Boosting Symbolic Execution for Heap-based Vulnerability Detection and Exploit Generation
Haoxin Tu · 2023
Heap-based vulnerabilities such as buffer overflow and use after free are severe flaws in various software systems. Detecting heap-based vulnerabilities and demonstrating their severity via generating exploits for them are of critical importance. Existing symbolic execution-based approaches have shown their potential in the above tasks. However, they still have some fundamental limitations in path exploration, memory modeling, and environment modeling, which significantly impede existing symbolic execution engines from efficiently and effectively detecting and exploiting heap-based vulnerabilities. The objective of this thesis is to design and implement a boosted symbolic execution engine named HEAPX to facilitate the automatic detection and exploitation of heap-based vulnerabilities. Specifically, a new path exploration strategy, a new memory model, and a new environment modeling solution are expected to be designed in HEAPX, so that the new boosted symbolic execution engine can detect heap-based vulnerabilities and generate working exploits for them more efficiently and effectively.