Redundant constraints elimination for symbolic execution

Constraints solving plays an important role in symbolic execution, in fact, it often dominates the cost in overall performance. However, in despite of significant advances have been made over the past few years, constraint solving is still the main bottleneck in symbolic execution. Reducing the overhead before calling a constraint solver by exploring domain specific insights have been an important issue. In this paper, we propose a novel approach to eliminate redundant constraints in symbolic execution. We transform the redundant constraints detection problem in a path condition (PC) to the redundancy identification problem in linear programming (LP), and then used an efficient algorithm to eliminate the redundant constraints in LP. We integrated our approach into Green, which is based on symbolic Pathfinder, a popular symbolic execution framework for JAVA programs. We have evaluated the tool on a set of publicly available software and the results showed that our approach can significantly reduce redundant constraints and ultimately improve the performance of symbolic execution.