A test-case generation method for concurrent programs including task-types
Tetsuro Katayama, Zengo Furukawa, Kentaro Ushijima · 2002
A task-type in concurrent programs is a template of task-instances which are dynamically generated in execution of the programs. The number of the generated task-instances may not be statically determined. Therefore, test-cases generated by our previously proposed method may not be realized in execution of the programs. This paper proposes an enhanced test-case generation method. Concurrent programs are classified into four kinds under two viewpoints, whether the number of task-instances generated from a task-type can be determined statically or dynamically, and whether or not the task-instances communicate with each other. The enhanced method uses a modified Event InterActions Graph (EIAG). The EIAG consists of event graphs and interactions. An event graph is an abstract control flow graph of a program unit. The interactions represent interactions between the units. Because the modified EIAG represents more accurate behavior of a program than the EIAG does, test-cases generated by the enhanced method increase the possibility of realizing themselves. The test-cases in the method increase in number. The method, however, is practical in the sense that the number of task-instances generated from a task-type is restricted.