In some cases, 2 or 3 calls are performed to functions for unusual
widths. Instead, perform 2 calls for different widths to split the
workload.
The 8+16 and 4+8 widths for respectively 8 and more than 8 bits can't
be processed that way without modifications: some calls use unaligned
buffers, and having branches to handle this was resulting in no
micro-benchmark benefit.
For block_w == 12 (around 1% of the pixels of the sequence):
Before:
12758 decicycles in epel_uni, 4093 runs, 3 skips
19389 decicycles in qpel_uni, 8187 runs, 5 skips
22699 decicycles in epel_bi, 32743 runs, 25 skips
34736 decicycles in qpel_bi, 32733 runs, 35 skips
After:
11929 decicycles in epel_uni, 4096 runs, 0 skips
18131 decicycles in qpel_uni, 8184 runs, 8 skips
20065 decicycles in epel_bi, 32750 runs, 18 skips
31458 decicycles in qpel_bi, 32753 runs, 15 skips
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
* cehoyos/master:
fate: Fix ffprobe tests with --target-path set.
doc/ffmpeg: Try to clarify that the input option -r is not the same as -framerate.
Do not print a useless error number if mov header reading fails.
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit '5caf039ba2b4be067568a30146f29008d8db28d0':
Prepare for 11_beta2 Release
Conflicts:
RELEASE
Not merged as theres no FFmpeg 11_beta2
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit 'f9f34cb9983ec6f4ef119c34b726d3b39c143110':
ogg: Use separate classes for the aliases
Conflicts:
libavformat/oggenc.c
See: 2ccc6ff03a
Merged-by: Michael Niedermayer <michaelni@gmx.at>
This option facilitates testing shared libarary builds: for instance
fate builders do no longer need to set LD_LIBRARY_PATH as the binaries will
get the right search paths hardcoded into their executable file.
This option is only meant to be used for testing purposes: The installed
libraries must not move around in the file system, and doing so will
cause a lot of subtle problems. For more information why using RPATH is
dangerous, please refer to
https://blog.flameeyes.eu/2010/06/the-why-and-how-of-rpath
Unfortunately this was not explicitly documented and thus
might be risky.
But all uses I could find in FFmpeg and one in VLC had a memleak
in these cases, and I could not find any that relied on the previous
behaviour.
Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>
~560 → ~500 decicycles
This is following the comments from Michael in
https://ffmpeg.org/pipermail/ffmpeg-devel/2014-August/160599.html
Using 2 registers for accumulator didn't help. On the other hand,
some re-ordering between the movs and psadbw allowed going ~538 to ~500.
Not having allocated it is not a good reason to leave the object
in an undetermined state. Though a particular setting like the
AV_EF_* flags could be useful to control that behaviour.
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Also correctly namespace other functions in vidstabutils, and decrease
difference from Libav.
Initial-patch-by: Vittorio Giovara <vittorio.giovara@gmail.com>
Signed-off-by: Timothy Gu <timothygu99@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
* commit '3526ab891c28396ada8b58bf7647309bab30de1d':
qt-faststart: Undefine fseeko/ftello before defining them
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit '051aadeed104ecbe8ee4850ec2d7e5394f5e1ccd':
ogg: Provide aliases for Speex, Opus and audio-only ogg
Conflicts:
Changelog
libavformat/oggenc.c
libavformat/version.h
See: 2ccc6ff03a
Merged-by: Michael Niedermayer <michaelni@gmx.at>
In some cases, in particular if several blocks are needed because of
the channel layout (e.g. 2.1), the information used to write the
trailing bits terminating the sample data was not reset.
This would cause potential desync on the decoder, although decoded
samples were actually mostly fine.
Fixes ticket #3879.
* commit '4d6c5152849e23a4cc0f6a6ac2880c01ebcd301b':
electronicarts: do not fail on zero-sized chunks
Conflicts:
libavformat/electronicarts.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit 'dc4b2e7d33903a6b9380e8a84b22b3a20facbb08':
rv34: use ff_mpeg_update_thread_context only when decoder is fully initialized
Merged-by: Michael Niedermayer <michaelni@gmx.at>
Clément Bœsch
Luca Barbato
Paul B Mahol
Christophe Gisquet
Michael Niedermayer
ThomasVolkert
Carl Eugen Hoyos
James Almer
Clément Bœsch
Reinhard Tartler
Reimar Döffinger
Muhammad Faiz
Timothy Gu
Diego Biurrun