Recent Papers Related To Fuzzing

This website is only used for collecting and grouping the related paper. If there are any paper need to be updated, you can contribute PR.
Please check the web https://wcventure.github.io/FuzzingPaper/, as the .md file shown in Github is cropped.
Advertisement: Our ICTT (Guangzhou) research group is accepting applications for master’s, doctoral, postdoctoral, and research assistant positions. We welcome hardworking, serious, and innovative young people who are interested in joining our research group.

Contributors

Starchart

All Papers (Classification according to Publication)

ISSTA 2024
- An Empirical Examination of Fuzzer Mutator Performance
- AsFuzzer: Differential Testing of Assemblers with Error-Driven Grammar Inference
- Atlas: Automating Cross-Language Fuzzing on Android Closed-Source Libraries
- Dance of the ADS: Orchestrating Failures through Historically-Informed Scenario Fuzzing
- DDGF: Dynamic Directed Greybox Fuzzing with Path Profiling
- Enhancing ROS System Fuzzing through Callback Tracing
- Fuzzing JavaScript Interpreters with Coverage-Guided Reinforcement Learning for LLM-Based Mutation
- Fuzzing MLIR Compiler Infrastructure via Operation Dependency Analysis
- How Effective Are They? Exploring Large Language Model Based Fuzz Driver Generation
- Logos: Log Guided Fuzzing for Protocol Implementations
- Midas: Mining Profitable Exploits in On-Chain Smart Contracts via Feedback-Driven Fuzzing and Differential Analysis
- Tacoma: Enhanced Browser Fuzzing with Fine-Grained Semantic Alignment
- Prospector: Boosting Directed Greybox Fuzzing for Large-Scale Target Sets with Iterative Prioritization
- FRIES: Fuzzing Rust Library Interactions via Efficient Ecosystem-Guided Target Generation
- Logos: Log Guided Fuzzing for Protocol Implementations
- Policy Testing with MDPFuzz (Replicability Study)
- Enhancing ROS System Fuzzing through Callback Tracing
- Sleuth: A Switchable Dual-Mode Fuzzer to Investigate Bug Impacts Following a Single PoC
ESEC/FSE 2024
- BRF: Fuzzing the eBPF runtime
- Evaluating Directed Fuzzers: Are We Heading in the Right Direction?
SP 2024
- Everything is Good for Something: Counterexample-Guided Directed Fuzzing via Likely Invariant Inference
- LABRADOR: Response Guided Directed Fuzzing for Black-box IoT Devices
- SoK: Prudent Evaluation Practices for Fuzzing
- Titan: Efficient Multi-target Directed Greybox Fuzzing
- Chronos: Finding Timeout Bugs in Practical Distributed Systems by Deep-Priority Fuzzing with Transient Delay
- DY Fuzzing: Formal Dolev-Yao Models Meet Cryptographic Protocol Fuzz Testing
- SyzGen++: Dependency Inference for Augmenting Kernel Driver Fuzzing
- Towards Smart Contract Fuzzing on GPU
- LLMIF: Augmented Large Language Model for Fuzzing IoT Devices
- SATURN: Host-Gadget Synergistic USB Driver Fuzzing
- To Boldly Go Where No Fuzzer Has Gone Before: Finding Bugs in Linux' Wireless Stacks through VirtIO Devices
- Predecessor-aware Directed Greybox Fuzzing
- AFGen: Whole-Function Fuzzing for Applications and Libraries
- SyzTrust: State-aware Fuzzing on Trusted OS Designed for IoT Devices
- LLMIF: Augmented Large Language Model for Fuzzing IoT Devices
ICISSP 2024
- Fuzzing Matter(s): A White Paper for Fuzzing the Matter Protocol
ICSE 2024
- Fuzz4All: Universal Fuzzing with Large Language Models
- Large Language Models are Edge-Case Generators: Crafting Unusual Programs for Fuzzing Deep Learning Libraries
- ECFuzz: Effective Configuration Fuzzing for Large-Scale Systems
- EDEFuzz: A Web API Fuzzer for Excessive Data Exposures
- FuzzSlice: Pruning False Positives in Static Analysis Warnings through Function-Level Fuzzing
- Learning and Repair of Deep Reinforcement Learning Policies from Fuzz-Testing Data
- Reorder Pointer Flow in Sound Concurrency Bug Prediction
- Are We There Yet? Unraveling the State-of-the-Art Smart Contract Fuzzers
- Crossover in Parametric Fuzzing
- Extrapolating Coverage Rate in Greybox Fuzzing
- FuzzInMem: Fuzzing Programs via In-memory Structures
- RPG: Rust Library Fuzzing with Pool-based Fuzz Target Generation and Generic Support
- Sedar: Obtaining High-Quality Seeds for DBMS Fuzzing via Cross-DBMS SQL Transfer
- SpecBCFuzz: Fuzzing LTL Solvers with Boundary Conditions
NDSS 2024
- Large Language Model guided Protocol Fuzzing
- DeepGo: Predictive Directed Greybox Fuzzing
- ShapFuzz: Efficient Fuzzing via Shapley-Guided Byte Selection
- EnclaveFuzz: Finding Vulnerabilities in SGX Applications
- MOCK: Optimizing Kernel Fuzzing Mutation with Context-aware Dependency
- EnclaveFuzz: Finding Vulnerabilities in SGX Applications
- Predictive Context-sensitive Fuzzing
- ReqsMiner: Automated Discovery of CDN Forwarding Request Inconsistencies and DoS Attacks with Grammar-based Fuzzing
USENIX Security 2024
- Atropos: Effective Fuzzing of Web Applications for Server-Side Vulnerabilities
- ResolverFuzz: Automated Discovery of DNS Resolver Vulnerabilities with Query-Response Fuzzing
- Cascade: CPU Fuzzing via Intricate Program Generation
- Towards Generic Database Management System Fuzzing
- SHiFT: Semi-hosted Fuzz Testing for Embedded Applications
- SDFuzz: Target States Driven Directed Fuzzing
- Critical Code Guided Directed Greybox Fuzzing for Commits
- EL3XIR: Fuzzing COTS Secure Monitors
- MultiFuzz: A Multi-Stream Fuzzer For Testing Monolithic Firmware
- WhisperFuzz: White-Box Fuzzing for Detecting and Locating Timing Vulnerabilities in Processors
- HYPERPILL: Fuzzing for Hypervisor-bugs by leveraging the Hardware Virtualization Interface
- Fuzzing BusyBox: Leveraging LLM and Crash Reuse for Embedded Bug Unearthing
- From One Thousand Pages of Specification to Unveiling Hidden Bugs: Large Language Model Assisted Fuzzing of Matter IoT Devices
ASPLOS 2024
- Greybox Fuzzing for Concurrency Testing
OOPSLA 2024
- WhiteFox: White-box Compiler Fuzzing Empowered by Large Language Models
Others 2024
- Crash Report Accumulation During Continuous Fuzzing with CASR (IVMEM 2024)
- CrossFuzz: Cross-contract fuzzing for smart contract vulnerability detection
MobiSys 2023
- No More Companion Apps Hacking but One Dongle: Hub-Based Blackbox Fuzzing of IoT Firmware
ESORICS 2023
- ResolFuzz: Differential Fuzzing of DNS Resolvers
- Intelligent Zigbee Protocol Fuzzing via Constraint-Field Dependency Inference
- LFuzz: Exploiting Locality-Enabled Techniques for File-System Fuzzing
- SPLITS: Split Input-to-State Mapping for Effective Firmware Fuzzing
- Forkfuzz: Leveraging the Fork-Awareness in Coverage-Guided Fuzzing
OOPSLA 2023
- Accelerating Fuzzing through Prefix-Guided Execution
- Towards Better Semantics Exploration for Browser Fuzzing
PLDI 2023
- Fuzzing Loop Optimizations in Compilers for C++ and Data-Parallel Languages
- WasmRef-Isabelle: a Verified Monadic Interpreter and Industrial Fuzzing Oracle for WebAssembly
S&P 2023
- SelectFuzz: Efficient Directed Fuzzing with Selective Path Exploration
- UTOPIA: Automatic Generation of Fuzz Driver using Unit Tests
- TEEzz: Fuzzing Trusted Applications on COTS Android Devices
- Toss a Fault to Your Witcher: Applying Grey-box Coverage-Guided Mutational Fuzzing to Detect SQL and Command Injection Vulnerabilities
- Finding Specification Blind Spots via Fuzz Testing
- SEGFUZZ: Segmentizing Thread Interleaving to Discover Kernel Concurrency Bugs through Fuzzing
- RSFuzzer: Discovering Deep SMI Handler Vulnerabilities in UEFI Firmware with Hybrid Fuzzing
- ODDFUZZ: Discovering Java Deserialization Vulnerabilities via Structure-Aware Directed Greybox Fuzzing
- VIDEZZO: Dependency-aware Virtual Device Fuzzing
- DEVFUZZ: Automatic Device Model-Guided Device Driver Fuzzing
CCS 2023
- HOPPER: Interpretative Fuzzing for Libraries
- Greybox Fuzzing of Distributed Systems
- Fuzz on the Beach: Fuzzing Solana Smart Contracts
- DSFuzz: Detecting Deep State Bugs with Dependent State Exploration
- Profile-guided System Optimizations for Accelerated Greybox Fuzzing
- NestFuzz: Enhancing Fuzzing with Comprehensive Understanding of Input Processing Logic
- Greybox Fuzzing of Distributed Systems
- SyzDirect: Directed Greybox Fuzzing for Linux Kernel
- PyRTFuzz: Detecting Bugs in Python Runtimes via Two-Level Collaborative Fuzzing
- Poster: Combining Fuzzing with Concolic Execution for IoT Firmware Testing
USENIX SEC 2023
- FishFuzz: Catch Deeper Bugs by Throwing Larger Nets
- MINER: A Hybrid Data-Driven Approach for REST API Fuzzing
- KextFuzz: Fuzzing macOS Kernel EXTensions on Apple Silicon via Exploiting Mitigations
- MorFuzz: Fuzzing Processor via Runtime Instruction Morphing enhanced Synchronizable Co-simulation
- CarpetFuzz: Automatic Program Option Constraint Extraction from Documentation for Fuzzing
- PolyFuzz: Holistic Greybox Fuzzing of Multi-Language Systems
- GLeeFuzz: Fuzzing WebGL Through Error Message Guided Mutation
- autofz: Automated Fuzzer Composition at Runtime
- Fuzztruction: Using Fault Injection-based Fuzzing to Leverage Implicit Domain Knowledge
- DynSQL: Stateful Fuzzing for Database Management Systems with Complex and Valid SQL Query Generation
- FuzzJIT: Oracle-Enhanced Fuzzing for JavaScript Engine JIT Compiler
- AIFORE: Smart Fuzzing Based on Automatic Input Format Reverse Engineering
- Automated Exploitable Heap Layout Generation for Heap Overflows Through Manipulation Distance-Guided Fuzzing
- Bleem: Packet Sequence Oriented Fuzzing for Protocol Implementations
- BoKASAN: Binary-only Kernel Address Sanitizer for Effective Kernel Fuzzing
- DDRace: Finding Concurrency UAF Vulnerabilities in Linux Drivers with Directed Fuzzing
- Forming Faster Firmware Fuzzers
- Intender: Fuzzing Intent-Based Networking with Intent-State Transition Guidance\
- MTSan: A Feasible and Practical Memory Sanitizer for Fuzzing COTS Binaries
ISSTA 2023
