When running regression tests in paralle, this one occasionally fails
because uniq exits with status 0. I believe this is because the test is
a bit racy: it assumes that true(1) will exit before uniq writes to
standard out.

It is easy to reproduce this failure by running the test in a loop and
running a buildkernel in the background. Make the race smaller by
sleeping for a bit. Maybe there's a cleverer way to handle this?