Automatically Inferring Quantified Loop Invariants by Algorithmic Learning from Simple Templates

Abstract. By combining algorithmic learning, decision procedures, predicate abstraction, and simple templates, we present an automated technique for finding quantified loop invariants. Our technique can find arbitrary first-order invariants (modulo a fixed set of atomic propositions and an underlying SMT solver) in the form of the given template and exploits the flexibility in invariants by a simple randomized mechanism. The proposed technique is able to find quantified invariants for loops from the Linux source, as well as for the benchmark code used in the previous works. Our contribution is a simpler technique than the previous works yet with a reasonable derivation power.

• Type: Conference
• Year: 2010
• Where: APLAS
• Authors: Soonho Kong, Yungbum Jung, Cristina David, Bow-Yaw Wang, Kwangkeun Yi
• Keywords: APLAS 2010, Arbitrary First-order Invariants, Decision Procedures, Programming Languages, Simple Randomized Mechanism