Efficient Testing of Concurrent Programs with Abstraction-Guided Symbolic Execution

ion-Guided Symbolic Execution Neha Rungta, Eric G Mercer and Willem Visser∗ Dept. of Computer Science, Brigham Young University, Provo, UT 84602, USA ∗SEVEN Networks, 901 Marshall Street, Redwood City, CA 94063, USA Abstract. Exhaustive search techniques such as model checking and Exhaustive search techniques such as model checking and symbolic execution are insufficient to detect errors in concurrent programs. In this work we present an abstraction-guided symbolic execution technique that quickly detects errors in concurrent programs that arise from thread schedules and input data values. An abstract system is generated that contains a set of key program locations that are relevant in testing the feasibility of a possible error in the program. We guide a symbolic execution along locations in the abstract system in an effort to generate a corresponding feasible execution trace to the error location. A combination of heuristics are used to automatically rank thread and data non-determinism in order to guide the execution. We demonstrate empirically that abstraction-guided symbolic execution generates feasible execution paths in the actual system to find concurrency errors in a few seconds where exhaustive symbolic execution fails to find the same errors in an hour.