Digger:采用新路径记录方法的fuzzing工具

Digger: a Fuzzer with an Improved Path-Recording Method

  • 摘要: fuzzing和混合执行是二进制代码分析和测试领域的两个重要技术.为了探索结合这两种技术的潜力,并找到代码中的深层缺陷,本文提出了一种新的路径记录方法,以推进混合执行来辅助fuzzing,并在工具Digger中实现了该方法,Digger能够分析测试接收文件输入或使用额外参数的目标程序.在实际应用程序(如coreutils,catdvi等)上评估了Digger,结果表明,通过使用新的路径记录方法,相比现有成熟工具Driller,Digger能够达到更高的代码覆盖率并找到更多的缺陷.

     

    Abstract: Fuzzing and concolic execution are two important techniques in the area of binary analyzing and testing,and they are able to complement each other well.To find deep bugs in the code,a new path-recording approach was proposed to make concolic execution assist fuzzing.And the approach was implemented in a tool called Digger.The Digger was arranged to deal with target programs that require input from files or with extra arguments.The Digger was evaluated on an example program as well as real-world applications (e.g.,coreutils,catdvi,etc.).Experimental results show that by using the new path-recording approach,Digger can reach higher code coverage and find more bugs than a well-established tool,i.e.,Driller.

     

/

返回文章
返回