Found-by: ubitux
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
(cherry picked from commit cfce6f7efd28130bf0dd409b2367ca0f8c9b2417)
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
The function otherwise would initialize the context without setting context_initialized
alternatively we could set context_initialized
Fixes valgrind anomalies related to ticket 3928
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
(cherry picked from commit 0d0f7f0ba43f64312ae4a05d97afecf1b7b1330c)
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
(cherry picked from commit 2762323c37511fbbc98b164c07620b9ebc59ec68)
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
* qatar/master:
mpeg: Drop unused parameters from ff_draw_horiz_band()
Conflicts:
libavcodec/mpegvideo.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit 'ebfe622bb1ca57cecb932e42926745cba7161913':
mpegvideo: drop support for real (non-emulated) edges
Conflicts:
libavcodec/mpegvideo.c
libavcodec/mpegvideo_motion.c
libavcodec/wmv2.c
If this is slower on a major platform then it should be investigated
and potentially reverted.
See: 8fc52a5ef94712d900fc8fe7503cf9c9ba635143
See: 3969b4b861ce8152e3195e8f8c3437abd273b90a
Merged-by: Michael Niedermayer <michaelni@gmx.at>
Several decoders disable those anyway and they are not measurably faster
on x86. They might be somewhat faster on other platforms due to missing
emu edge SIMD, but the gain is not large enough (and those decoders
relevant enough) to justify the added complexity.
* commit 'aec25b1c4650944d32706bfd40eb02bbd5587303':
mpegvideo: split the encoding-only parts of frame_start() into a separate function
Conflicts:
libavcodec/mpegvideo.c
libavcodec/mpegvideo_enc.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit 'feded990e3ef9af4a0b827d5b6d8fe86f0b94942':
mpegvideo: set reference/pict_type on generated reference frames
Conflicts:
libavcodec/mpegvideo.c
The picture type is left at P type as it was explicitly set before
Merged-by: Michael Niedermayer <michaelni@gmx.at>
With interlaced vc1 it was possible that accesses could happen outside these
tables before this.
Regression since 017e234c204f8ffb5f85a073231247881be1ac6f
Reproduced with a sample from Mateusz "j00ru" Jurczyk and Gynvael Coldwind
Fixes (again) : 480i30__codec_WVC1__mode_2__framerate_29.970__type_2__preproc_17.SIGFPE.bfa.390.wmv
No releases are affected by this
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Otherwise the generic code will unref them, which can then result in
last_picture_ptr == current_picture_ptr, which causes deadlocks at least
in rv40.
Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC:libav-stable@libav.org
* commit 'b06c8bce02b15115a4789252365df2dda0c4713c':
mpegvideo: remove an unneeded call to avcodec_get_frame_defaults().
Merged-by: Michael Niedermayer <michaelni@gmx.at>
This is a temporary workaround to allow deprecating
avcodec_get_frame_defaults(). The proper solution will be using a
properly allocated AVFrame in Picture.
* commit '0b0a7a751de02464a33717e70352f696372ba1c4':
mpegvideo: move encode-only parts of common_end() to encode_end()
Conflicts:
libavcodec/mpegvideo.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit '381a722562bcc0b623acf6a00a583fe989bcb72a':
mpegvideo: split the encoding-only parts of ff_MPV_frame_end() into a separate function
Conflicts:
libavcodec/mpegvideo.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit 'b7254288d222013e20539c530b1ec5d324ed5352':
mpegvideo: do not update last_non_b_pict_type in update_thread_context()
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit 'bedf952bb80ae26427854dcde56c139ecb87d4a0':
mpegvideo: move setting encoding-only vars from common_defaults() to encode_defaults()
Merged-by: Michael Niedermayer <michaelni@gmx.at>
