TESTING AND METRICS FOR CONCURRENT COMPUTATION BASED ON RENDEZVOUS MECHANISM

Abstract

Testing and metrics are two important approaches to assure the reliability and quality of software industries. Testing and metrics of sequential programs have been fairly sophisticated processes, with various methodologies and tool savailable for use in building and demonstrating the correctness of a program. The emergence of concurrent programming in the recent years, however, introduces new testing problems and difficulties that cannot be solved by testing techinques of traditional sequential programs. One of the difficult tasks is that concurrent programs can have many instances of execution for the same set of input data. Many testing methodologies of concurrent program are proposed to solve controlled execution and determinism. However, there are few discussions of concurrent software testing from an in ter-task perspective. Yet, the common characteristics of concurrent programming are explicit identification of the large grain parallel compputation units (tasks), and the explicit inter-task communication via a rendezvous-style mechanism. This paper focuses testing on the concurrent porrgramming through a task decomposition emchanism. Four testing criteria are proposed to test a concurrent program. Programmers can choose an appropriate testing strategy depending on the properties of the concurrent programs. Associated with the strategies, four equations are provides to measure the complexity of the concurrent programs.