The designer of a computing system should adopt explicit criteria for accepting or rejecting proposed system features. Three possible criteria af this kind are input recordability, input specifiability, and asynchrono...
详细信息
The designer of a computing system should adopt explicit criteria for accepting or rejecting proposed system features. Three possible criteria af this kind are input recordability, input specifiability, and asynchronous reproducibility of output. These criteria imply that a user can, if he desires, either know or control all the influences affecting the content and extent of his computer's output. To define the scope of the criteria, the notion of an abstract machine of a programming language and the notion of a virtual computer are explained. Examples of applications of the criteria concern the reading of a time-of-day clock, the synchronization of parallel processes, protection in multipragrammed systems, and the assignment of capability indexes. [ABSTRACT FROM AUTHOR]
Testing and debugging are very important tasks in software development. Fault localization is a very critical activity in the debugging process and also is one of the most difficult and time-consuming activities. The ...
详细信息
Testing and debugging are very important tasks in software development. Fault localization is a very critical activity in the debugging process and also is one of the most difficult and time-consuming activities. The demand for effective fault localization techniques that can aid developers to the location of faults is high. In this paper, a fault localization technique based on complex network theory named FLCN-S is proposed to improve localization effectiveness on single -fault subject programs. The proposed technique diagnoses and ranks faulty program statements based on their behavioral anomalies and distance between statements in failed tests execution by utilizing two network centrality measures (degree centrality and closeness centrality). The proposed technique is evaluated on a well-known standard benchmark (Siemens test suite) and four Unix real -life utility subject programs ( gzip, sed, flex, and grep ). Overall, the results show that FLCN-S is significantly more effective in locating faults in comparison with other techniques. Furthermore, we observed that both degree and closeness centrality play a vital role in the identification of faults.
Based on analyzing QF and QS, a software debugging method based on event injection is designed, in this way, the software debugging can be more initiative, more targeted and automatically, this method can improve the ...
详细信息
Based on analyzing QF and QS, a software debugging method based on event injection is designed, in this way, the software debugging can be more initiative, more targeted and automatically, this method can improve the efficiency of software debugging. The paper holds certain referential value and practical significance in promoting the development of Quantum programming.
A fast conditional breakpoint is considered to be an essential process control capability of advanced high-performance parallel debuggers. The fundamental principles for finding and implementing fast conditional break...
详细信息
A fast conditional breakpoint is considered to be an essential process control capability of advanced high-performance parallel debuggers. The fundamental principles for finding and implementing fast conditional breakpoints are introduced and the design problems associated with their implementations such as code generation, code instrumentation and source program mapping are discussed. Dynamic code instrumentation, combined with hybrid code generation and source breakpoint identification, provides a well-suited solution to those problems. Using an enhanced version of the Dyninst run-time code patching library, a fast conditional breakpoint in a parallel debugger, XBUSTER, is implemented. Compared with GDB experimental results show that XBUSTER can debug a program with a higher efficiency. Compared with other fast conditional breakpoint implementations based on dynamic instrumentation technique, XBUSTER is more portable and functional.
Because of complexity of parallel debugging and versatility of object model, it makes the organic combination of the different parallel debugged tools difficult. It is necessary to propose the unified debugged object ...
详细信息
Because of complexity of parallel debugging and versatility of object model, it makes the organic combination of the different parallel debugged tools difficult. It is necessary to propose the unified debugged object model to simplify debugging work. The parallel virtual machine (PVM) system was analyzed and the concrete actions were defined, which was used to describe the unified PVM debugged object model. The relation of view and program, the relation of event and program, and description of the basic event were elaborated. The debugging can be performed by using different debugged methods on same debugged object. It makes parallel debugging more easy.
Because the program is running in a distributed way, both the concurrent features and the runtime environments should be taken into account when testing a distributed program. A Java oriented distributed program test ...
详细信息
Because the program is running in a distributed way, both the concurrent features and the runtime environments should be taken into account when testing a distributed program. A Java oriented distributed program test system is put forward, which is called JDPT (Java-oriented distributed program testing system). JDPT defines events based on the runtime environments, records execution processes of the distributed program as event sequences, and defines event sequencing constraints to check the validity of those feasible set of event sequences. By the technology, the JDPT can effectively estimate the correctness of concurrent executions of the tested program within distributed environments, and can be used to test Java programs running upon multiple platforms.
Parallel programs can not be debugged as sequential ones repeatly because of their nondetermination, so tracing and replay become the key technique of parallel debugging, and the most overhead of this technique is tra...
详细信息
Parallel programs can not be debugged as sequential ones repeatly because of their nondetermination, so tracing and replay become the key technique of parallel debugging, and the most overhead of this technique is tracing the events. This paper improves the optimal tracing method provided by Rober Netzer etc. by combining the characteristics of PVM (parallel virtual machine), makes the events tracing optimal even when non_transitive message race appears. The improved method has been implemented in PVM_based parallel clebugger.
暂无评论