Experiments with Non-Termination Analysis for Java Bytecode

Abstract

Non-termination analysis proves that programs, or parts of a program, do not terminate. This is important since non-termination is often an unexpected behaviour of computer programs and exposes a bug in their code. While research has found ways of proving non-termination of logic programs and of term rewriting systems, this is hardly the case for imperative programs. In this paper, we describe and experiment with a technique for proving non-termination of imperative, bytecode programs by relating their non-termination to that of a (constraint) logic program. Moreover, we show that our non-termination test effectively helps a termination test, by avoiding expensive search for termination proofs of those portions of the code where such proofs do not exist.