Closed astuyve closed 1 month ago
All modified and coverable lines are covered by tests :white_check_mark:
Project coverage is 71.97%. Comparing base (
abc6cd6
) to head (f68d2aa
).
Benchmark execution time: 2024-10-02 20:28:09
Comparing candidate commit f68d2aa in PR branch aj/dsd-disable-miri
with baseline commit abc6cd6 in branch main
.
Found 1 performance improvements and 1 performance regressions! Performance is the same for 49 metrics, 2 unstable metrics.
execution_time
[+25.709ns; +35.613ns] or [+2.156%; +2.986%]execution_time
[-67.574ns; -61.207ns] or [-2.450%; -2.219%]Omitted due to size.
What does this PR do?
Disables miri for dogstatsd. Also uses
collect
instead offor_each
+push
which Levi thinks could avoidmemcpy
s.Motivation
dogstatsd tests were really slow under miri in CI, like 20,000 times slower (higher than normal for miri).
Additional Notes
Anything else we should know when reviewing?
How to test the change?
Describe here in detail how the change can be validated.