面向细粒度随机化的ROP攻击响应方案OA

Return-oriented programming attack(ROP)can bypass data execute disable(DEP)defense and pose a major threat to software security.ROP detection methods on runtime induce the false positives and false negatives,and it is difficult to infer ROP attacks from the context information of program crashes.Therefore,this paper proposes a ROP attack response scheme oriented to fine-grained randomization.In this scheme,an exception capture component was embedded in the program to capture rich crash context information in time.Meanwhile,15 ROP attack indicators associated with ROP payloads and attack modes were designed.This scheme used simulated attack instances and real vulnerability instances to verify the scheme indicators,used fuzzy testing to verify the attack differences of the indicators,and proposed ROP attack utilization modes for different vulnerability types.This scheme built a bridge between fine-grained randomization security mechanism and rapid response to ROP attacks.