This paper presents a novel hybrid finite-domain constraint solving engine for RTL circuits. We describe how DPLL search is modified for search in combined integer and Boolean domains by using efficient finite-domain constraint propagation. This enables efficient combination of Boolean SAT and linear integer arithmetic solving techniques. We automatically use control and data-path abstraction in RTL descriptions. We use conflict-based learning using the variables on the boundary of control and data-path for additional performance benefits. Finally, we analyze the hybrid constraint solver experimentally using some example circuits.

No abstract