Previously ignoring a test with subtests had no effect, as the failures are reported against the subtests.