An Emperical Evaluation Of Adequacy Criteria For Testing Concurrent Programs

A concurrent program contains two or more threads that execute concurrently and work together to perform some task. Concurrency increases the efficiency of a program. Testing of concurrent programs has been a challenging task because of the inherent non-determinism. Most approaches proposed for concurrent program testing employ, explicitly or implicitly, a coverage criterion to measure test adequacy. In order to apply those approaches, we must first choose a criterion that suits best for our programs. There is a need for quantitative results of evaluation regarding the effectiveness of the various coverage criteria used for the testing of concurrent programs. Such an evaluation can be useful to select a particular criterion for a specific set of concurrent programs. This thesis tries to evaluate the effectiveness of some of these criteria and puts forth the quantitative results for such an evaluation.