Sorry, I misread the logs - it was the wait4 variant of the test that succeeded in the test run after the wait6 one failed. In any case, this simply looks like flaky/racy tests, or ptrace(), (and certainly unrelated to the listed commits). kre