This is necessary since the switch to floating point
input means there is no longer sufficient input data
validation.
There is a good chance that other encoders are affected
by similar issues.
This problem can be triggered by trying to encode
extremely large values and probably also with Inf and
possibly also NaNs.
Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>
Odd chunk size makes no sense for stereo and incomplete chunks are
not supported.
Reported-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org
Simplify a little, assume empty frames are acceptable and
do not pointlessly reinit the bytestream2 contexts using
possibly wrong size values.
Reported-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org
* qatar/master:
doc: Mention the target_samples and ld variables for fate configs
fate.sh: Allow specifying --as via a specific variable
Conflicts:
doc/fate.texi
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit '2d2d6a4883479403798f4ed46941d5b365823570':
lavf: add a raw WavPack muxer.
apetag: add support for writing APE tags
matroskaenc: support muxing WavPack
Conflicts:
libavformat/Makefile
libavformat/allformats.c
libavformat/apetag.h
libavformat/version.h
libavformat/wvenc.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit '794ca87d2bff2513118de8b97595b3e23070e67d':
wvdec: split block header parsing into a separate file
Conflicts:
libavformat/Makefile
libavformat/wvdec.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit '89806691b1c39181c63d95e0fddc30f11e2a7b04':
wavpack: check that all the channels were coded.
wavpack: check that there aren't too many blocks per packet
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit '7d039e70a5ff23a7deaa866684d2e8872acc5169':
wavpack: extract channel information from the bitstream
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit '78f75b6fa421dd39a715588e9487579f1ce5bada':
wavpack: extract sample rate from the bitstream
configure: support gcc-4.8 instrumentation
Conflicts:
libavcodec/wavpack.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit '4a27a52a1f74016095b7aee1b4a422cf62217ade':
fate: Don't use files from SRC_PATH in the actual tests
indeo4: reuse context block VLC for band instead of defaulting
Conflicts:
tests/Makefile
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit '728214992e3698305550c1762f973d2ac567f016':
vc1dec: Remove interlaced warning
vc1dec: Fix mixed field/frame intensity compensation
vc1dec: Match addressing between compensation and MC in vc1_mc_4mv_chroma4
vc1dec: Handle top and bottom blocks in vc1_mc_4mv_chroma4() differently if needed
vc1dec: Fix doxy for vc1_mc_4mv_chroma4()
vc1dec: Drop old use_ic code from vc1_b_mc
vc1: Use shuffled use_ic instead of equally shuffled mv_mode
vc1dec: Implement intensity compensation for vc1_interp_mc()
Conflicts:
Changelog
libavcodec/vc1dec.c
libavcodec/version.h
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit '93b1281264b87961f53c3e9c134cc2727ecd91ed':
vc1dec: Shuffle field MVs after decoding, not before
Merged-by: Michael Niedermayer <michaelni@gmx.at>
the float conditions are largely changed to check the DWT
instead of the bitexact flag, which previously was wrong
(DWT53 is always int)
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
The modern MSVC for ARM always builds for thumb, and it can't be
disabled.
Also just use the default arch instead of trying to map the -march
parameter to MSVC's -arch parameter (which only takes the values
ARMv7VE and VFPv4).
Signed-off-by: Martin Storsjö <martin@martin.st>
The ist->pts value at this point does not necessarily match the actual packet
pts, which is what should be used to decide whether to discard the packet.
Without this change, some video packets that have pts on or after the start
time are getting discarded when they should not be.
Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
The decoder did not start predicted frames with missing references but
called the end of frame handling on them anyway. This caused an
assertion in the VA API HW accelorator since it tried to render a
picture with uninitialized buffers.
The quantization code needs more work, not so much work
merging but more work investigating what is correct.
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
