Symbolic Abstract Heaps for Polymorphic Information-flow Guard Inference (Extended Version) (2211.03450v1)

Abstract: In the realm of sound object-oriented program analyses for information-flow control, very few approaches adopt flow-sensitive abstractions of the heap that enable a precise modeling of implicit flows. To tackle this challenge, we advance a new symbolic abstraction approach for modeling the heap in Java-like programs. We use a store-less representation that is parameterized with a family of relations among references to offer various levels of precision based on user preferences. This enables us to automatically infer polymorphic information-flow guards for methods via a co-reachability analysis of a symbolic finite-state system. We instantiate the heap abstraction with three different families of relations. We prove the soundness of our approach and compare the precision and scalability obtained with each instantiated heap domain by using the IFSpec benchmarks and real-life applications.