- Large Language Models are Zero-Shot Fuzzers: Fuzzing Deep-Learning Libraries via Large Language Models
- GrayC: Greybox Fuzzing of Compilers and Analysers for C
- 1dFuzz: Reproduce 1-Day Vulnerabilities with Directed Differential Fuzzing
- Automated Program Repair from Fuzzing Perspective
- Detecting State Inconsistency Bugs in DApps via On-Chain Transaction Replay and Fuzzing
- Fuzzing Deep Learning Compilers with HirGen
- Fuzzing Embedded Systems using Debug Interfaces
- Green Fuzzer Benchmarking
- Green Fuzzing: A Saturation-Based Stopping Criterion using Vulnerability Prediction
- Guiding Greybox Fuzzing with Mutation Testing
- Icicle: A Re-designed Emulator for Grey-Box Firmware Fuzzing
- ItyFuzz: Snapshot-Based Fuzzer for Smart Contract
- Rare Path Guided Fuzzing
ICSE 2023
- Evaluating and Improving Hybrid Fuzzing
- Fuzzing Automatic Differentiation in Deep-Learning Libraries
- JITfuzz: Coverage-guided Fuzzing for JVM Just-in-Time Compilers
- Learning Seed-Adaptive Mutation Strategies for Greybox Fuzzing
- Operand-Variation-Oriented Differential Analysis for Fuzzing Binding Calls in PDF Readers
- Regression Fuzzing for Deep Learning Systems
- Large Language Models are Few-shot Testers: Exploring LLM-based General Bug Reproduction
- Validating SMT Solvers via Skeleton Enumeration Empowered by Historical Bug-Triggering Inputs
NDSS 2023
- Assessing the Impact of Interface Vulnerabilities in Compartmentalized Software
- FUZZILLI: Fuzzing for JavaScript JIT Compiler Vulnerabilities
- No Grammar, No Problem: Towards Fuzzing the Linux Kernel without System-Call Descriptions
- DARWIN: Survival of the Fittest Fuzzing Mutators
- LOKI: State-Aware Fuzzing Framework for the Implementation of Blockchain Consensus Protocols
- OBSan: An Out-Of-Bound Sanitizer to Harden DNN Executables
Euro S&P 2023 / Euro S&P Workshops 2023
- EF/CF: High Performance Smart Contract Fuzzing for Exploit Generation
- Fuzzing SGX Enclaves via Host Program Mutations
- The Bandit’s States: Modeling State Selection for Stateful Network Fuzzing as Multi-armed Bandit Problem
ICISSP/ICPRAM 2023
- Rainfuzz: Reinforcement-Learning Driven Heat-Maps for Boosting Coverage-Guided Fuzzing
- SWaTEval: An Evaluation Framework for Stateful Web Application Testing
- Evaluating the Fork-Awareness of Coverage-Guided Fuzzers
Others 2023
- CIDFuzz: Fuzz testing for continuous integration
OOPSLA 2022
- Satisfiability Modulo Fuzzing: A Synergistic Combination of SMT Solving and Fuzzing
ASE 2022
- Effectively Generating Vulnerable Transaction Sequences in Smart Contracts with Reinforcement Learning-guided Fuzzing
- Efficient Greybox Fuzzing to Detect Memory Errors
- FuzzerAid: Grouping Fuzzed Crashes Based On Fault Signatures
- Fuzzle: Making a Puzzle for Fuzzers
- Griffin: Grammar-Free DBMS Fuzzing
- HTFuzz: Heap Operation Sequence Sensitive Fuzzing
- LawBreaker: An Approach for Specifying Traffic Laws and Fuzzing Autonomous Vehicles
- QATest: A Uniform Fuzzing Framework for Question Answering Systems
- So Many Fuzzers, So Little Time - Experience from Evaluating Fuzzers on the Contiki-NG Network (Hay)StackVirt
- Auto Off-Target: Enabling Thorough and Scalable Testing for Complex Software Systems
ISSTA 2022
- sAlmost Correct Invariants: Synthesizing Inductive Invariants by Fuzzing Proof
- DocTer: Documentation-Guided Fuzzing for Testing Deep Learning API Functions
- Efficient Greybox Fuzzing of Applications in Linux-based IoT Devices via Enhanced User-mode Emulation
- MDPFuzz: Testing Models Solving Markov Decision Processes
- PrIntFuzz: Fuzzing Linux Drivers via Automated Virtual Device Simulation
- SLIME: Program-sensitive Energy Allocation for Fuzzing
- SnapFuzz: High-Throughput Fuzzing of Network Applications
- TensileFuzz: Facilitating Seed Input Generation in Fuzzing via String Constraint Solving
- WASAI: Uncovering Vulnerabilities in Wasm Smart Contracts
- Unicorn: Detect Runtime Error in Time-Series Databases With Hybrid Input Synthesis
ESEC/FSE 2022
- Minerva: Browser API Fuzzing with Dynamic Mod-Ref Analysis
- RoboFuzz: Fuzzing Robotic Systems over Robot Operating System (ROS) for Finding Correctness Bugs
- SEDiff: Scope-Aware Differential Fuzzing to Test Internal Function Models in Symbolic Execution
- Fuzzing Deep-Learning Libraries via Automated Relational API Inference
CCS 2022
- DriveFuzz: Discovering Autonomous Driving Bugs through Driving Quality-Guided Fuzzing
- JIT-Picking: Differential Fuzzing of JavaScript Engines
- LibAFL: A Framework to Build Modular and Reusable Fuzzers
- MC^2: Rigorous and Efficient Directed Greybox Fuzzing
- SFuzz: Slice-based Fuzzing for Real-Time Operating Systems
- SpecDoctor: Differential Fuzz Testing to Find Transient Execution Vulnerabilities
S&P 2022
- Effective Seed Scheduling for Fuzzing with Graph Centrality Analysis
- BEACON: Directed Grey-Box Fuzzing with Provable Path Pruning
- JIGSAW: Efficient and Scalable Path Constraints Fuzzing
- PATA: Fuzzing with Path Aware Taint Analysis
- FuzzUSB: Hybrid Stateful Fuzzing of USB Gadget Stacks
NDSS 2022
- Cooper: Testing the Binding Code of Scripting Languages with Cooperative Mutation
- EMS: History-Driven Mutation for Coverage-based Fuzzing
- Semantic-Informed Driver Fuzzing Without Both the Hardware Devices and the Emulators
- MobFuzz: Adaptive Multi-objective Optimization in Gray-box Fuzzing
- Context-Sensitive and Directional Concurrency Fuzzing for Data-Race Detection
- datAFLow: Towards a Data-Flow-Guided Fuzzer
USENIX SEC 2022
- SyzScope: Revealing High-Risk Security Impacts of Fuzzer-Exposed Bugs in Linux kernel
- MundoFuzz: Hypervisor Fuzzing with Statistical Coverage Testing and Grammar Inference
- TheHuzz: Instruction Fuzzing of Processors Using Golden-Reference Models for Finding Software-Exploitable Vulnerabilities
- Morphuzz: Bending (Input) Space to Fuzz Virtual Devices
- Fuzzware: Using Precise MMIO Modeling for Effective Firmware Fuzzing
- FuzzOrigin: Detecting UXSS vulnerabilities in Browsers through Origin Fuzzing
- Drifuzz: Harvesting Bugs in Device Drivers from Golden Seeds
- Fuzzing Hardware Like Software
- BrakTooth: Causing Havoc on Bluetooth Link Manager via Directed Fuzzing
- Stateful Greybox Fuzzing
- AmpFuzz: Fuzzing for Amplification DDoS Vulnerabilities
- SGXFuzz: Efficiently Synthesizing Nested Structures for SGX Enclave Fuzzing
- FRAMESHIFTER: Manipulating HTTP/2 Frame Sequences with Fuzzing
- FIXREVERTER: A Realistic Bug Injection Methodology for Benchmarking Fuzz Testing
- StateFuzz: System Call-Based State-Aware Linux Driver Fuzzing
ICSE 2022
- μAFL: Non-intrusive Feedback-driven Fuzzing for Microcontroller Firmware
- BeDivFuzz: Integrating Behavioral Diversity into Generator-based Fuzzing
- CONFETTI: Amplifying Concolic Guidance for Fuzzers
- Demystifying the Dependency Challenge in Kernel Fuzzing
- Evaluating and Improving Neural Program-Smoothing-based Fuzzing
- Fuzzing Class Specifications
- GraphFuzz: Library API Fuzzing with Lifetime-aware Dataflow Graphs
- Linear-time Temporal Logic guided Greybox Fuzzing
- Muffin: Testing Deep Learning Libraries via Neural Architecture Fuzzing
- One Fuzzing Strategy to Rule Them All
- On the Reliability of Coverage-Based Fuzzer Benchmarking
- Path Transitions Tell More: Optimizing Fuzzing Schedules via Runtime Program States
- R2Z2: Detecting Rendering Regressions in Web Browsers through Differential Fuzz Testing
- Semantic Image Fuzzing of AI Perception Systems
- Free Lunch for Testing: Fuzzing Deep-Learning Libraries from Open Source
- WindRanger: A Directed Greybox Fuzzer driven by Deviation Basic Block
- MOREST: Model-based RESTful API Testing with Execution Feedback
- Controlled Concurrency Testing via Periodical Scheduling
- Combinatorial Testing of RESTful APIs
- Automated Testing of Software that Uses Machine Learning APIs
- FADATest: Fast and Adaptive Performance Regression Testing of Dynamic Binary Translation Systems
- Nessie: Automatically Testing JavaScript APIs with Asynchronous Callbacks
SANER 2022
- FuzzingDriver: the Missing Dictionary to Increase Code Coverage in Fuzzers
Others 2022
- UltraFuzz: Towards Resource-saving in Distributed Fuzzing (TSE 2022)
- Fuzzing with automatically controlled interleavings to detect concurrency bugs (The Journal of Systems & Software)
- ROZZ: Property-based Fuzzing for Robotic Programs in ROS (ICRA 2022)
- Efficient ECU Analysis Technology through Structure-aware CAN Fuzzing (Access 2022)
- Smart Contract Fuzzing for Enterprises: The Language Agnostic Way (COMSNETS 2022)
- spotFuzzer: Static Instrument and Fuzzing Windows COTs (2022)
- Mutation Analysis: Answering the Fuzzing Challenge (2022)
- SNPSFuzzer: A Fast Greybox Fuzzer for Stateful Network Protocols using Snapshots (2022)
- Better Pay Attention Whilst Fuzzing (2022)
- looking for lacunae in bitcoin cores fuzzing efforts
- FEFuzzer: Hybrid Files Fuzzing Tool (2022)
- Sydr-Fuzz: Continuous Hybrid Fuzzing and Dynamic Analysis for Security Development Lifecycle (ISPRAS Open 2022)
- Improving HPC Security with Targeted Syscall Fuzzing (S-HPC 2022)
ASE 2021
- CorbFuzz: Checking Browser Security Policies with Fuzzing
- Scalable Fuzzing of Program Binaries with E9AFL
- SMARTIAN : Enhancing Smart Contract Fuzzing with Static and Dynamic Data-Flow Analyses
- Towards Systematic and Dynamic Task Allocation for Collaborative Parallel Fuzzing
- Improving Configurability of Unit-level Continuous Fuzzing: An Industrial Case Study with SAP HANA
- Understanding and Detecting Performance Bugs in Markdown Compilers
- Rust Library Fuzzing via API Dependency Graph Traversal
- InstruGuard: Find and Fix Instrumentation Errors for Coverage-based Greybox Fuzzing
ACSAC 2021
- Westworld: Fuzzing-Assisted Remote Dynamic Symbolic Execution of Smart Apps on IoT Cloud Platforms
- ICS3Fuzzer: A Framework for Discovering Protocol Implementation Bugs in ICS Supervisory Software by Fuzzing
