If a task fails due to misconfiguration (e.g. because it is missing a required attribute) inside an xharness testcase, this error is treated as a test failure and not as a "fatal" error that would cause xharness to exit. Instead, the test in which the error has occurred is maked as failed and xharness continues with subsequent tests.
As a result this may make it difficult to figure out why and where test run has failed, as the error message may be in the middle of a very long log, or the task is reported as <unknown>. See