summaryrefslogtreecommitdiff
path: root/share
diff options
context:
space:
mode:
authorMassimiliano Culpo <massimiliano.culpo@gmail.com>2018-01-10 17:14:17 +0100
committerAdam J. Stewart <ajstewart426@gmail.com>2018-01-10 10:14:17 -0600
commit596d463714d46929f81db62b171c5a772679bb12 (patch)
treec6b15a9fcfc07af79db3cd686c391094146d6215 /share
parentd17a10c6ac07f372bb27ef61b25edb0407c08b83 (diff)
downloadspack-596d463714d46929f81db62b171c5a772679bb12.tar.gz
spack-596d463714d46929f81db62b171c5a772679bb12.tar.bz2
spack-596d463714d46929f81db62b171c5a772679bb12.tar.xz
spack-596d463714d46929f81db62b171c5a772679bb12.zip
Travis: use --concurrency=multiprocessing only on build tests (#6872)
On a local workstation, it seems that tracking multiple processes during coverage may result in malformed coverage reports for unit tests and not for build tests. Given that multiple processes make a difference in coverage mainly for build tests, try to disable the tracking for unit tests to see if we get more stable coverage results.
Diffstat (limited to 'share')
-rwxr-xr-xshare/spack/qa/setup.sh6
1 files changed, 5 insertions, 1 deletions
diff --git a/share/spack/qa/setup.sh b/share/spack/qa/setup.sh
index 98c79a0457..4b3f98e330 100755
--- a/share/spack/qa/setup.sh
+++ b/share/spack/qa/setup.sh
@@ -11,10 +11,14 @@ SPACK_ROOT="$QA_DIR/../../.."
. "$SPACK_ROOT/share/spack/setup-env.sh"
# Set up some variables for running coverage tests.
-if [[ "$COVERAGE" == true ]]; then
+if [[ "$COVERAGE" == "true" && "$TEST_SUITE" == "unit" ]]; then
coverage=coverage
coverage_run="coverage run"
coverage_combine="coverage combine"
+elif [[ "$COVERAGE" == "true" && "$TEST_SUITE" == "build" ]]; then
+ coverage=coverage
+ coverage_run="coverage run --concurrency=multiprocessing"
+ coverage_combine="coverage combine"
else
coverage=""
coverage_run=""