(In reply to Chris Wilson from comment #1) > Just the test doesn't understand the new format. Should we keep this as a regression and mark it as fixed once parser is corrected?