fixes use of uninitialized values if the filter is still unstable after using
the previous frame lpc coefficients.
(cherry picked from commit 9a3f10695a)
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Also define a codec capability for codecs that can handle
parameters changed externally between decoded packets.
Signed-off-by: Martin Storsjö <martin@martin.st>
(cherry picked from commit fe75dc8583)
Conflicts:
libavcodec/avcodec.h
libavcodec/version.h
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Signed-off-by: Martin Storsjö <martin@martin.st>
(cherry picked from commit 195c123cc8)
Conflicts:
libavcodec/version.h
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Fixes Ticket809
Bug found by: Oana Stratulat
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
(cherry picked from commit a40f6a5c69)
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
overreads in mpeg12 are not possible
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
(cherry picked from commit 549749c68e)
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
in ffmpegs h264 decoder.
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
(cherry picked from commit 38331d2036)
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
When turned on, H264/CAVLC gets ~15% (CVPCMNL1_SVA_C.264) slower for
ultra-high-bitrate files, or ~2.5% (CVFI1_SVA_C.264) for lower-bitrate
files. Other codecs are affected to a lesser extent because they are
less optimized; e.g., VC-1 slows down by less than 1% (all on x86).
The patch generated 3 extra instructions (cmp, cmovae and mov) per
call to get_bits().
The performance penalty on ARM is within the error margin for most
files, up to 4% in extreme cases such as CVPCMNL1_SVA_C.264.
Based on work (for GCI) by Aneesh Dogra <lionaneesh@gmail.com>, and
inspired by patch in Chromium by Chris Evans <cevans@chromium.org>.
(cherry picked from commit 8cfbbd928c)
Conflicts:
configure
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
The A32 bitstream reader variant is only used on ARMv5 and for
Prores due to the larger bit cache this decoder requires.
In benchmarks on ARMv5 (Marvell Sheeva) with gcc 4.6, the only
statistically significant difference between ALT and A32 is
a 4% advantage for ALT in FLAC decoding. There is thus no (longer)
any reason to keep the A32 reader from this point of view.
This patch adds an option to the ALT reader increasing the bit
cache to 32 bits as required by the Prores decoder. Benchmarking
shows no significant change in speed on Intel i7. Again, the
A32 reader fails to justify its existence.
Signed-off-by: Mans Rullgard <mans@mansr.com>
(cherry picked from commit a1e98f198e)
Conflicts:
libavcodec/get_bits.h
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
This fixes a deadlock VLC triggered with multithreaded decoding. The
wait forces one of the current waiters to wake and not the thread
which calls pthread_cond_signal() itself.
Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
(cherry picked from commit 2003a29284)
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Fixes Ticket804
Bug found by: Oana Stratulat
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
(cherry picked from commit 134aaa79f7)
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Fixes Ticket806
Bug found by Oana Stratulat
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
(cherry picked from commit fc739b79ad)
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Interlaced content for most codec requires it.
This patch is a stop-gap pending a serious rework to support
codecs with non 16 pixel macroblocks.
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
(cherry picked from commit 37c0dc626d)
Conflicts:
libavcodec/utils.c
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Fixes Ticket803
Bug found by: Oana Stratulat
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
(cherry picked from commit f27930cd9a)
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Fixes Ticket801
Bug found by: Oana Stratulat
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
(cherry picked from commit a8469223f6)
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Fixes a null pointer dereference.
Fixes 2nd half of Ticket800
Bug found by: Oana Stratulat
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
(cherry picked from commit 65f0f9183b)
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Fixes Ticket792
Bug found by Oana Stratulat
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
(cherry picked from commit 03a9c9932d)
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Fixes second half of Ticket794
Bug found by: Oana Stratulat
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
(cherry picked from commit f85334f58e)
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Fixes half of Ticket794
Bug found by: Oana Stratulat
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
(cherry picked from commit 608708009f)
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
The previous code ended in multiple different infinite
loops. See stl_ten_1_big.sfd as example with and without zzuf
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
(cherry picked from commit 7859740c6a)
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
No difference in PSNR or bitrate in the printed precission with the matrix lobby scene at 322x242
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
(cherry picked from commit 56bf2c2a19)
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Fixes Ticket784
Found by: ami_stuff
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
(cherry picked from commit e09ffa4524)
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
channels would be 0 sometimes and would cause floating point exception
Fixes bugzilla #124
Signed-off-by: Justin Ruggles <justin.ruggles@gmail.com>
(cherry picked from commit e614fac2e6)
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Fixes bug170
Bug found by: Gautam Gupta
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
(cherry picked from commit f37b2d5a68)
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Fixes Ticket780
Bug Found by: cosminamironesei
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
(cherry picked from commit 9af6abdc17)
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
In the case that (frame_flags & 0x03) == 3, hybrid_maxclip
may have had a signed integer overflow.
Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
Signed-off-by: Anton Khirnov <anton@khirnov.net>
(cherry picked from commit bb9747c8ee)
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
It doesn't make much sense to clip pre-shift,
nor is it correct for proper decoding.
Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
Signed-off-by: Anton Khirnov <anton@khirnov.net>
(cherry picked from commit 365e1ec26d)
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Bug found by: Oana Stratulat
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
(cherry picked from commit 6d8e6fe9db)
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
width and height might get passed as 0 and would cause floating point
exceptions in decode_frame.
Fixes bugzilla #149
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
(cherry picked from commit 4af0262f7d)
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
dont leave its value at something random.
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
(cherry picked from commit 902c090413)
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
