ExpGen: A 2-Step Vulnerability Exploitability Evaluation Solution for Binary Programs under ASLR Environment

Current automatic exploit generation solutions generally adopt an 1-step exploit generation philosophy and neglect the potential difference between analysis-time environment and runtime environment. Therefore, they usually fail in evaluating exploitability for vulnerable programs running in an ASLR...

Full description

Bibliographic Details
Published in:Applied Sciences
Main Authors: Hui Huang, Yuliang Lu, Zulie Pan, Kailong Zhu, Lu Yu, Liqun Zhang
Format: Article
Language:English
Published: MDPI AG 2022-06-01
Subjects:
Online Access:https://www.mdpi.com/2076-3417/12/13/6593
Description
Summary:Current automatic exploit generation solutions generally adopt an 1-step exploit generation philosophy and neglect the potential difference between analysis-time environment and runtime environment. Therefore, they usually fail in evaluating exploitability for vulnerable programs running in an ASLR environment. We propose ExpGen, a 2-step vulnerability-exploitability evaluation solution for binary programs running in an ASLR environment, with three novel techniques introduced, separately partial-exploit sensitive-POC generation, exploitation context sensitive analysis-time exploit generation, and runtime exploit relocation. ExpGen firstly generates an analysis-time exploit that can carry out all the desired exploitation steps through applying the first two techniques in an iterative manner, then dynamically gaps the address-space layout differences between the analysis-time environment and runtime environment by adopting the runtime exploit-relocation technique, making the analysis-time exploit dynamically adaptable to the runtime exploitation session. Using a benchmark containing six test programs, 10 CTF&RHG programs and four real-world applications with known vulnerabilities, we demonstrate that ExpGen can effectively generate partial exploit input that carries out some address-leakage event and provide a complete automated exploitability evaluation workflow on vulnerable programs running in the ASLR environment.
ISSN:2076-3417