Assessing the Effectiveness of Software-Based Self-Test Programs for Static Cell-Aware Test

Software-Based Self-Test (SBST) is vastly adopted as a hardware safety mechanism for the in-field test of safety-critical systems in the form of Software Test Libraries (STLs). Typically, an STL's diagnostic coverage is evaluated on the stuck-at fault model. As various defect-oriented fault models exist and are used for manufacturing testing, such as the popular cell-aware test (CAT), there is a need to evaluate the effectiveness of SBST when such models are targeted. This work targets static CAT faults. We evaluated the fault coverage of open-available STLs for a RISC-V SoC. We used results stemming from stuck-at fault simulation and gate-exhaustive simulation to elaborate on the obtained results.