SOSP 2021
- HEALER: Relation Learning Guided Kernel Fuzzing
S&P 2021
- DiFuzzRTL: Differential Fuzz Testing to Find CPU Bugs
- STOCHFUZZ: Sound and Cost-effective Fuzzing of Stripped Binaries by Incremental and Stochastic Rewriting
- NTFUZZ: Enabling Type-Aware Kernel Fuzzing on Windows with Static Binary Analysis
- DIANE: Identifying Fuzzing Triggers in Apps to Generate Under-constrained Inputs for IoT Devices
- One Engine to Fuzz 'em All: Generic Language Processor Testing with Semantic Validation
ISSTA 2021
- Fuzzing SMT Solvers via Two-Dimensional Input Space Exploration
- Gramatron: Effective Grammar-Aware Fuzzing
- QFuzz: Quantitative Fuzzing for Side Channels
- Seed Selection for Successful Fuzzing
- ProFuzzBench - A Benchmark for Stateful Protocol Fuzzing
ESEC/FSE 2021
- Skeletal Approximation Enumeration for SMT Solver Testing
- Estimating Residual Risk in Greybox Fuzzing
- HeteroFuzz: Fuzz Testing to Detect Platform Dependent Divergence for Heterogeneous Applications
- FuzzBench: An Open Fuzzer Benchmarking Platform and Service
CCS 2021
- Snipuzz: Black-box Fuzzing of IoT Firmware via Message Snippet Inference
- Hardware Support to Improve Fuzzing Performance and Precision
- SoFi: Reflection-Augmented Fuzzing for JavaScript Engines
- T-Reqs: HTTP Request Smuggling with Differential Fuzzing
- V-SHUTTLE: Scalable and Semantics-Aware Hypervisor Fuzzing
- Same Coverage, Less Bloat: Accelerating Binary-only Fuzzing with Coverage-preserving Coverage-guided Tracing
- Regression Greybox Fuzzing
- SyzGen: Automated Generation of Syscall Specification of Closed-Source macOS Drivers
OOPSLA 2021
- Fuzzing Channel-Based Concurrency Runtimes using Types and Effects
- Fully Automated Functional Fuzzing of Android Apps for Detecting Non-crashing Logic Bugs
PLDI 2021
- Automated Conformance Testing for JavaScript Engines via Deep Compiler Fuzzing
ICSE 2021
- Fuzzing Symbolic Expressions
- Graph-based Fuzz Testing for Deep Learning Inference Engines
- Growing A Test Corpus with Bonsai Fuzzing
- Industry Practice of Coverage-Guided Enterprise-Level DBMS Fuzzing
- IntelliGen: Automatic Driver Synthesis for Fuzz Testing
USENIX ATC 2021
- RIFF: Reduced Instruction Footprint for Coverage-Guided Fuzzing
- TCP-Fuzz: Detecting Memory and Semantic Bugs in TCP Stacks with Fuzzing
USENIX Security 2021
- Constraint-guided Directed Greybox Fuzzing
- UNIFUZZ: A Holistic and Pragmatic Metrics-Driven Platform for Evaluating Fuzzers
- Nyx: Greybox Hypervisor Fuzzing using Fast Snapshots and Affine Types
- Breaking Through Binaries: Compiler-quality Instrumentation for Better Binary-only Fuzzing
- The Use of Likely Invariants as Feedback for Fuzzers
- ICSFuzz: Manipulating I/Os and Repurposing Binary Code to Enable Instrumented Fuzzing in ICS Control Applications
- Android SmartTVs Vulnerability Discovery via Log-Guided Fuzzing
- APICraft: Fuzz Driver Generation for Closed-source SDK Libraries
- SyzVegas: Beating Kernel Fuzzing Odds with Reinforcement Learning
- Token-Level Fuzzing
NDSS 2021
- Favocado: Fuzzing the Binding Code of JavaScript Engines Using Semantically Correct Test Cases
- WINNIE: Fuzzing Windows Applications with Harness Synthesis and Fast Cloning
- PGFUZZ: Policy-Guided Fuzzing for Robotic Vehicles
- Reinforcement Learning-based Hierarchical Seed Scheduling for Greybox Fuzzing
ICST 2021
- Industrial Oriented Evaluation of Fuzzing Techniques
- Learning-Based Fuzzing of IoT Message Broker
- RiverFuzzRL - an open-source tool to experiment with reinforcement learning for fuzzing
SANER 2021
- OTA: An Operation-oriented Time Allocation Strategy for Greybox Fuzzing
JFLA 2022
- Alt-Ergo-Fuzz: A fuzzer for the Alt-Ergo SMT solver
Others 2021
- FuSeBMC v. 4: Smart Seed Generation for Hybrid Fuzzing (2021)
- REST API Fuzzing by Coverage Level Guided Blackbox Testing (2021)
- State Selection Algorithms and Their Impact on The Performance of Stateful Network Protocol Fuzzing (2021)
- DIAR: Removing Uninteresting Bytes from Seeds in Software Fuzzing (2021)
- CatchBackdoor: Backdoor Testing by Critical Trojan Neural Path Identification via Differential Fuzzing (2021)
- Emmutaler: Fuzzing the iOS Boot Loader (Thesis 2021)
- Concolic-Fuzzing of JavaScript Programs using GraalVM and Truffle (SKILL 2021)
- BanditFuzz: Fuzzing SMT Solvers with Multi-agent Reinforcement Learning (FM 2021)
- My Fuzzer Beats Them All! Developing a Framework for Fair Evaluation and Comparison of Fuzzers (2021)
- Evaluating Code Coverage for Kernel Fuzzers via Function Call Graph (Access 2021)
- A Tight Integration of Symbolic Execution and Fuzzing (short paper 2021)
- DeltaFuzz: Historical Version Information Guided Fuzz Testing (Journal of Computer Science and Technology 2021)
- ovAFLow: Detecting Memory Corruption Bugs with Fuzzing-based Taint Inference (Journal of Computer Science and Technology 2021)
- Machine Learning Guided Cross-Contract Fuzzing (2021)
- SANRAZOR: Reducing Redundant Sanitizer Checks in C/C++ Programs (OSDI 2021)
- webFuzz: Grey-Box Fuzzing for Web Applications (ESORICS 2021)
- Vulnerability-Oriented Fuzz Testing for Connected Autonomous Vehicle Systems (IEEE Transactions on Reliability 2021)
- Duo: Differential Fuzzing for Deep Learning Operators (IEEE Transactions on Reliability 2021)
- Rtkaller: State-aware Task Generation for RTOS Fuzzing (EMSOFT 2021)
- Neural Network Guided Evolutionary Fuzzing for Finding Traffic Violations of Autonomous Vehicles (arXiv 2021)
- Fuzzing: Cyberphysical System Testing for Security and Dependability (Computer 2021)
- ESRFuzzer: an enhanced fuzzing framework for physical SOHO router devices to discover multi-Type vulnerabilities (Cybersecurity 2021)
- ICPFuzzer: proprietary communication protocol fuzzing by using machine learning and feedback strategies (Cybersecurity 2021)
- Hashing Fuzzing: Introducing Input Diversity to Improve Crash Detection (TSE 2021)
- KCFuzz: Directed Fuzzing Based on Keypoint Coverage (ICAIS 2021)
- Token-Level Fuzzing (WiSec'21)
- SHFuzz: A hybrid fuzzing method assisted by static analysis for binary programs (China Communications 2021)
- RapidFuzz: Accelerating fuzzing via Generative Adversarial Networks (Neurocomputing 2021)
- HFContractFuzzer: Fuzzing Hyperledger Fabric Smart Contracts for Vulnerability Detection (EASE 2021)
- Fuzzing With Optimized Grammar-Aware Mutation Strategies (Access 2021)
- FIRM-COV: High-Coverage Greybox Fuzzing for IoT Firmware via Optimized Process Emulation (Access 2021)
- Extended grammar-based fuzzing algorithm for JavaScript Engines (2021)
- CoCoFuzzing: Testing Neural Code Models with Coverage-Guided Fuzzing (2021)
- FUZZOLIC: Mixing fuzzing and concolic execution (Computers&Security 2021)
- DirectFuzz: Automated Test Generation for RTL Designs using Directed Graybox Fuzzing (DAC 2021)
- CollabFuzz: A Framework for Collaborative Fuzzing (EuroSec 2021)
- HDR-Fuzz: Detecting Buffer Overruns using AddressSanitizer Instrumentation and Fuzzing (2021)
- Unleashing Fuzzing Through Comprehensive, Efficient, and Faithful Exploitable-Bug Exposing (2021)
- Improving Web Application Vulnerability Detection Leveraging Ensemble Fuzzing (ENASE 2021)
- Coverage-guided binary fuzzing with rev.ng and llvm libfuzzer
- An Empirical Study of OSS-Fuzz Bugs (MSR 2021)
- MooFuzz: Many-Objective Optimization Seed Schedule for Fuzzer (Mathematics 2021)
- PMFuzz: Test Case Generation for Persistent Memory Programs (ASPLOS 2021)
- A Priority Based Path Searching Method for Improving Hybrid Fuzzing (Computers & Security 2021)
- CMFuzz: context-aware adaptive mutation for fuzzers (Empirical Software Engineering 2021)
- Refined Grey-Box Fuzzing with Sivo (arXiv 2021)
- Constructing More Complete Control Flow Graphs Utilizing Directed Gray-Box Fuzzing (MDPI 2021)
- Symbolic Security Predicates: Hunt Program Weaknesses (ISPRAS Open 2021)
- Towards Symbolic Pointers Reasoning in Dynamic Symbolic Execution (IVMEM 2021)
CCS 2020
- FREEDOM: Engineering a State-of-the-Art DOM Fuzzer
- SQUIRREL: Testing Database Management Systems with Language Validity and Coverage Feedback
ASE 2020
- BigFuzz: Efficient Fuzz Testing for Data Analytics using Framework Abstraction
- MoFuzz: A Fuzzer Suite for Testing Model-Driven Software Engineering Tools
- Zeror: Speed Up Fuzzing with Coverage-sensitive Tracing and Scheduling
- Generating Highly-structured Input Data by Combining Search-based Testing and Grammar-based Fuzzing
ISSTA 2020
- Active Fuzzing for Testing and Securing Cyber-Physical Systems
- Learning Input Tokens for Effective Fuzzing
- WEIZZ: Automatic Grey-Box Fuzzing for Structured Binary Formats
ESEC/FSE 2020
- Fuzzing: On the Exponential Cost of Vulnerability Discovery
- Boosting Fuzzer Efficiency: An Information Theoretic Perspective
- CrFuzz: Fuzzing Multi-purpose Programs through Input Validation
- Detecting Critical Bugs in SMT Solvers using Blackbox Mutational Fuzzing
- Intelligent REST API Data Fuzzing
- MTFuzz: Fuzzing with a Multi-task Neural Network
ACSAC 2020
- DPIFuzz: A Differential Fuzzing Framework to Detect DPI Elusion Strategies for QUIC
- Cupid: Automatic Fuzzer Selection for Collaborative Fuzzing
- OmniFuzz: A Flexible Framework for Expediting Bug Finding by Leveraging Past (Mis-)Behavior to Discover New Bugs
- RUSTY: A Fuzzing Tool for Rust
PLDI 2020
- Validating SMT Solvers via Semantic Fusion
USENIX Security 2020
- Analysis of DTLS Implementations Using Protocol State Fuzzing
- Frankenstein: Advanced Wireless Fuzzing to Exploit New Bluetooth Escalation Targets
- SpecFuzz: Bringing Spectre-type vulnerabilities to the surface
- FuzzGen: Automatic Fuzzer Generation
- MUZZ: Thread-aware Grey-box Fuzzing for Effective Bug Hunting in Multithreaded Programs
- Montage: A Neural Network Language Model-Guided JavaScript Engine Fuzzer
