Modular Demand-Driven Analysis of Semantic Difference for Program Versions

In this work we present a modular and demand-driven analysis of the semantic difference between program versions. Our analysis characterizes initial states for which final states in the program versions differ. It also characterizes states for which the final states are identical. Such characterizations are useful for regression verification, for revealing security vulnerabilities and for identifying changes in the program’s functionality. Syntactic changes in program versions are often small and local and may apply to procedures that are deep in the call graph. Our approach analyses only those parts of the programs that are affected by the changes. Moreover, the analysis is modular, processing a single pair of procedures at a time. Called procedures are not inlined. Rather, their previously computed summaries and difference summaries are used. For efficiency, procedure summaries and difference summaries can be abstracted and may be refined on demand. We have compared our method to well established tools and observed speedups of one order of magnitude and more. Furthermore, in many cases our tool proves equivalence or finds differences while the others fail to do so.