diff options
| author | drebs <drebs@leap.se> | 2017-07-07 18:17:38 -0300 |
|---|---|---|
| committer | drebs <drebs@leap.se> | 2017-07-08 07:48:47 -0300 |
| commit | f21d9e1a4e6d0a2ed6d7af83191e39a2a883241b (patch) | |
| tree | a3d917c93c4ed7f11889182094172924d2759a9c /testing/tox.ini | |
| parent | 16d8677ad67ceb0cdf78f3b0a7d6b4e4dae4b624 (diff) | |
[benchmarks] run benchmarks twice, for time and resources
We noticed that instrumentation added for watching resources has an
impact in time statistics (i.e. it increases average and stddev). This
commit makes the benchmark tests run twice: once for measuring time and
a second time for measuring resources.
Diffstat (limited to 'testing/tox.ini')
| -rw-r--r-- | testing/tox.ini | 2 |
1 files changed, 2 insertions, 0 deletions
diff --git a/testing/tox.ini b/testing/tox.ini index 9aaa1ae3..88b27d48 100644 --- a/testing/tox.ini +++ b/testing/tox.ini @@ -76,7 +76,9 @@ commands = # we must make sure that installed pysqlcipher was built with the HAVE_USLEEP # flag, or we might have problems with concurrent db access. ./ensure-pysqlcipher-has-usleep.sh +# run benchmarks twice: once for time and a second time for resources py.test --benchmark-only {posargs} + py.test --benchmark-only --watch-resources {posargs} passenv = HOST_HOSTNAME [testenv:code-check] |