- GREYONE: Data Flow Sensitive Fuzzing
- FuzzGuard: Filtering out Unreachable Inputs in Directed Grey-box Fuzzing through Deep Learning
- ParmeSan: Sanitizer-guided Greybox Fuzzing
- EcoFuzz: Adaptive Energy-Saving Greybox Fuzzing as a Variant of the Adversarial Multi-Armed Bandit
- FANS: Fuzzing Android Native System Services via Automated Interface Analysis
- Fuzzing Error Handling Code using Context-Sensitive Software Fault Injection
- USBFuzz: A Framework for Fuzzing USB Drivers by Device Emulation
- AFL++: Combining Incremental Steps of Fuzzing Research (USENIX Woot2020)
ICSE 2020
- Typestate-Guided Fuzzer for Discovering Use-after-Free Vulnerabilities
- MemLock: Memory Usage Guided Fuzzing
- Ankou: Guiding Grey-box Fuzzing towards Combinatorial Difference
- JVM Fuzzing for JIT-Induced Side-Channel Detection
- Targeted Greybox Fuzzing with Static Lookahead Analysis
- Fuzz Testing based Data Augmentation to Improve Robustness of Deep Neural Networks
- sFuzz: An Efficient Adaptive Fuzzer for Solidity Smart Contracts
- HyDiff: Hybrid Differential Software Analysis
- Automatically Testing String Solvers
NDSS 2020
- HYPER-CUBE: High-Dimensional Hypervisor Fuzzing
- HotFuzz: Discovering Algorithmic Denial-of-Service Vulnerabilities Through Guided Micro-Fuzzing
- HFL: Hybrid Fuzzing on the Linux Kernel
- Not All Coverage Measurements Are Equal: Fuzzing by Coverage Accounting for Input Prioritization
S&P 2020
- SAVIOR: Towards Bug-Driven Hybrid Testing
- RetroWrite: Statically Instrumenting COTS Binaries for Fuzzing and Sanitization
- IJON: Exploring Deep State Spaces via Fuzzing
- PANGOLIN: Incremental Hybrid Fuzzing with Polyhedral Path Abstraction
- KRace: Data Race Fuzzing for Kernel File Systems
- Fuzzing JavaScript Engines with Aspect-preserving Mutation
SANER 2020
- ETHPLOIT: From Fuzzing to Efficient Exploit Generation against Smart Contracts
- Sequence directed hybrid fuzzing
ICST 2020
- Language-Agnostic Generation of Compilable Test Programs
- ct-fuzz: Fuzzing for Timing Leaks
- AFLNET: A Greybox Fuzzer for Network Protocols
ASIACCS 2020
- PathAFL: Path-Coverage Assisted Fuzzing
Others 2020
- Reproducible Crashes: Fuzzing Pharo by Mutating the Test Methods (IWST20 2020)
- Finding race conditions in Kernels: from fuzzing to symbolic exection (2020)
- BanditFuzz: Fuzzing SMT Solvers with Reinforcement Learning (2020)
- Fuzzing Technique in Web Applications and Beyond (MCTE 2020)
- Vulnerability Detection in SIoT Applications: A Fuzzing Method on their Binaries (IEEE Transactions on Network Science and Engineering 2020)
- MaxAFL: Maximizing Code Coverage with a Gradient-Based Optimization Technique (Electronics 2020)
- CSEFuzz: Fuzz Testing Based on Symbolic Execution (Access 2020)
- A Quantitative Comparison of Covera (AST 2020)
- Finding Bugs in File Systems with an Extensible Fuzzing Framework (TOS 2020)
- ICS Protocol Fuzzing: Coverage Guided Packet Crack and Generation (DAC 2020)
- Finding Security Vulnerabilities in Network Protocol Implementations (Arxiv 2020)
- Coverage Guided Differential Adversarial Testing of Deep Learning Systems
- Fw-fuzz: A code coverage-guided fuzzing framework for network protocols on firmware
- Greybox Fuzzing Based on Ant Colony Algorithm
- MEUZZ: Smart Seed Scheduling for Hybrid Fuzzing
- Binary-level Directed Fuzzing for Use-After-Free Vulnerabilities
- Smart seed selection-based effective black box fuzzing for IIoT protocol
- RDFuzz: Accelerating Directed Fuzzing with Intertwined Schedule and Optimized Mutation
- A deep convolution generative adversarial networks based fuzzing framework for industry control protocols
- TOFU: Target-Oriented FUzzer
- BaseSAFE: Baseband SAnitized Fuzzing through Emulation
- Sydr: Cutting Edge Dynamic Symbolic Execution (ISPRAS Open 2020)
ACSAC 2019
- Opening Pandoras Box through ATFuzzer: Dynamic Analysis of AT Interface for AndroidSmartphones
OOPSLA 2019
- FuzzFactory: Domain-Specific Fuzzing with Waypoints
- Compiler Fuzzing: How Much Does It Matter
TSE 2019
- The Art, Science, and Engineering of Fuzzing: A Survey
- Smart Greybox Fuzzing
Access 2019
- NeuFuzz: Efficient Fuzzing With Deep Neural Network
- LearnAFL: Greybox Fuzzing With Knowledge Enhancement
CCS 2019
- Intriguer: Field-Level Constraint Solving for Hybrid Fuzzing
- Learning to Fuzz from Symbolic Execution with Application to Smart Contracts
- Matryoshka: fuzzing deeply nested branches
- Different is Good: Detecting the Use of Uninitialized Variables through Differential Replay
- Gollum: Modular and Greybox Exploit Generation for Heap Overflows in Interpreters
- Poster: Fuzzing IoT Firmware via Multi-stage Message Generation
S&P 2019
- NEUZZ: Efficient Fuzzing with Neural Program Smoothing
- Fuzzing File Systems via Two-Dimensional Input Space Exploration
- ProFuzzer: On-the-fly Input Type Probing for Better Zero-day Vulnerability Discovery
- Razzer: Finding Kernel Race Bugs through Fuzzing
- Full-speed Fuzzing: Reducing Fuzzing Overhead through Coverage-guided Tracing
USENIX Security 2019
- MOPT: Optimize Mutation Scheduling for Fuzzers
- Antifuzz: impeding fuzzing audits of binary executables
- FUZZIFICATION: Anti-Fuzzing Technique
- EnFuzz: Ensemble Fuzzing with Seed Synchronization among Diverse Fuzzers
- GRIMOIRE : Synthesizing Structure while Fuzzing
- RVFuzzer: Finding Input Validation Bugs in Robotic Vehicles through Control-Guided Random Testing
- FIRM-AFL: High-Throughput Greybox Fuzzing of IoT Firmware via Augmented Process Emulation
- Unicorefuzz: On the Viability of Emulation for Kernelspace Fuzzing
ASE 2019
- Learning-Guided Network Fuzzing for Testing Cyber-Physical System Defences
NDSS 2019
- REDQUEEN: Fuzzing with Input-to-State Correspondence
- PeriScope: An Effective Probing and Fuzzing Framework for the Hardware-OS Boundary
- Life after Speech Recognition: Fuzzing Semantic Misinterpretation for Voice Assistant Applications
- Send Hardest Problems My Way: Probabilistic Path Prioritization for Hybrid Fuzzing
- CodeAlchemist: Semantics-Aware Code Generation to Find Vulnerabilities in JavaScript Engines
- NAUTILUS: Fishing for Deep Bugs with Grammars
ICSE 2019
- DifFuzz: Differential Fuzzing for Side-Channel Analysis
- Deep Differential Testing of JVM Implementations
- REST-ler: Stateful REST API Fuzzing
- SLF: Fuzzing without Valid Seed Inputs
- Superion: Grammar-Aware Greybox Fuzzing
- Hunting for bugs in code coverage tools via randomized differential testing
ESEC/FSE 2019
- Cerebro: Context-aware Adaptive Fuzzing for Effective Vulnerability Detection
- Just Fuzz It: Solving Floating-Point Constraints Using Coverage-guided Fuzzing
- FUDGE: Fuzz Driver Generation at Scale
ISSTA 2019
- Semantic Fuzzing with Zest
- DeepHunter: A Coverage-Guided Fuzz Testing Framework for Deep Neural Networks
- Deferred Concretization in Symbolic Execution via Fuzzing
PLDI 2019
- Parser-Directed Fuzzing
ASIACCS 2019
- Ptrix: Efficient Hardware-Assisted Fuzzing for COTS Binary
- A Feature-Oriented Corpus for understanding, Evaluating and Improving Fuzz Testing
ICST 2019
- MemFuzz: Using Memory Accesses to Guide Fuzzing
- SeqFuzzer: An Industrial Protocol Fuzzing Framework in Deep Learning Perspective
ISSRE 2019
- Engineering a Better Fuzzer with SynergicallyIntegrated Optimizations
- Fuzzing Error Handling Code in Device Drivers Based on Software Fault Injection
Other 2019
- Fuzz Testing of Multithreaded Applications Based on Waiting (2019)
- Leveraging Textual Specifications for Grammar-Based Fuzzing of Network Protocols (AAAI 2019)
- Fuzzing JavaScript Environment APIs with Interdependent Function Calls (IFM 2019)
- DeepFuzzer: Accelerated Deep Greybox Fuzzing (TDSC 2019)
- CAGFuzz: Coverage-Guided Adversarial Generative Fuzzing Testing of Deep Learning Systems
- Suzzer: A Vulnerability-Guided Fuzzer Based on Deep Learning
- ConFuz: A Concurrency Fuzzer
- INSTRCR: Lightweight instrumentation optimization based on coverage-guided fuzz testing
- HFuzz: Towards automatic fuzzing testing of NB-IoT core network protocols implementations
- Study and Comparison of General Purpose Fuzzers
- From proof-of-concept to exploitable
- Sequence coverage directed greybox fuzzing
- Field-aware Evolutionary Fuzzing Based on Input Specifications and Vulnerability Metrics
- Fuzzing IPC with Knowledge Inference
- Be Sensitive and Collaborative: Analyzing Impact of Coverage Metrics in Greybox Fuzzing
- Hydra: An Extensible Fuzzing Framework for Finding Semantic Bugs in File Systems
- Exploring Effective Fuzzing Strategies to Analyze Communication Protocols
- V-Fuzz: Vulnerability-Oriented Evolutionary Fuzzing
S&P 2018
- T-Fuzz: fuzzing by program transformation
- Angora: Efficient Fuzzing by Principled Search
- CollAFL: Path Sensitive Fuzzing
USENIX Security 2018
- FUZE: Towards Facilitating Exploit Generation for Kernel Use-After-Free Vulnerabilities
- MoonShine: Optimizing OS Fuzzer Seed Selection with Trace Distillation
- QSYM : A Practical Concolic Execution Engine Tailored for Hybrid Fuzzing
CCS 2018
- Evaluating fuzz testing
- Hawkeye Towards a Desired Directed Grey-box Fuzzer
- Revery: From Proof-of-Concept to Exploitable
NDSS 2018
- IoTFuzzer: Discovering Memory Corruptions in IoT Through App-based Fuzzing
- What You Corrupt Is Not What You Crash: Challenges in Fuzzing Embedded Devices
- Enhancing Memory Error Detection for Large-Scale Applications and Fuzz Testing
- INSTRIM Lightweight Instrumentation for Coverage-guided Fuzzing
ICSE 2018
- SAFL: increasing and accelerating testing coverage with symbolic execution and guided fuzzing
ESEC/FSE 2018
- Singularity: Pattern Fuzzing for Worst Case Complexity
- PAFL: Extend Fuzzing Optimizations of Single Mode to Industrial Parallel Mode
- FOT: A Versatile, Configurable, Extensible Fuzzing Framework
