The current assumption is that Path.home(), e.g. /home/ubuntu will be located on the ancestry path of the setup_repro_env.py script's absolute path. This is used to locate the test_repos.conf file by walking the file tree, starting the script's file location. We should ensure that the file discovery in discover_test_repos_conf() is robust enough to handle this or else we risk an infinite loop.

The conditions for triggering this loop would be:
1. Repo is cloned somewhere not on the home path
2. Relative paths between test_repos.conf and setup_repro_env.py have been changed from the mongo master locations, e.g. the paths of test_repos.conf and/or setup_repro_env.py were altered in some way such that setup_repro_env.py does not have an ancestor directory that contains test_repos.conf