Abstract: For buffer overflow vulnerability caused by circulating copies of memory, this paper proposes a context-sensitive inter-process analysis and detection model. Through a series of static analysis of binary code, inter-process analysis using data provided by the inter-process analysis, it mines the buffer overflow prone points. This detection model is based on BinNavi open platform, implemented as a plug, is able to screen accurately spill-prone points and effectively reduces false and negative cases.

Key words: buffer overflow, loop, prone points, process analysis, function summary

摘要： 针对循环拷贝内存引发的缓冲区溢出漏洞，提出了一种上下文相关的过程间分析检测模型，通过对二进制代码进行一系列的静态分析，使用过程间分析提供的数据交互关系，对缓冲区溢出易发点进行挖掘。这种检测模型基于BinNavi的开放平台，以插件形式实现，能够对溢出易发点进行精确的筛选，有效地减少误报漏报情况。

关键词: 缓冲区溢出, 循环, 易发点, 过程间分析, 函数摘要