- DLFuzz: Differential Fuzzing Testing of Deep Learning Systems
ASE 2018
- ContractFuzzer: Fuzzing Smart Contracts for Vulnerability Detection
- FairFuzz: A Targeted Mutation Strategy for Increasing Greybox Fuzz Testing Coverage
- Differential Program Analysis with Fuzzing and Symbolic Execution
ISSTA 2018
- Badger: Complexity Analysis with Fuzzing and Symbolic Execution
- Compiler Fuzzing through Deep Learning
- PerfFuzz: Automatically Generating Pathological Inputs
CAV 2018
- StringFuzz: A fuzzer for string solvers
ACSAC 2018
- TIFF: Using Input Type Inference To Improve Fuzzing
- A Heuristic Framework to Detect Concurrency Vulnerabilities
ICCAD 2018
- RFUZZ: Coverage-Directed Fuzz Testing of RTL on FPGAs
Other 2018
- FuzzerGym: A Competitive Framework for Fuzzing and Learning
- MTF-Storm: a high performance fuzzer for Modbus-TCP
S&P 2017
- NEZHA: Efficient Domain-Independent Differential Testing
- Skyfire: Data-Driven Seed Generation for Fuzzing
USENIX Security 2017
- kAFL: Hardware-Assisted Feedback Fuzzing for OS Kernels
- CAB-Fuzz: Practical Concolic Testing Techniques for COTS Operating Systems
CCS 2017
- Directed Greybox Fuzzing
- Designing New Operating Primitives to Improve Fuzzing Performance
- DIFUZE: Interface aware fuzzing for kernel drivers
- SemFuzz: Semantics-based Automatic Generation of Proof-of-Concept Exploits
- SlowFuzz: Automated Domain-Independent Detection of Algorithmic Complexity Vulnerabilities
- IMF: Inferred Model-based Fuzzer
NDSS 2017
- VUzzer: Application-aware Evolutionary Fuzzing
ESEC/FSE 2017
- Steelix: Program-State Based Binary Fuzzing
ASE 2017
- Learn&Fuzz: Machine Learning for Input Fuzzing
- In-memory fuzzing for binary code similarity analysis
ISSRE 2017
- Chizpurfle: A Gray-Box Android Fuzzer for Vendor Service Customizations
Other 2017
- Bbuzz: A Bit-aware Fuzzing Framework for Network Protocol Systematic Reverse Engineering and Analysis
CCS 2016
- Coverage-based Greybox Fuzzing as Markov Chain
- Systematic Fuzzing and Testing of TLS Libraries
NDSS 2016
- Driller: Argumenting Fuzzing Through Selective Symbolic Execution
PLDI 2016
- Coverage-Directed Differential Testing of JVM Implementations
S&P 2015
- Program-Adaptive Mutational Fuzzing
USENIX Security 2015
- Protocol State Fuzzing of TLS Implementations
Others
- Test Data Generation for Stateful Network Protocol Fuzzing Using a Rule-Based State Machine
- PULSAR: Stateful Black-Box Fuzzing of Proprietary Network Protocols
- A Modbus-TCP Fuzzer for testing internetworked industrial systems
- SECFUZZ: Fuzz-testing Security Protocols
- AutoFuzz: Automated Network Protocol Fuzzing Framewor

All Papers (Classification according to Subject)

Recent Papers Related To Fuzzing
- Contributors
- Starchart
All Papers (Classification according to Publication)
All Papers (Classification according to Subject)
Survey/Review
- A Survey of Protocol Fuzzing
- Demystify the Fuzzing Methods: A Comprehensive Survey (ACM CSUR 2023)
- Fuzzing: Challenges and Reflections
- The Progress, Challenges, and Perspectives of Directed Greybox Fuzzing (STVR)
- Fuzzing: Hack, Art, and Science (CACM 2020)
- Survey of Directed Fuzzy Technology
- A Review of Machine Learning Applications in Fuzzing
- A systematic review of fuzzing based on machine learning techniques
- The Art, Science, and Engineering of Fuzzing: A Survey
- Fuzzing: Art, Science, and Engineering
- Fuzzing: a survey
- Fuzzing: State of the art
- Fuzzing: A Survey for Roadmap
- Survey of Software Fuzzing Techniques
- A Review of Fuzzing Tools and Methods
- Embedded fuzzing: a review of challenges, tools, and solutions
Differential Fuzzing
- JIT-Picking: Differential Fuzzing of JavaScript Engines (CCS 2022)
- SpecDoctor: Differential Fuzz Testing to Find Transient Execution Vulnerabilities
- SEDiff: Scope-Aware Differential Fuzzing to Test Internal Function Models in Symbolic Execution (FSE 2022)
- T-Reqs: HTTP Request Smuggling with Differential Fuzzing (CCS 2021)
- CatchBackdoor: Backdoor Testing by Critical Trojan Neural Path Identification via Differential Fuzzing (2021)
- Duo: Differential Fuzzing for Deep Learning Operators (IEEE Transactions on Reliability 2021)
- DiFuzzRTL: Differential Fuzz Testing to Find CPU Bug (S&P 2021)
- DPIFuzz: A Differential Fuzzing Framework to Detect DPI Elusion Strategies for QUIC (ACSAC 2020)
- DifFuzz: Differential Fuzzing for Side-Channel Analysis (ICSE 2019)
- Deep Differential Testing of JVM Implementations (ICSE 2019)
- Hunting for bugs in code coverage tools via randomized differential testing (ICSE 2019)
- Different is Good: Detecting the Use of Uninitialized Variables through Differential Replay (CCS 2019)
- Differential Program Analysis with Fuzzing and Symbolic Execution (ASE 2018)
- NEZHA: Efficient Domain-Independent Differential Testing (S&P 2017)
- Coverage-Directed Differential Testing of JVM Implementations (PLDI 2016)
Evaluate Fuzzing
- SoK: Prudent Evaluation Practices for Fuzzing (SP 2024)
- FIXREVERTER: A Realistic Bug Injection Methodology for Benchmarking Fuzz Testing (USENIX Security2022)
- On the Reliability of Coverage-Based Fuzzer Benchmarking (ICSE 2022)
- Mutation Analysis: Answering the Fuzzing Challenge (2022)
- Evaluating Code Coverage for Kernel Fuzzers via Function Call Graph (Access 2021)
- FuzzBench: An Open Fuzzer Benchmarking Platform and Service (FSE 2021)
- An Empirical Study of OSS-Fuzz Bugs (MSR 2021)
- Industrial Oriented Evaluation of Fuzzing Techniques (ICST 2021)
- UNIFUZZ: A Holistic and Pragmatic Metrics-Driven Platform for Evaluating Fuzzers (USENIX Security2021)
- My Fuzzer Beats Them All! Developing a Framework for Fair Evaluation and Comparison of Fuzzers (2021)
- A Quantitative Comparison of Covera (AST 2020)
- Fuzzing: On the Exponential Cost of Vulnerability Discovery (FSE 2020)
- A Feature-Oriented Corpus for understanding, Evaluating and Improving Fuzz Testing (ASIACCS 2019)
- Be Sensitive and Collaborative: Analyzing Impact of Coverage Metrics in Greybox Fuzzing (RAID 2019)
- Study and Comparison of General Purpose Fuzzers
- Evaluating Fuzz Testing (CCS 2018)
Instrumentation
- InstruGuard: Find and Fix Instrumentation Errors for Coverage-based Greybox Fuzzing (ASE 2021)
- RIFF: Reduced Instruction Footprint for Coverage-Guided Fuzzing (USENIX ATC 2021)
- Hashing Fuzzing: Introducing Input Diversity to Improve Crash Detection (TSE 2021)
- RetroWrite: Statically Instrumenting COTS Binaries for Fuzzing and Sanitization (S&P 2020)
- INSTRCR: Lightweight instrumentation optimization based on coverage-guided fuzz testing (CCET 2019)
- Full-speed Fuzzing: Reducing Fuzzing Overhead through Coverage-guided Tracing (S&P 2019)
- INSTRIM Lightweight Instrumentation for Coverage-guided Fuzzing (NDSS 2018 workshop)
- SyzGen: Automated Generation of Syscall Specification of Closed-Source macOS Drivers (CCS 2021)
IoT or protocols fuzzing
- LABRADOR: Response Guided Directed Fuzzing for Black-box IoT Devices (SP 2024)
- PrIntFuzz: Fuzzing Linux Drivers via Automated Virtual Device Simulation (ISSTA 2022)
- SnapFuzz: High-Throughput Fuzzing of Network Applications (ISSTA 2022)
- Efficient Greybox Fuzzing of Applications in Linux-based IoT Devices via Enhanced User-mode Emulation (ISSTA 2022)
- FuzzUSB: Hybrid Stateful Fuzzing of USB Gadget Stacks (FSE 2022)
- BrakTooth: Causing Havoc on Bluetooth Link Manager via Directed Fuzzing (USENIX SEC 2022)
- Drifuzz: Harvesting Bugs in Device Drivers from Golden Seeds (USENIX SEC 2022)
- StateFuzz: System Call-Based State-Aware Linux Driver Fuzzing (USENIX SEC 2022)
- SNPSFuzzer: A Fast Greybox Fuzzer for Stateful Network Protocols using Snapshots (2022)
- State Selection Algorithms and Their Impact on The Performance of Stateful Network Protocol Fuzzing (2021)
- ICS3Fuzzer: A Framework for Discovering Protocol Implementation Bugs in ICS Supervisory Software by Fuzzing (ACSAC 2021)
- Westworld: Fuzzing-Assisted Remote Dynamic Symbolic Execution of Smart Apps on IoT Cloud Platforms (ACSAC 2021)
- ProFuzzBench - A Benchmark for Stateful Protocol Fuzzing (ISSTA 2021)
- TCP-Fuzz: Detecting Memory and Semantic Bugs in TCP Stacks with Fuzzing (USENIX ATC 2021)
- ICPFuzzer: proprietary communication protocol fuzzing by using machine learning and feedback strategies (Cybersecurity 2021)
- Fuzzing With Optimized Grammar-Aware Mutation Strategies (Access 2021)
- FIRM-COV: High-Coverage Greybox Fuzzing for IoT Firmware via Optimized Process Emulation (Access 2021)
- DIANE: Identifying Fuzzing Triggers in Apps to Generate Under-constrained Inputs for IoT Devices (S&P 2020)
- Snipuzz: Black-box Fuzzing of IoT Firmware via Message Snippet Inference (CCS 2021)
- Learning-Based Fuzzing of IoT Message Brokers (ICST 2021)
- RiverFuzzRL - an open-source tool to experiment with reinforcement learning for fuzzing (ICST 2021)
- Vulnerability Detection in SIoT Applications: A Fuzzing Method on their Binaries (IEEE Transactions on Network Science and Engineering 2020)
- Analysis of DTLS Implementations Using Protocol State Fuzzing (USENIX Security2020)
- Frankenstein: Advanced Wireless Fuzzing to Exploit New Bluetooth Escalation Targets (USENIX Security2020)
- A deep convolution generative adversarial networks based fuzzing framework for industry control protocols
- ICS Protocol Fuzzing: Coverage Guided Packet Crack and Generation (DAC 2020)
- AFLNET: A Greybox Fuzzer for Network Protocols (ICST 2020)
- Finding Security Vulnerabilities in Network Protocol Implementations (Arxiv 2020)
- Smart seed selection-based effective black box fuzzing for IIoT protocol (2020)
- Fw-fuzz: A code coverage-guided fuzzing framework for network protocols on firmware (2020)
- BaseSAFE: Baseband SAnitized Fuzzing through Emulation (WiSec 2020)
- Poster: Fuzzing IoT Firmware via Multi-stage Message Generation (CCS 2019)
