diff options
| author | Alp Mestanogullari <alpmestan@gmail.com> | 2019-11-12 18:59:07 +0100 |
|---|---|---|
| committer | Ben Gamari <ben@smart-cactus.org> | 2019-11-12 18:40:19 -0500 |
| commit | 643d42fcf68f9b96bc8829d20bcd8f5fef84a78c (patch) | |
| tree | dacd233765c3a026af4ad50846ad6ba0a09c281f /testsuite/driver/testlib.py | |
| parent | f8ec32d79ff876686cb191151c5a151773ae6495 (diff) | |
| download | haskell-643d42fcf68f9b96bc8829d20bcd8f5fef84a78c.tar.gz | |
testsuite: don't collect compiler stats in collect_runtime_residency
We instead want to collect the runtime stats (with collect_stats, instead of
collect_compiler_stats).
This should fix a number of perf tests failures we have been seeing, where
we suddenly started measuring metrics we didn't intend to measure, which
tend to fall outside of the acceptance window.
Metric Decrease:
lazy-bs-alloc
T3586
Metric Increase:
space_leak_001
T4801
T5835
T12791
Diffstat (limited to 'testsuite/driver/testlib.py')
| -rw-r--r-- | testsuite/driver/testlib.py | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/testsuite/driver/testlib.py b/testsuite/driver/testlib.py index 35bb4ad32e..79d504a845 100644 --- a/testsuite/driver/testlib.py +++ b/testsuite/driver/testlib.py @@ -583,7 +583,7 @@ RESIDENCY_OPTS = '+RTS -A256k -i0 -h -RTS' # See Note [Measuring residency]. def collect_runtime_residency(tolerance_pct: float): return [ - collect_compiler_stats(['peak_megabytes_allocated', 'max_bytes_used'], tolerance_pct), + collect_stats(['peak_megabytes_allocated', 'max_bytes_used'], tolerance_pct), extra_run_opts(RESIDENCY_OPTS), # The nonmoving collector does not support -G1 omit_ways([WayName(name) for name in ['nonmoving', 'nonmoving_thr', 'nonmoving_thr_ghc']]) |