frame, but the other one doesn't.
2015-09-05vp9: sync segmentation.absolute_vals between threads.Ronald S. Bultje
2015-09-05vp9: always sync segmentation.feat between threads.Ronald S. Bultje
2015-09-04vp9: don't erase values for {lf,ref,skip,q}_enabled if segmentation=0.Ronald S. Bultje
Instead, use segmentation.enabled before accessing each of these values.
2015-09-04vp9: fix segmentation map retention across keyframe boundaries.Ronald S. Bultje
2015-09-04vp9: fix RGB chroma subsampling.Ronald S. Bultje
2015-09-04vp9: read reserved bit in RGB header.Ronald S. Bultje
2015-09-04vp9: clip intermediates in dequant calculations.Ronald S. Bultje
This makes values consistent with libvpx.
2015-09-04vp9: fix segmentation map retention if segmentation is turned off.Ronald S. Bultje
2015-09-04vp9: use resetctx in the same way as libvpx.Ronald S. Bultje
2015-09-04vp9: fix indentation.Ronald S. Bultje
2015-09-04vp9: deal with the case where update_map=0 but we're key/intraonly.Ronald S. Bultje
2015-09-04vp9: reset segmentation information on context reset.Ronald S. Bultje
2015-07-27Merge commit 'def97856de6021965db86c25a732d78689bd6bb0'Michael Niedermayer
* commit 'def97856de6021965db86c25a732d78689bd6bb0': lavc: AV-prefix all codec capabilities Conflicts: cmdutils.c ffmpeg.c ffplay.c libavcodec/8svx.c libavcodec/aacenc.c libavcodec/ac3dec.c libavcodec/adpcm.c libavcodec/alac.c libavcodec/atrac3plusdec.c libavcodec/bink.c libavcodec/dnxhddec.c libavcodec/dvdec.c libavcodec/dvenc.c libavcodec/ffv1dec.c libavcodec/ffv1enc.c libavcodec/fic.c libavcodec/flacdec.c libavcodec/flacenc.c libavcodec/flvdec.c libavcodec/fraps.c libavcodec/frwu.c libavcodec/gifdec.c libavcodec/h261dec.c libavcodec/hevc.c libavcodec/iff.c libavcodec/imc.c libavcodec/libopenjpegdec.c libavcodec/libvo-aacenc.c libavcodec/libvorbisenc.c libavcodec/libvpxdec.c libavcodec/libvpxenc.c libavcodec/libx264.c libavcodec/mjpegbdec.c libavcodec/mjpegdec.c libavcodec/mpegaudiodec_float.c libavcodec/msmpeg4dec.c libavcodec/mxpegdec.c libavcodec/nvenc_h264.c libavcodec/nvenc_hevc.c libavcodec/pngdec.c libavcodec/qpeg.c libavcodec/ra288.c libavcodec/rv10.c libavcodec/s302m.c libavcodec/sp5xdec.c libavcodec/takdec.c libavcodec/tiff.c libavcodec/tta.c libavcodec/utils.c libavcodec/v210dec.c libavcodec/vp6.c libavcodec/vp9.c libavcodec/wavpack.c libavcodec/yop.c Merged-by: Michael Niedermayer <michael@niedermayer.cc>
2015-07-27lavc: AV-prefix all codec capabilitiesVittorio Giovara
Express bitfields more simply. Signed-off-by: Vittorio Giovara <vittorio.giovara@gmail.com>
2015-07-16lavc: Add properties field to AVCodecContext.Carl Eugen Hoyos
The new field can hold information about losslessness and closed captions for now.
2015-06-30vp9/update_prob: prevent out of bounds table readJames Zern
the max value of the lookup in expanded form is: (((1 << 7) - 1) << 1) - 65 + 1 + 64 = 254 add one entry of padding to inv_map_table[] to prevent out of bounds access with non-conforming / fuzzed bitstreams Signed-off-by: James Zern <jzern@google.com> Reviewed-by: "Ronald S. Bultje" <rsbultje@gmail.com> Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
2015-06-25vp9: fix profile check for intraonly frames.Ronald S. Bultje
Reviewed-by: James Zern <jzern@google.com> Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
2015-06-19vp9: don't retain NULL as segmentation_mapAndreas Cadhalpun
This fixes segmentation faults, which were introduced in commit 4ba8f327. Reviewed-by: Ronald S. Bultje <rsbultje@gmail.com> Signed-off-by: Andreas Cadhalpun <Andreas.Cadhalpun@googlemail.com>
2015-06-16doc: avoid incorrect phrase 'allows to'Andreas Cadhalpun
Also fix typo found by Lou Logan: Sacrifying -> Sacrificing Reviewed-by: Lou Logan <lou@lrcd.com> Signed-off-by: Andreas Cadhalpun <Andreas.Cadhalpun@googlemail.com>
2015-06-08vp9: change type of tile_size from unsigned to int64_tAndreas Cadhalpun
Otherwise the check 'tile_size < size' treats a negative size as unsigned, causing the check to pass. This subsequently leads to segmentation faults. This was originally fixed as part of Libav commit 72ca83, so the original author is one of the following developers: Anton Khirnov <anton@khirnov.net> Diego Biurrun <diego@biurrun.de> Luca Barbato <lu_zero@gentoo.org> Martin Storsjö <martin@martin.st> Reviewed-by: Ronald S. Bultje <rsbultje@gmail.com> Signed-off-by: Andreas Cadhalpun <Andreas.Cadhalpun@googlemail.com>
2015-06-01vp9: clamp final zero MV if find_ref_mvs() found no suitable candidates.Ronald S. Bultje
This may actually result in a non-zero MV. Fixes ticket 4583. Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
2015-05-31vp9: Parse subsampling and report missing featureVittorio Giovara
2015-05-18Merge remote-tracking branch 'rbultje/vp9-bugfixes'Michael Niedermayer
* rbultje/vp9-bugfixes: vp9: match another find_ref_mvs() bug in libvpx. vp9: fix scaled motion vector clipping for sub8x8 blocks. vp9: improve signbias check. vp9: don't allow compound references if error_resilience is enabled. vp9: clamp segmented lflvl before applying ref/mode deltas. vp9: reset loopfilter mode/ref deltas on keyframe. vp9: fix crash when playing back 440/440 content with width%64<56. vp9: extend loopfilter workaround for vp9 h/v mix-up to work for 422. vp9: clip motion vectors in the same way as libvpx does. vp9: set skip flag if the block had no coded coefficients. vp9: apply mv scaling workaround only when subsampling is enabled. vp9: read all 4x4 blocks in sub8x8 blocks individually with scalability. vp9: fix segmentation map referencing upon framesize change. vp9: disable more pmulhrsw optimizations in idct16/32. vp9: disable all pmulhrsw in 8/16 iadst x86 optimizations. Merged-by: Michael Niedermayer <michaelni@gmx.at>
2015-05-16vp9: match another find_ref_mvs() bug in libvpx.Ronald S. Bultje
If we find a second non-sub8x8 motion vector for a non-first sub8x8 block, and the clamped value is identical to the first non-sub8x8 motion vector, then the resulting nearmv motion vector is forced to zero.
2015-05-16vp9: fix scaled motion vector clipping for sub8x8 blocks.Ronald S. Bultje
To match the obscure clipping bug behaviour in libvpx.
generated by cgit v1.2.3 (git 2.25.1) at 2024-08-11 08:12:39 +0300