- SeqFuzzer: An Industrial Protocol Fuzzing Framework in Deep Learning Perspective (ICST 2019)
- SPFuzz: A Hierarchical Scheduling Framework for Stateful Network Protocol Fuzzing (IEEE Access 2019)
- HFuzz: Towards automatic fuzzing testing of NB-IoT core network protocols implementations (FGCS 2019)
- FIRM-AFL: High-Throughput Greybox Fuzzing of IoT Firmware via Augmented Process Emulation (USENIX Security2019)
- Exploring Effective Fuzzing Strategies to Analyze Communication Protocols (FEAST 2019)
- Leveraging Textual Specifications for Grammar-Based Fuzzing of Network Protocols (AAAI 2019)
- MTF-Storm: a high performance fuzzer for Modbus-TCP (ETFA 2018)
- Advancing Protocol Fuzzing for Industrial Automation and Control Systems (ICISSP 2018)
- IoTFuzzer: Discovering Memory Corruptions in IoT Through App-based Fuzzing (NDSS 2018)
- Bbuzz: A Bit-aware Fuzzing Framework for Network Protocol Systematic Reverse Engineering and Analysis (MCC 2017)
- Test Data Generation for Stateful Network Protocol Fuzzing Using a Rule-Based State Machine (2016)
- Protocol State Fuzzing of TLS Implementations (USENIX Security2015)
- A Modbus-TCP Fuzzer for testing internetworked industrial systems (ETFA 2015)
- PULSAR: Stateful Black-Box Fuzzing of Proprietary Network Protocols (Springer, Cham, 2015)
- SECFUZZ: Fuzz-testing Security Protocols (AST 2012)
- Extension of SPIKE for Encrypted Protocol Fuzzing (2011)
- AutoFuzz: Automated Network Protocol Fuzzing Framework (IJCSNS 2010)
Fuzzing with LLMs
- Large Language Model guided Protocol Fuzzing (NDSS 2024)
- Large Language Models are Zero-Shot Fuzzers:Fuzzing Deep-Learning Libraries via Large Language Models (ISSTA 2023)
- Large Language Models are Few-shot Testers: Exploring LLM-based General Bug Reproduction (ICSE 2023)
SMT Fuzzing
- Validating SMT Solvers via Skeleton Enumeration Empowered by Historical Bug-Triggering Inputs (ICSE 2023)
- Alt-Ergo-Fuzz: A fuzzer for the Alt-Ergo SMT solver (JFLA 2022)
- BanditFuzz: Fuzzing SMT Solvers with Multi-agent Reinforcement Learning (FM 2021)
- Skeletal Approximation Enumeration for SMT Solver Testing (FSE 2021)
- Fuzzing SMT Solvers via Two-Dimensional Input Space Exploration (ISSTA 2021)
- Detecting Critical Bugs in SMT Solvers Using Blackbox Mutational Fuzzing (FSE 2020)
- On the Unusual Effectiveness of Type-aware Mutations for Testing SMT Solvers
- BanditFuzz: Fuzzing SMT Solvers with Reinforcement Learning (2020)
- Validating SMT Solvers via Semantic Fusion (PLDI 2020)
- Automatically Testing String Solvers (ICSE 2020)
- StringFuzz: A fuzzer for string solvers (CAV 2018)
Anti Fuzzing
- Antifuzz: impeding fuzzing audits of binary executables (USENIX Security2019)
- FUZZIFICATION: Anti-Fuzzing Technique (USENIX Security2019)
Kernel Fuzzing
- No Grammar, No Problem: Towards Fuzzing the Linux Kernel without System-Call Descriptions (NDSS 2023)
- SFuzz: Slice-based Fuzzing for Real-Time Operating Systems (CCS 2022)
- Demystifying the Dependency Challenge in Kernel Fuzzing (ICSE 2022)
- Semantic-Informed Driver Fuzzing Without Both the Hardware Devices and the Emulators (ICSE 2022)
- SyzScope: Revealing High-Risk Security Impacts of Fuzzer-Exposed Bugs in Linux kernel (USENIX SEC'22)
- HEALER: Relation Learning Guided Kernel Fuzzing (SOSP 2021)
- NTFUZZ: Enabling Type-Aware Kernel Fuzzing on Windows with Static Binary Analysis(S&P 2021)
- Finding Bugs in File Systems with an Extensible Fuzzing Framework (TOS 2020)
- Finding race conditions in Kernels: from fuzzing to symbolic exection (2020)
- Hydra: An Extensible Fuzzing Framework for Finding Semantic Bugs in File Systems (SOSP 2019)
- Fuzzing File Systems via Two-Dimensional Input Space Exploration (S&P 2019)
- Unicorefuzz: On the Viability of Emulation for Kernelspace Fuzzing (USENIX WOOT'19)
- Razzer: Finding Kernel Race Bugs through Fuzzing (S&P 2019)
- MoonShine: Optimizing OS Fuzzer Seed Selection with Trace Distillation (USENIX Security2018)
- FUZE: Towards Facilitating Exploit Generation for Kernel Use-After-Free Vulnerabilities (USENIX Security2018)
- kAFL: Hardware-Assisted Feedback Fuzzing for OS Kernels (Usenix Security2017)
- DIFUZE: Interface aware fuzzing for kernel drivers (CCS 2017)
- IMF: Inferred Model-based Fuzzer (CCS 2017)
Hybrid Fuzzing
- Evaluating and Improving Hybrid Fuzzing (ICSE 2023)
- Sydr-Fuzz: Continuous Hybrid Fuzzing and Dynamic Analysis for Security Development Lifecycle (ISPRAS Open 2022)
- TensileFuzz: Facilitating Seed Input Generation in Fuzzing via String Constraint Solving (ISSTA 2022)
- CONFETTI: Amplifying Concolic Guidance for Fuzzers (ICSE 2022)
- FuSeBMC v. 4: Smart Seed Generation for Hybrid Fuzzing (2021)
- A Tight Integration of Symbolic Execution and Fuzzing (short paper 2021)
- Symbolic Security Predicates: Hunt Program Weaknesses (ISPRAS Open 2021)
- Towards Symbolic Pointers Reasoning in Dynamic Symbolic Execution (IVMEM 2021)
- FUZZOLIC: Mixing fuzzing and concolic execution (Computers&Security 2021)
- Concolic-Fuzzing of JavaScript Programs using GraalVM and Truffle (SKILL 2021)
- SHFuzz: A hybrid fuzzing method assisted by static analysis for binary programs (China Communications 2021)
- A Priority Based Path Searching Method for Improving Hybrid Fuzzing (Computers & Security 2021)
- Sydr: Cutting Edge Dynamic Symbolic Execution (ISPRAS Open 2020)
- CSEFuzz: Fuzz Testing Based on Symbolic Execution (Access 2020)
- Sequence directed hybrid fuzzing (SANER 2020)
- HFL: Hybrid Fuzzing on the Linux Kernel (NDSS 2020)
- PANGOLIN: Incremental Hybrid Fuzzing with Polyhedral Path Abstraction (S&P 2020)
- SAVIOR: Towards Bug-Driven Hybrid Testing (S&P 2020)
- Deferred Concretization in Symbolic Execution via Fuzzing (ISSTA 2019)
- Send Hardest Problems My Way: Probabilistic Path Prioritization for Hybrid Fuzzing (NDSS 2019)
- Intriguer: Field-Level Constraint Solving for Hybrid Fuzzing (CCS 2019)
- DeepFuzzer: Accelerated Deep Greybox Fuzzing (TDSC 2019)
- QSYM: A Practical Concolic Execution Engine Tailored for Hybrid Fuzzing (USENIX Security2018)
- Angora: Efficient Fuzzing by Principled Search (S&P 2018)
- SAFL: increasing and accelerating testing coverage with symbolic execution and guided fuzzing (ICSE 2018)
- CAB-Fuzz: Practical Concolic Testing Techniques for COTS Operating Systems (Usenix Security2017)
- Driller: Argumenting Fuzzing Through Selective Symbolic Execution (NDSS 2016)
- Hybrid Fuzz Testing - Discovering Software Bugs via Fuzzing and Symbolic Execution (2012)
- Hybrid concolic testing (2007)
Mutation\Coverage\Path
- GrayC: Greybox Fuzzing of Compilers and Analysers for C (ISSTA 2023)
- DARWIN: Survival of the Fittest Fuzzing Mutators (NDSS 2023)
- Rainfuzz: Reinforcement-Learning Driven Heat-Maps for Boosting Coverage-Guided Fuzzing (ICPRAM 2023)
- Evaluating the Fork-Awareness of Coverage-Guided Fuzzers (ICISSP 2023)
- One Fuzzing Strategy to Rule Them All (ICSE 2022)
- BeDivFuzz: Integrating Behavioral Diversity into Generator-based Fuzzing (ICSE 2022)
- FuzzingDriver: the Missing Dictionary to Increase Code Coverage in Fuzzers (SANER 2022)
- EMS: History-Driven Mutation for Coverage-based Fuzzing (NDSS 2022)
- OTA: An Operation-oriented Time Allocation Strategy for Greybox Fuzzing (SANER 2021)
- MaxAFL: Maximizing Code Coverage with a Gradient-Based Optimization Technique (Electronics 2020)
- PathAFL: Path-Coverage Assisted Fuzzing (ASIA CCS 2020)
- Zeror: Speed Up Fuzzing with Coverage-sensitive Tracing and Scheduling (ASE 2020)
- Not All Coverage Measurements Are Equal: Fuzzing by Coverage Accounting for Input Prioritization (NDSS 2020)
- Matryoshka: fuzzing deeply nested branches (CCS 2019)
- REDQUEEN: Fuzzing with Input-to-State Correspondence (NDSS2019)
- T-Fuzz: fuzzing by program transformation (S&P 2018)
- FairFuzz: A Targeted Mutation Strategy for Increasing Greybox Fuzz Testing Coverage (ASE 2018)
- VUzzer: Application-aware Evolutionary Fuzzing (NDSS 2017)
Grammars \ Semantic \ Context-aware Fuzzing
- FUZZILLI: Fuzzing for JavaScript JIT Compiler Vulnerabilities (NDSS 2023)
- FRAMESHIFTER: Manipulating HTTP/2 Frame Sequences with Fuzzing (Usenix Security2020)
- SGXFuzz: Efficiently Synthesizing Nested Structures for SGX Enclave Fuzzing (Usenix Security2022)
- Unicorn: Detect Runtime Error in Time-Series Databases With Hybrid Input Synthesis (ISSTA 2022)
- Cooper: Testing the Binding Code of Scripting Languages with Cooperative Mutation (NDSS 2022)
- Fuzzing Class Specifications (ICSE 2022)
- Efficient ECU Analysis Technology through Structure-aware CAN Fuzzing (Access 2022)
- Semantic Image Fuzzing of AI Perception Systems
- SoFi: Reflection-Augmented Fuzzing for JavaScript Engines (CCS 2021)
- V-SHUTTLE: Scalable and Semantics-Aware Hypervisor Fuzzing (CCS 2021)
- Token-Level Fuzzing (WiSec 2021)
- Extended grammar-based fuzzing algorithm for JavaScript Engines (2021)
- Gramatron: Effective Grammar-Aware Fuzzing (ISSTA 2021)
- One Engine to Fuzz 'em All: Generic Language Processor Testing with Semantic Validation (S&P 2021)
- Growing A Test Corpus with Bonsai Fuzzing (ICSE 2021)
- Favocado: Fuzzing the Binding Code of JavaScript Engines Using Semantically Correct Test Cases (NDSS 2021)
- CMFuzz: context-aware adaptive mutation for fuzzers (Empirical Software Engineering 2021)
- Generating Highly-structured Input Data by Combining Search-based Testing and Grammar-based Fuzzing (ASE 2020)
- Montage: A Neural Network Language Model-Guided JavaScript Engine Fuzzer (Usenix Security2020)
- Fuzzing JavaScript Engines with Aspect-preserving Mutation (S&P 2020)
- Language-Agnostic Generation of Compilable Test Programs (ICST 2020)
- Smart Greybox Fuzzing (TSE 2019)
- Semantic Fuzzing with Zest (ISSTA 2019)
- Field-aware Evolutionary Fuzzing Based on Input Specifications and Vulnerability Metrics (2019)
