Concurrent Program Verification with Invariant-Guided Underapproximation

Abstract

Automatic verification of concurrent programs written in low-level languages like ANSI-C is an important task as multi-core architectures are gaining widespread adoption. Formal verification, although very valuable for this domain, rapidly runs into the state-explosion problem due to multiple thread interleavings. Recently, Bounded Model Checking (BMC) has been used for this purpose, which does not scale in practice. In this work, we develop a method to further constrain the search space for BMC techniques using underapproximations of data flow of shared memory and lazy demand-driven refinement of the approximation. A novel contribution of our method is that our underapproximation is guided by likely data-flow invariants mined from dynamic analysis and our refinement is based on proof-based learning. We have implemented our method in a prototype tool. Initial experiments on benchmark examples show potential performance benefit.