Poster - Program Reachability for Vulnerability and Malware Analysis
November 2020 • Poster
This project automates analysis of binary code, choosing inputs that trigger specific behavior and reduce the time spent performing complex software analysis.
Software Engineering Institute
Manually coercing specific portions of executable code to run presents a number of challenges, such as determining the unknown input conditions required to trigger the desired behavior, eliminating non-determinism, and coping with missing dependencies complicate this effort. CMU SEI researchers developed capabilities within the CMU SEI’s Pharos binary code analysis framework to address these challenges by identifying the specific program inputs and environments needed to reach an execution of interest to an analyst, which is called path finding. Finding paths in an executable can be especially useful for bypassing run-time anti-analysis checks in the code.