A Runtime Monitoring Based Fuzzing Framework for Temporal Properties

Abstract

The detection of the bugs specified in temporal properties is difficult for the existing fuzzers. These bugs are triggered when the program executions contain some specific sequences of operations. This extended abstract reports our recent progress of a runtime monitoring-based fuzzing framework towards the bugs expressed as temporal properties. Specifically, we propose two novel algorithms for preserving input mutants and mutating the input seed to improve fuzzing's efficiency. We have implemented a prototype for Java programs and carried out experiments on real-world open-source Java programs. The preliminary experimental results indicate the promising of our fuzzing method.