Comment 2 for bug 1942113

The difference I can see between the passing and failing run is that in the failing run, systemd and glibc are upgraded before the "rebooting testbed after setup commands that affected boot" and in the passing run they are upgraded after that step. That doesn't make a lot of sense but maybe the test depends on some boot time actions that the newer libc/systemd are not doing correctly?

The fact that this doesn't reproduce outside the autopkgtest infrastructure makes me less worried about the consequences of hinting past this but also having systemd hinted is not a great place to be.