diff options
| author | Yann Collet <cyan@fb.com> | 2018-02-20 21:05:22 (GMT) |
|---|---|---|
| committer | Yann Collet <cyan@fb.com> | 2018-02-20 21:09:13 (GMT) |
| commit | ae3dededed1c6c1a9b4c631d9edba690c4abb59c (patch) | |
| tree | 247aa24692ecbc82e1979269350e663a7a570c81 /circle.yml | |
| parent | 1a233c5f0fb665fa6b65c856b150bdef92654b42 (diff) | |
| download | lz4-ae3dededed1c6c1a9b4c631d9edba690c4abb59c.zip lz4-ae3dededed1c6c1a9b4c631d9edba690c4abb59c.tar.gz lz4-ae3dededed1c6c1a9b4c631d9edba690c4abb59c.tar.bz2 | |
ensure bench speed measurement is more accurate for small inputs
Previous method would produce too many time() invocations,
becoming a significant fraction of workload measured.
The new strategy is to use time() only once per batch,
and dynamically resize batch size so that each round lasts approximately 1 second.
This only matters for small inputs.
Measurement for large files (such as silesia.tar) are much less impacted
(though decoding speed is so fast that even medium-size files will notice an improvement).
Diffstat (limited to 'circle.yml')
0 files changed, 0 insertions, 0 deletions