Files
ffmpeg/libavutil
Giorgio Vazzana d0a34aeedf md5: optimize second round by using 4-operation form of G()
4-operation form is preferred over 3-operation because it breaks a long
dependency chain, thus allowing a superscalar processor to execute more
operations in parallel.
The idea was taken from: http://www.zorinaq.com/papers/md5-amd64.html

AMD Athlon(tm) II X3 450 Processor, x86_64

$ for i in $(seq 1 4); do ./avutil_md5_test2; done
size: 1048576  runs: 1024  time:    5.821 +- 0.019
size: 1048576  runs: 1024  time:    5.822 +- 0.019
size: 1048576  runs: 1024  time:    5.841 +- 0.018
size: 1048576  runs: 1024  time:    5.821 +- 0.018

$ for i in $(seq 1 4); do ./avutil_md5_test2; done
size: 1048576  runs: 1024  time:    5.646 +- 0.019
size: 1048576  runs: 1024  time:    5.646 +- 0.018
size: 1048576  runs: 1024  time:    5.642 +- 0.019
size: 1048576  runs: 1024  time:    5.641 +- 0.019

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
2013-05-20 04:27:34 +02:00
..
2012-10-28 14:33:41 +01:00
2012-10-28 14:33:41 +01:00
2012-07-16 19:30:13 +00:00
2012-12-05 20:25:32 +01:00
2013-03-07 01:12:04 +01:00
2013-03-07 01:12:04 +01:00
2013-04-13 23:48:26 +02:00
2012-10-31 17:01:52 +01:00
2013-01-24 00:49:47 +01:00
2013-03-19 00:47:00 +01:00
2013-05-15 23:53:40 +02:00
2013-05-13 21:42:37 +02:00
2012-08-09 01:44:15 +02:00
2013-05-17 21:16:02 +02:00
2013-05-13 13:52:31 +02:00
2013-05-13 13:52:31 +02:00
2013-05-15 12:10:09 +02:00
2013-05-13 21:42:37 +02:00
2013-05-17 11:32:53 +00:00
2013-05-17 11:32:53 +00:00
2013-04-15 20:31:06 +02:00
2012-10-28 14:33:41 +01:00
2012-10-31 03:10:35 +01:00
2012-10-28 14:33:41 +01:00
2013-05-17 11:32:53 +00:00