- Parser-Directed Fuzzing (PLDI 2019)
- GRIMOIRE: Synthesizing Structure while Fuzzing (USENIX Security2019)
- Life after Speech Recognition: Fuzzing Semantic Misinterpretation for Voice Assistant Applications (NDSS 2019)
- SLF: Fuzzing without Valid Seed Inputs (ICSE 2019)
- Superion: Grammar-Aware Greybox Fuzzing (ICSE 2019)
- ProFuzzer: On-the-fly Input Type Probing for Better Zero-day Vulnerability Discovery (S&P 2019)
- CodeAlchemist: Semantics-Aware Code Generation to Find Vulnerabilities in JavaScript Engines (NDSS 2019)
- NAUTILUS: Fishing for Deep Bugs with Grammars (NDSS 2019)
- TIFF: Using Input Type Inference To Improve Fuzzing (ACSAC 2018)
- Skyfire: Data-Driven Seed Generation for Fuzzing (S&P 2017)
Exploit Generation
- ETHPLOIT: From Fuzzing to Efficient Exploit Generation against Smart Contracts (SANER2020)
- Gollum: Modular and Greybox Exploit Generation for Heap Overflows in Interpreters (CCS 2019)
- From proof-of-concept to exploitable (Cybersecurity 2019)
- Revery: From Proof-of-Concept to Exploitable (CCS 2018)
- SemFuzz: Semantics-based Automatic Generation of Proof-of-Concept Exploits (CCS 2017)
- ExploitMeter: Combining Fuzzing with Machine Learning for Automated Evaluation of Software Exploitability (PAC 2017)
Parallel / Ensemble Fuzzing
- UltraFuzz: Towards Resource-saving in Distributed Fuzzing (TSE 2022)
- Towards Systematic and Dynamic Task Allocation for Collaborative Parallel Fuzzing (ASE 2021 NIER)
- CollabFuzz: A Framework for Collaborative Fuzzing (EuroSec 2021)
- Improving Web Application Vulnerability Detection Leveraging Ensemble Fuzzing (ENASE 2021)
- Cupid: Automatic Fuzzer Selection for Collaborative Fuzzing (ACSAC 2020)
- EnFuzz: Ensemble Fuzzing with Seed Synchronization among Diverse Fuzzers (USENIX Security2019)
- PAFL: Extend FuzzingOptimizations of Single Mode to Industrial Parallel Mode (ESEC/FSE 2018)
Sanitizer-guided Fuzzing
- FishFuzz: Catch Deeper Bugs by Throwing Larger Nets (USENIX Security2023)
- ParmeSan: Sanitizer-guided Greybox Fuzzing (USENIX Security2020)
State / Sequence Guided Fuzzing
- The Bandit’s States: Modeling State Selection for Stateful Network Fuzzing as Multi-armed Bandit Problem
- SWaTEval: An Evaluation Framework for Stateful Web Application Testing (ICISSP 2023)
- Stateful Greybox Fuzzing (USENIX Security 2022)
- Linear-time Temporal Logic guided Greybox Fuzzing (ICSE 2022)
- Rtkaller: State-aware Task Generation for RTOS Fuzzing (EMSOFT 2021)
- Typestate-Guided Fuzzer for Discovering Use-after-Free Vulnerabilities (ICSE 2020)
- IJON: Exploring Deep State Spaces via Fuzzing (S&P 2020)
- MemFuzz: Using Memory Accesses to Guide Fuzzing (ICST 2019)
Rust Fuzzing
- Rust Library Fuzzing via API Dependency Graph Traversal (ASE 2021)
- RUSTY: A Fuzzing Tool for Rust (ACSAC 2020)
Regression Fuzzing
- R2Z2: Detecting Rendering Regressions in Web Browsers through Differential Fuzz Testing (ICSE 2022)
- FADATest: Fast and Adaptive Performance Regression Testing of Dynamic Binary Translation Systems (ICSE 2022)
- Regression Greybox Fuzzing (CCS 2021)
- DeltaFuzz: Historical Version Information Guided Fuzz Testing (Journal of Computer Science and Technology 2021)
- HyDiff: Hybrid Differential Software Analysis (ICSE 2020)
Directed Fuzzing
- Everything is Good for Something: Counterexample-Guided Directed Fuzzing via Likely Invariant Inference (SP 2024)
- LABRADOR: Response Guided Directed Fuzzing for Black-box IoT Devices (SP 2024)
- Titan: Efficient Multi-target Directed Greybox Fuzzing (SP 2024)
- FishFuzz: Catch Deeper Bugs by Throwing Larger Nets (USENIX Security2023)
- MC^2: Rigorous and Efficient Directed Greybox Fuzzing (CCS 2022)
- WindRanger: A Directed Greybox Fuzzer driven by Deviation Basic Block (ICSE 2022)
- BEACON: Directed Grey-Box Fuzzing with Provable Path Pruning (S&P 2022)
- Improving Configurability of Unit-level Continuous Fuzzing: An Industrial Case Study with SAP HANA (ASE 2021 Industry)
- KCFuzz: Directed Fuzzing Based on Keypoint Coverage (ICAIS 2021)
- Constraint-guided Directed Greybox Fuzzing (USENIX Security2021)
- Constructing More Complete Control Flow Graphs Utilizing Directed Gray-Box Fuzzing (MDPI 2021)
- Binary-level Directed Fuzzing for Use-After-Free Vulnerabilities (RAID 2020)
- Ankou: Guiding Grey-box Fuzzing towards Combinatorial Difference (ICSE 2020)
- RDFuzz: Accelerating Directed Fuzzing with Intertwined Schedule and Optimized Mutation (2020)
- TOFU: Target-Oriented FUzzer (Arxiv 2020)
- Sequence coverage directed greybox fuzzing (ICPC 2019)
- Hawkeye: Towards a Desired Directed Grey-box Fuzzer (CCS 2018)
- RFUZZ: Coverage-Directed Fuzz Testing of RTL on FPGAs (ICCAD 2018)
- Directed Greybox Fuzzing (CCS 2017)
Addressing Collision:
- CollAFL: Path Sensitive Fuzzing (S&P 2018)
Performance Fuzzing
- Understanding and Detecting Performance Bugs in Markdown Compilers (ASE 2021)
- HotFuzz: Discovering Algorithmic Denial-of-Service Vulnerabilities Through Guided Micro-Fuzzing (NDSS 2020)
- MemLock: Memory Usage Guided Fuzzing (ICSE2020)
- Singularity: Pattern Fuzzing for Worst Case Complexity (FSE 2018)
- PerfFuzz: Automatically Generating Pathological Inputs (ISSTA 2018)
- Badger: Complexity Analysis with Fuzzing and Symbolic Execution (ISSTA 2018)
- SlowFuzz: Automated Domain-Independent Detection of Algorithmic Complexity Vulnerabilities (CCS 2017)
Enhancing Memory Error:
- OBSan: An Out-Of-Bound Sanitizer to Harden DNN Executables (NDSS 2023)
- SANRAZOR: Reducing Redundant Sanitizer Checks in C/C++ Programs (OSDI 2021)
- Unleashing Fuzzing Through Comprehensive, Efficient, and Faithful Exploitable-Bug Exposing
- HDR-Fuzz: Detecting Buffer Overruns using AddressSanitizer Instrumentation and Fuzzing (2021)
- Enhancing Memory Error Detection for Large-Scale Applications and Fuzz Testing (NDSS 2018)
- AddressSanitizer: A Fast Address Sanity Checker (USENIX Security2012)
Schedule (Power & Mutation)
- SLIME: Program-sensitive Energy Allocation for Fuzzing (ISSTA 2022)
- MobFuzz: Adaptive Multi-objective Optimization in Gray-box Fuzzing (NDSS 2022)
- Effective Seed Scheduling for Fuzzing with Graph Centrality Analysis (S&P 2022)
- Path Transitions Tell More: Optimizing Fuzzing Schedules via Runtime Program States (ICSE 2022)
- FEFuzzer: Hybrid Files Fuzzing Tool (2022)
- Seed Selection for Successful Fuzzing (ISSTA 2021)
- MooFuzz: Many-Objective Optimization Seed Schedule for Fuzzer (Mathematics 2021)
- EcoFuzz: Adaptive Energy-Saving Greybox Fuzzing as a Variant of the Adversarial Multi-Armed Bandit (USENIX Security2020)
- MEUZZ: Smart Seed Scheduling for Hybrid Fuzzing (2020)
- Greybox Fuzzing Based on Ant Colony Algorithm (AINA 2020)
- Suzzer: A Vulnerability-Guided Fuzzer Based on Deep Learning (Inscrypt 2019)
- MOPT: Optimize Mutation Scheduling for Fuzzers (USENIX Security2019)
- Cerebro: Context-aware Adaptive Fuzzing for Effective Vulnerability Detection (FSE 2019)
- Coverage-based Greybox Fuzzing as Markov Chain (CCS 2016)
- Program-Adaptive Mutational Fuzzing (S&P 2015)
Learning-based Fuzzing
- Evaluating and Improving Neural Program-Smoothing-based Fuzzing (ICSE 2022)
- Better Pay Attention Whilst Fuzzing (2022)
- Automated Conformance Testing for JavaScript Engines via Deep Compiler Fuzzing (PLDI 2021)
- Reinforcement Learning-based Hierarchical Seed Scheduling for Greybox Fuzzing (NDSS 2021)
- Neural Network Guided Evolutionary Fuzzing for Finding Traffic Violations of Autonomous Vehicles (arXiv 2021)
- OmniFuzz: A Flexible Framework for Expediting Bug Finding by Leveraging Past (Mis-)Behavior to Discover New Bugs (ACSAC 2020)
- Learning Input Tokens for Effective Fuzzing (ISSTA 2020)
- MTFuzz: Fuzzing with a Multi-task Neural Network (FSE 2020)
- FuzzGuard: Filtering out Unreachable Inputs in Directed Grey-box Fuzzing through Deep Learning (USENIX Security2020)
- LearnAFL: Greybox Fuzzing With Knowledge Enhancement (Access 2019)
- NeuFuzz: Efficient Fuzzing With Deep Neural Network (Access 2019)
- Learning-Guided Network Fuzzing for Testing Cyber-Physical System Defences (ASE 2019)
- NEUZZ: Efficient Fuzzing with Neural Program Smoothing (S&P 2019)
- V-Fuzz: Vulnerability-Oriented Evolutionary Fuzzing (Arxiv 2019)
- Compiler Fuzzing through Deep Learning (ISSTA 2018)
- Deep Reinforcement Fuzzing (SPW 2018)
- FuzzerGym: A Competitive Framework for Fuzzing and Learning (arxiv 2018)
- Learn&Fuzz: Machine Learning for Input Fuzzing (ASE 2017)
Fuzzing Deep Learning Compilers and Inference Engines
- Graph-based Fuzz Testing for Deep Learning Inference Engines (ICSE 2021)
- Coverage-Guided Tensor Compiler Fuzzing with Joint IR-Pass Mutation (OOPSLA 2022)
- NNSmith: Generating Diverse and Valid Test Cases for Deep Learning Compilers (ASPLOS 2023)
- Fuzzing Deep Learning Compilers with HirGen (ISSTA 2023)
Fuzzing Deep Learning Libraries
- DocTer: Documentation-Guided Fuzzing for Testing Deep Learning API Functions (ISSTA 2022)
- Fuzzing Deep-Learning Libraries via Automated Relational API Inference (ESEC/FSE 2022)
- Free Lunch for Testing: Fuzzing Deep-Learning Libraries from Open Source (ICSE 2022)
- Muffin: Testing Deep Learning Libraries via Neural Architecture Fuzzing (ICSE 2022)
Fuzzing Machine Learning Model
- MDPFuzz: Testing Models Solving Markov Decision Processes (ISSTA 2022)
- RapidFuzz: Accelerating fuzzing via Generative Adversarial Networks (Neurocomputing 2021)
- CoCoFuzzing: Testing Neural Code Models with Coverage-Guided Fuzzing (2021)
- Fuzz Testing based Data Augmentation to Improve Robustness of Deep Neural Networks (ICSE 2020)
- Coverage Guided Differential Adversarial Testing of Deep Learning Systems (TNSE 2020)
- CAGFuzz: Coverage-Guided Adversarial Generative Fuzzing Testing of Deep Learning Systems (2019)
- DeepHunter: A Coverage-Guided Fuzz Testing Framework for Deep Neural Networks (ISSTA 2019)
