计算机与现代化 ›› 2013, Vol. 1 ›› Issue (3): 74-77.doi:

• 数据库与数据挖掘 • 上一篇    下一篇

基于符号执行和数据挖掘的路径可达性检测研究

范 彧   

  1. 上海交通大学软件学院,上海200240
  • 收稿日期:2012-11-16 修回日期:1900-01-01 出版日期:2013-04-03 发布日期:2013-04-03

Research on Path Feasibility Detection Based on Symbolic Execution and Data Mining

FAN Yu   

  1. School of Software, Shanghai Jiao Tong University, Shanghai 200240, China
  • Received:2012-11-16 Revised:1900-01-01 Online:2013-04-03 Published:2013-04-03

摘要: 要把静态分析技术真正应用于软件工程的实践中,程序中的不可到达路径依旧是一个巨大的障碍。在缺乏不可到达路径信息的情况下,数据流信息只能以一种非常保守的方式被利用,因而使进一步的工作,例如测试数据生成或者软件缺陷检测变得更加困难和低效。本文提出一种融合符号执行以及数据挖掘的混合型方法来检测程序中的不可到达路径,充分利用符号执行的精确性以及数据挖掘的高效性,这种方法被证明是有效并且实用的。

关键词: 符号执行, 数据挖掘, 程序路径, 可达性, 不可到达路径

Abstract: The presence of infeasible program paths remains a big obstacle in applying static analysis to software engineering activities. Without the knowledge of the infeasible path problem, data flow information can only be treated in a rather conservative way, which makes further works such as test data generation or bug finding much harder and more inefficient. This paper presents a hybrid approach, combining symbolic execution and data mining techniques for infeasible path detection. Taking advantages of the precise of symbolic execution and the efficiency of data mining, this approach proves to be effective and practical.

Key words: symbolic execution, data mining, program path, feasibility, infeasible path