Project/Area Number |
01460151
|
Research Category |
Grant-in-Aid for General Scientific Research (B)
|
Allocation Type | Single-year Grants |
Research Field |
計算機工学
|
Research Institution | Kyushu University |
Principal Investigator |
USHIJIMA Kazuo Kyushu University, Faculty of Engineering Professor, 工学部, 教授 (40037750)
|
Co-Investigator(Kenkyū-buntansha) |
YOSHIDA Norihiko Kyushu University, Faculty of Engineering Research Associate, 工学部, 助手 (00182775)
FURUKAWA Zengo Kyushu University, Faculty of Engineering Lecturer, 工学部, 講師 (30190135)
HIRABARU Masaki Kyushu University, Faculty of Engineering Lecturer, 工学部, 講師 (10192717)
CHENG Jingde Kyushu University, Faculty of Engineering Associate Professor, 工学部, 助教授 (30217228)
ARAKI Keijiro Kyushu University, Faculty of Engineering Associate Professor, 工学部, 助教授 (40117057)
|
Project Period (FY) |
1989 – 1990
|
Project Status |
Completed (Fiscal Year 1990)
|
Budget Amount *help |
¥7,000,000 (Direct Cost: ¥7,000,000)
Fiscal Year 1990: ¥500,000 (Direct Cost: ¥500,000)
Fiscal Year 1989: ¥6,500,000 (Direct Cost: ¥6,500,000)
|
Keywords | Concurrent Programs / Specification / Validation / Testing / Debugging / Deadlock / Distributed Cooperative Processing / Distributed RunーTime Environment / 実時間処理 / プログラム変換 |
Research Abstract |
We investigated validation methods for concurrent programs and validation support tools, and obtained the following results on semantics, testing methods, development methodology, and implementation of concurrent programs. 1. We implemented an eventーdriven execution monitor EDEN for concurrent Ada programs, which monitors tasking behavior of a concurrent Ada program with minimal interference. We designed an algorithm to detect all kinds of deadlocks in concurrent Ada programs, and realized it in the monitoring tool EDEN. 2. We proposed two kinds of criteria for testing a concurrent Ada program, i.e., a criterion inside each parallel process and a criterion among processes. We examined program errors according to the criteria. We also implemented prototype tools to measure testing coverage and to generate test cases. 3. A concurrent system can be regarded as a network of parallel processes which are connected with streams of messages, signals or events. We found out that the functional programming language Miranda works well as an executable specification language to describe and analyze concurrent systems in terms of stream processing functions. 4. We proposed a framework for distributed cooperative processing. A distributed cooperative processing system consists of autonomous processes, each of which is associated with a field. Several kinds of group behaviors are well described within the framework. We also implemented a prototype system which realizes such processes on a distributed system. 5. We designed a distributed programming language and its processor, which makes a single parallel program as a whole run on a distributed system. We emulated a distributed runーtime system which manages such a parallel program on a distributed system consisting of workstations connected with LAN.
|