It has already been demonstrated that the de Bruijn method has benefits
over the current implementation: commit 971d12b7f9.
That commit implemented it for long long, this extends it to the int version.
Tested with FATE.
Signed-off-by: Ganesh Ajjanagadde <gajjanagadde@gmail.com>
Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
This reduces the memory & cache need from 256 to 64 bytes
the code also seems faster with this change
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
This uses Stein's binary GCD algorithm:
https://en.wikipedia.org/wiki/Binary_GCD_algorithm
to get a roughly 4x speedup over Euclidean GCD on standard architectures
with a compiler intrinsic for ctzll, and a roughly 2x speedup otherwise.
At the moment, the compiler intrinsic is used on GCC and Clang due to
its easy availability.
Quick note regarding overflow: yes, subtractions on int64_t can, but the
llabs takes care of that. The llabs is also guaranteed to be safe, with
no annoying INT64_MIN business since INT64_MIN being a power of 2, is
shifted down before being sent to llabs.
The binary GCD needs ff_ctzll, an extension of ff_ctz for long long (int64_t). On
GCC, this is provided by a built-in. On Microsoft, there is a
BitScanForward64 analog of BitScanForward that should work; but I can't confirm.
Apparently it is not available on 32 bit builds; so this may or may not
work correctly. On Intel, per the documentation there is only an
intrinsic for _bit_scan_forward and people have posted on forums
regarding _bit_scan_forward64, but often their documentation is
woeful. Again, I don't have it, so I can't test.
As such, to be safe, for now only the GCC/Clang intrinsic is added, the rest
use a compiled version based on the De-Bruijn method of Leiserson et al:
http://supertech.csail.mit.edu/papers/debruijn.pdf.
Tested with FATE, sample benchmark (x86-64, GCC 5.2.0, Haswell)
with a START_TIMER and STOP_TIMER in libavutil/rationsl.c, followed by a
make fate.
aac-am00_88.err:
builtin:
714 decicycles in av_gcd, 4095 runs, 1 skips
de-bruijn:
1440 decicycles in av_gcd, 4096 runs, 0 skips
previous:
2889 decicycles in av_gcd, 4096 runs, 0 skips
Signed-off-by: Ganesh Ajjanagadde <gajjanagadde@gmail.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Intel compiler also defines __GNUC__, so the Intel specific intrinsics were not
really being used.
Reviewed-by: Michael Niedermayer <michaelni@gmx.at>
Signed-off-by: James Almer <jamrial@gmail.com>
This removes inline av_log2 and av_log2_16bit from the public API,
instead exporting them as regular functions. In-tree code still
gets the inline and otherwise optimised variants.
Signed-off-by: Mans Rullgard <mans@mansr.com>
GCC 4.3 and later do the right thing with the plain C code. Earlier
versions in 32-bit mode generate one extra instruction, needlessly
zeroing what would be the high half of the shifted value. At least
two gcc configurations miscompile the inline asm in some situations.
In 64-bit mode, all gcc versions generate imul r64, r64 followed by
shr. On Intel i7 and later, this imul is faster 32-bit mul. On
older Intel and all AMD, it is slightly slower. On Atom it is much
slower.
Considering where the FASTDIV macro is used, any overall negative
performance impact of this change should be negligible. If anyone
cares, they should file a bug against gcc and get the instruction
selection fixed.
Signed-off-by: Mans Rullgard <mans@mansr.com>
There is no point in having the user disable any fastdiv macros.
Besides the condition implementation was broken and only disabled
the C implementation, but no platform specific assembly versions.
This is a bit hackish. I will try to think of something nicer, but
this will do for now.
Originally committed as revision 22366 to svn://svn.ffmpeg.org/ffmpeg/trunk
Consistently apply this rule: the guard name is obtained from the
filename by stripping the leading "lib", converting '/' and '.' to
'_' and uppercasing the resulting name. Guard names in the root
directory have to be prefixed by "FFMPEG_".
Originally committed as revision 15120 to svn://svn.ffmpeg.org/ffmpeg/trunk
Patch by Zuxy Meng: zuxy meng gmail com
Original thread:
[FFmpeg-devel] [PATCH] Pure, const and malloc attributes to libavutil
Date: 03/18/2008 6:09 AM
Originally committed as revision 12489 to svn://svn.ffmpeg.org/ffmpeg/trunk
Ganesh Ajjanagadde
Ronald S. Bultje
Michael Niedermayer
Timothy Gu
James Almer
Michael Niedermayer
Matthew Oliver
Reimar Döffinger
Justin Ruggles
Mans Rullgard
Diego Biurrun
Luca Barbato
Stefano Sabatini
Zuxy Meng
Benoit Fouet