Test Cases Generating And Prioritizating Of Embedded Software Defect Detecting

Embedded software defect detection is a very important field of software defectdetection research. Contrasting to traditional software defect detection technology,embedded software defect detection consider not only the use of traditional softwaretesting methods, but also the need of real-time and hardware-related issues whiledealing with high reliability requirements of embedded software. Defect detection forembedded software has been widely studied and applied in academia and industry inrecent years.As a key factor driving software testing and defect detection, test casesgeneration and prioritization technology has been a research focus. Currently, most ofthis type of tools on the market is suit to PC platforms, also need the support of sourcecode, while the ones for binary embedded software relatively immature. This paperfirstly researched test cases generation and prioritization used in embedded softwaredefect detection system, and then use the test case in test cases set sequentially todetect software defects dynamically for the purpose of defect detection in embeddedsoftware.The main work in this paper includes four aspects as follows:1、Common software defect detection technology, embedded software defectdetection technology and simulation technology in embedded software developmentare researched.2、Test cases generation method for embedded software defect detection systemare researched. In this paper, we used a test cases generation method based onsymbolic execution.The first step is statically analysis the instructions of the program,then instrument program code, and finally dynamic execution of the program andgenerating test cases by symbolic execution techniques and constraint solvingtechniques. And thus achieve the purpose of providing reliable test cases data forsubsequent software defect detection.3、Test cases prioritization method for embedded software defect detectionsystem are researched. Firstly, we get function calling relations through staticallyanalysis. Secondly,by running functional dependencies TCP algorithm to construct thefinal reachable matrix of order n. According to the final reachable matrix, we get test case prioritization and accomplish the aim of sorting test cases.4、The entire embedded software defect detection work environment was builtbase on the open-source Linux-based dynamic instrumentation framework Valgrind,STP constraint solver and QEMU simulation software. We developed the test casesmodule using C and python programming language. Then analyzed the coverage anddefect detection capabilities of test cases set. The function of defect detection inembedded software was realized mostly.