Running "make recheck" after the test suite fails hides intermittent
system test failures in GitLab CI. This makes it hard to identify which
branches are affected by a particular test failure mode and causes CI
results to be overly optimistic. Prevent "make recheck" from being run
when "make check" fails to ensure GitLab CI results properly reflect the
stability of the "main" branch.
- *setup_softhsm
script:
- cd bin/tests/system
- - make -j${TEST_PARALLEL_JOBS:-1} -k check V=1 || make -j${TEST_PARALLEL_JOBS:-1} -k recheck V=1
+ - make -j${TEST_PARALLEL_JOBS:-1} -k check V=1
after_script:
- cat bin/tests/system/test-suite.log