Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
The pathological benchmark runs too fast to be meaningful, and because it only contains a single path, it doesn't make use of multi-threading like the other test cases do. So, let's set up a new baseline by adding more paths to the test case and running it enough times for it to be within the same order of magnitude as its immediate neighbors. Before: user system total real pathological 0.000000 0.000000 0.000000 ( 0.002413) command-t 0.380000 0.000000 0.380000 ( 0.380633) chromium (subset) 0.850000 0.010000 0.860000 ( 0.441269) chromium (whole) 3.400000 0.010000 3.410000 ( 1.342450) After: user system total real pathological 0.110000 0.000000 0.110000 ( 0.112132) command-t 0.340000 0.000000 0.340000 ( 0.340255) chromium (subset) 0.800000 0.010000 0.810000 ( 0.398784) chromium (whole) 3.320000 0.010000 3.330000 ( 1.236914) As you can see there is some flutter between runs (the timings for the other test cases shouldn't have changed), but I am doing best-of-three runs for all of these, so that's as close as I can get to scientific. Signed-off-by: Greg Hurrell <greg@hurrell.net>
- Loading branch information