Research And Optimization On The Automated Test Case Generation System

With the development of software technology, software testing technology is increasingly important for software developers, and the quality of software has direct impact on company or personal information security. At present, software testing mainly generates some special values for software code, such as the values of various critical points, then uses those values to execute software code and test whether the program code is running like requirements specification, but the current software testing techniques, such as unit testing, most of them need testers to analyze program codes,and write the test driver for each module of software, which will take many manpower and material resources to write those drivers for testing.Symbolic execution is a typical test technology to generate test input data for software program. This technology uses to search many different program paths in a given time, and generates a set of test cases for each path. It can find many errors, which include assertion violation, uncaught exceptions, security vulnerabilities and memory crash. Now symbolic execution technology is still in the stage of academic research,and there are a number of symbolic execution tools designed by some research institutes or universities for research. Symbolic execution, as a way to automatically generate test cases, is a great software testing technology, and has a good prospects in the future. It is likely to be a main testing technology. Therefore, symbolic execution has a good research value in the future.Now the symbolic execution testing tools have existed can only be applied on PC,and for other platforms, symbolic execution technology has not been applied. It has the good research value if this technology can apply to the other platforms. This thesis mainly studies the application of symbolic execution technology in embedded platforms,and designs a symbolic execution tool which can be applied to the embedded platform.This thesis first introduces the background and present situation of symbolic execution,the theory of knowledge, the symbolic execution tools and the problems symbolic execution faced. Then it designs a symbolic execution tool that can be applied to embedded platform, and uses some programs to test whether this tool can run rightly.Finally, this thesis introduces some shortcomings of this tool and prospects the future work.