Enhanced RTL Test Point Insertion Flow with Improved Functional Block Testability

This paper introduces a novel methodology to extend Register Transfer Level (RTL) test point insertion (TPI) capabilities to functional blocks (FB) encompassing operators such as comparators, arithmetic and logical reduction operators, and shifters. The proposed methodology leverages a function-call-based technique for precise back-annotation of functional block expressions and the insertion of test point logic at targeted nodes. By effectively isolating complex regions within RTL designs and enabling the placement of control points and observe points, this approach achieves higher granularity in improving test coverage while preserving design optimization during synthesis. Additionally, it addresses challenges related to expression and logic sharing, ensuring correctness when RTL contexts overlap or diverge, with minimal impact on design area and performance, post-testpoint insertion. Furthermore, the methodology provides actionable feedback to RTL designers to identify locations unsuitable for test point insertion, thereby refining testability at RTL. Experimental results demonstrate the advantages of enhanced TPI to functional blocks which include reduced pattern count and minimal area overhead in comparison to gate level (GL) TPI.