Towards Practical Binary Code Similarity Detection: Vulnerability Verification via Patch Semantic Analysis

Vulnerability is a major threat to software security. It has been proven that binary code similarity detection approaches are efficient search for recurring vulnerabilities introduced by sharing in software. However, these suffer from high false-positive rates since they usually take the patched functions as vulnerable, and do not work well when binaries compiled with different compilation settings. To this end, we propose an approach, named Robin , confirm filtering out functions. powered lightweight symbolic execution solve set of function inputs can lead vulnerability-related code. then executes target same capture vulnerable or behaviors filtration. Experimental results show achieves accuracy patch across compilers compiler optimization levels respectively on 287 real-world 10 Based accurate detection, significantly reduces rate state-of-the-art vulnerability tools (by 94.3% average), making them more practical. additionally detects 12 new potentially