[FFmpeg-devel] [PATCH] md5: optimize second round by using 4-operation form of G()

[Giorgio Vazzana]

2013/5/18 Giorgio Vazzana <mywing81 at gmail.com>:
> Hello,
>
> here's another trick we could use to improve speed.
> Nicolas, if you have time, could you run the benchmark again to see if
> it's worth including this? Thanks.

Ok, here's some benchmark results (test based on Nicolas' code):

AMD Athlon(tm) II X3 450 Processor, x86_64

# without patch
$ 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

# with patch
$ 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