Exact pole assignment with regularity by output feedback in descriptor systems—Part 2†