A loop summarization method for dynamic binary program analysis

动态测试数据生成方法相对于传统Fuzz测试方法能有效的提高软件测试与漏洞分析的效率。本文针对动态测试数据生成过程中对循环进行处理时的路径覆盖效率较低与约束求解困难的问题，提出了一种使用归纳变量构建循环，并通过符号计算提取循环的方法。本文通过将循环应用于软件动态分析过程中，验证了使用循环能有效的提高约束求解与循环路径遍历的效率。 Compared to the traditional Fuzz testing ,dynamic test generation can improve the efficiency of software testing and vulnerability analysis .This paper focuses on the performance of path covering and constraint solving in dynamic test genera-tion ,and proposes a method which constructs loop summary with induce variables and abstracts loop summary by symbolic computa-tion .This paper uses the loop summary in software dynamic analysis ,and proves that the loop summary can improve the efficiency of loop constraint solving and loop path traversal .Compared to the traditional Fuzz testing, dynamic test generation can improve the efficiency of software testing and vulnerability analysis. This paper focuses on the performance of path covering and constraint solving in dynamic test generation, and proposes a method which constructs loop summary with induce variables and abstracts loop summary by symbolic computation. This paper uses the loop summary in software dynamic analysis, and proves that the loop summary can improve the efficiency of loop constraint solving and loop path traversal