Dependency
Published:
Demystifying the Dependency Challenge in Kernel Fuzzing
Overview
This project systematically studies the dependency challenge in kernel fuzzing — where kernel code is locked behind specific kernel states that fuzzers struggle to reach. It provides measurement tools and analysis framework to quantify how dependencies hinder coverage, along with techniques to improve fuzzer effectiveness through better state exploration.
Key Features
Comprehensive measurement study of kernel dependency coverage gaps
Static analysis for unresolved dependency identification
State-aware fuzzing techniques for improved coverage
Evaluation framework for assessing fuzzer effectiveness under dependencies
Technologies
Go / C++ / C / Python
LLVM
Syzkaller
Protobuf / gRPC
🏆 Google Research Paper Rewards
| 📄 Paper | 🐙 GitHub |