- TensorFuzz: Debugging Neural Networks with Coverage-Guided Fuzzing (ICML 2019)
- DLFuzz: Differential Fuzzing Testing of Deep Learning Systems (FSE 2018)
Data Flow Sensitive Fuzzing
- PATA: Fuzzing with Path Aware Taint Analysis (S&P 2022)
- datAFLow: Towards a Data-Flow-Guided Fuzzer (NDSS 2022)
- ovAFLow: Detecting Memory Corruption Bugs with Fuzzing-based Taint Inference (Journal of Computer Science and Technology 2021)
- DIAR: Removing Uninteresting Bytes from Seeds in Software Fuzzing (2021)
- GREYONE: Data Flow Sensitive Fuzzing (USENIX Security2020)
Binary Fuzzing
- spotFuzzer: Static Instrument and Fuzzing Windows COTs (2022)
- Emmutaler: Fuzzing the iOS Boot Loader (Thesis 2021)
- Same Coverage, Less Bloat: Accelerating Binary-only Fuzzing with Coverage-preserving Coverage-guided Tracing (CCS 2021)
- Scalable Fuzzing of Program Binaries with E9AFL (ASE 2021)
- STOCHFUZZ: Sound and Cost-effective Fuzzing of Stripped Binaries by Incremental and Stochastic Rewriting (S&P 2021)
- Coverage-guided binary fuzzing with rev.ng and llvm libfuzzer
- Breaking Through Binaries: Compiler-quality Instrumentation for Better Binary-only Fuzzing (USENIX Security2021)
- WEIZZ: Automatic Grey-box Fuzzing for Structured Binary Formats
- Ptrix: Efficient Hardware-Assisted Fuzzing for COTS Binary (ASIACCS 2019)
- Steelix: Program-State Based Binary Fuzzing (FSE 2017)
- In-memory fuzzing for binary code similarity analysis (ASE 2017)
Smart Contracts
- WASAI: Uncovering Vulnerabilities in Wasm Smart Contracts (ISSTA 2022)
- Smart Contract Fuzzing for Enterprises: The Language Agnostic Way (COMSNETS 2022)
- looking for lacunae in bitcoin cores fuzzing efforts (2022)
- Machine Learning Guided Cross-Contract Fuzzing (2021)
- SMARTIAN : Enhancing Smart Contract Fuzzing with Static and Dynamic Data-Flow Analyses (ASE 2021)
- HFContractFuzzer: Fuzzing Hyperledger Fabric Smart Contracts for Vulnerability Detection (EASE 2021)
- sFuzz: An Efficient Adaptive Fuzzer for Solidity Smart Contracts (ICSE 2020)
- Targeted Greybox Fuzzing with Static Lookahead Analysis (ICSE 2020)
- Learning to Fuzz from Symbolic Execution with Application to Smart Contracts (CCS 2019)
- ContractFuzzer: Fuzzing Smart Contracts for Vulnerability Detection (ASE 2018)
Constraint Solving
- JIGSAW: Efficient and Scalable Path Constraints Fuzzing (S&P 2022)
- Fuzzing Symbolic Expressions (ICSE 2021)
- Just Fuzz It: Solving Floating-Point Constraints Using Coverage-guided Fuzzing (FSE 2019)
Side-Channel Detection
- QFuzz: Quantitative Fuzzing for Side Channels (ISSTA 2021)
- JVM Fuzzing for JIT-Induced Side-Channel Detection (ICSE 2020)
- ct-fuzz: Fuzzing for Timing Leaks (ICST 2020)
Concurrency Fuzzing
- Context-Sensitive and Directional Concurrency Fuzzing for Data-Race Detection (NDSS 2022)
- Fuzzing with automatically controlled interleavings to detect concurrency bugs (The Journal of Systems & Software 2022)
- Controlled Concurrency Testing via Periodical Scheduling (ICSE 2022)
- Fuzzing Channel-Based Concurrency Runtimes using Types and Effects Slides (OOPSLA 2021)
- MUZZ: Thread-aware Grey-box Fuzzing for Effective Bug Hunting in Multithreaded Programs (USENIX Security2020)
- KRace: Data Race Fuzzing for Kernel File Systems (S&P 2020)
- ConFuzz: A Concurrency Fuzzer (2019)
- Fuzz Testing of Multithreaded Applications Based on Waiting (MICSECS 2019)
- A Heuristic Framework to Detect Concurrency Vulnerabilities (ACSAC 2018)
API Testing/Fuzzing
- Assessing the Impact of Interface Vulnerabilities in Compartmentalized Software (NDSS 2023)
- Minerva: Browser API Fuzzing with Dynamic Mod-Ref Analysis (FSE 2022)
- Combinatorial Testing of RESTful APIs (ICSE 2022)
- GraphFuzz: Library API Fuzzing with Lifetime-aware Dataflow Graphs (ICSE 2022)
- MOREST: Model-based RESTful API Testing with Execution Feedback (ICSE 2022)
- Automated Testing of Software that Uses Machine Learning APIs (ICSE 2022)
- Nessie: Automatically Testing JavaScript APIs with Asynchronous Callbacks (ICSE 2022)
- REST API Fuzzing by Coverage Level Guided Blackbox Testing (2021)
- FuzzGen: Automatic Fuzzer Generation (USENIX Security2020)
- Intelligent REST API Data Fuzzing (FSE 2020)
- FUDGE: Fuzz Driver Generation at Scale (FSE 2019)
- REST-ler: Stateful REST API Fuzzing (ICSE 2019)
- Fuzzing JavaScript Environment APIs with Interdependent Function Calls (IFM 2019)
- Systematic Fuzzing and Testing of TLS Libraries (CCS 2016)
Hardware Fuzzing
- CPU Fuzzing for Discovering Hardware-caused Information Leakage (2022)
- Fuzzing Hardware Like Software (USENIX SEC 2022)
- PeriScope: An Effective Probing and Fuzzing Framework for the Hardware-OS Boundary (NDSS2019)
Other Interesting Fuzzing
- LOKI: State-Aware Fuzzing Framework for the Implementation of Blockchain Consensus Protocols (NDSS 2023)
- DriveFuzz: Discovering Autonomous Driving Bugs through Driving Quality-Guided Fuzzing (CCS 2022)
- LibAFL: A Framework to Build Modular and Reusable Fuzzers (CCS 2022)
- RoboFuzz: Fuzzing Robotic Systems over Robot Operating System (ROS) for Finding Correctness Bugs (FSE 2022)
- Almost Correct Invariants: Synthesizing Inductive Invariants by Fuzzing Proofs (ISSTA 2022)
- ��AFL: Non-intrusive Feedback-driven Fuzzing for Microcontroller Firmware (ICSE 2022)
- CorbFuzz: Checking Browser Security Policies with Fuzzing (ASE 2021)
- Fully Automated Functional Fuzzing of Android Apps for Detecting Non-crashing Logic Bugs (OOPSLA 2021)
- webFuzz: Grey-Box Fuzzing for Web Applications (ESORICS 2021)
- Vulnerability-Oriented Fuzz Testing for Connected Autonomous Vehicle Systems ( IEEE Transactions on Reliability 2021)
- Hardware Support to Improve Fuzzing Performance and Precision (CCS 2021)
- Fuzzing: Cyberphysical System Testing for Security and Dependability (Computer 2021)
- ESRFuzzer: an enhanced fuzzing framework for physical SOHO router devices to discover multi-Type vulnerabilities (Cybersecurity 2021)
- Estimating Residual Risk in Greybox Fuzzing (FSE 2021)
- HeteroFuzz: Fuzz Testing to Detect Platform Dependent Divergence for Heterogeneous Applications (FSE 2021)
- DirectFuzz: Automated Test Generation for RTL Designs using Directed Graybox Fuzzing (DAC 2021)
- PMFuzz: Test Case Generation for Persistent Memory Programs (ASPLOS 2021)
- PGFUZZ: Policy-Guided Fuzzing for Robotic Vehicles (NDSS 2021)
- IntelliGen: Automatic Driver Synthesis for Fuzz Testing (ICSE 2021 SEIP)
- WINNIE: Fuzzing Windows Applications with Harness Synthesis and Fast Cloning (NDSS 2021)
- Industry Practice of Coverage-Guided Enterprise-Level DBMS Fuzzing (ICSE 2021)
- Refined Grey-Box Fuzzing with Sivo (arXiv 2021)
- Fuzzing Technique in Web Applications and Beyond (MCTE 2020)
- Nyx: Greybox Hypervisor Fuzzing using Fast Snapshots and Affine Types (USENIX Security2021)
- The Use of Likely Invariants as Feedback for Fuzzers (USENIX Security2021)
- FREEDOM: Engineering a State-of-the-Art DOM Fuzzer (CCS 2020)
- SQUIRREL: Testing Database Management Systems with Language Validity and Coverage Feedback (CCS 2020)
- BigFuzz: Efficient Fuzz Testing for Data Analytics using Framework Abstraction (ASE 2020)
- MoFuzz: A Fuzzer Suite for Testing Model-Driven Software Engineering Tools (ASE 2020)
- AFL++: Combining Incremental Steps of Fuzzing Research (USENIX Woot2020)
- Active Fuzzing for Testing and Securing Cyber-Physical Systems (ISSTA 2020)
- CrFuzz: Fuzzing Multi-purpose Programs through Input Validation (FSE 2020)
- SpecFuzz: Bringing Spectre-type vulnerabilities to the surface (USENIX Security2020)
- USBFuzz: A Framework for Fuzzing USB Drivers by Device Emulation (USENIX Security2020)
- Boosting Fuzzer Efficiency: An Information Theoretic Perspective (FSE 2020)
- Fuzzing Error Handling Code using Context-Sensitive Software Fault Injection (USENIX Security2020)
- FANS: Fuzzing Android Native System Services via Automated Interface Analysis (USENIX Security2020)
- Fuzzing IPC with Knowledge Inference (SRDS 2019)
- HYPER-CUBE: High-Dimensional Hypervisor Fuzzing (NDSS 2020)
- Reproducible Crashes: Fuzzing Pharo by Mutating the Test Methods (IWST20 2020)
- Opening Pandoras Box through ATFuzzer: Dynamic Analysis of AT Interface for Android Smartphones (ACSAC 2019)
- FuzzFactory: Domain-Specific Fuzzing with Waypoints (OOPSLA 2019)
- Compiler Fuzzing: How Much Does It Matter (OOPSLA2019)
- RVFuzzer: Finding Input Validation Bugs in Robotic Vehicles through Control-Guided Random Testing (USENIX Security2019)
- Engineering a Better Fuzzer with Synergically Integrated Optimizations (ISSRE 2019)
- Fuzzing Error Handling Code in Device Drivers Based on Software Fault Injection (ISSRE 2019)
- What You Corrupt Is Not What You Crash: Challenges in Fuzzing Embedded Devices (NDSS 2018)
- FOT: A Versatile, Configurable, Extensible Fuzzing Framework (FSE 2018)
- Designing New Operating Primitives to Improve Fuzzing Performance (CCS 2017)
- Chizpurfle: A Gray-Box Android Fuzzer for Vendor Service Customizations (ISSRE 2017)
- Fuzzware: Using Precise MMIO Modeling for Effective Firmware Fuzzing (USENIX Security2022)
- Auto Off-Target: Enabling Thorough and Scalable Testing for Complex Software Systems (ASE 2022)
  - Page Traffic Analysis

Survey/Review

A Survey of Protocol Fuzzing

Link
Paper

Abstract: Communication protocols form the bedrock of our interconnected world, yet vulnerabilities within their implementations pose significant security threats. Recent developments have seen a surge in fuzzing-based research dedicated to uncovering these vulnerabilities within protocol implementations. However, there still lacks a systematic overview of protocol fuzzing for answering the essential questions such as what the unique challenges are, how existing works solve them, etc. To bridge this gap, we conducted a comprehensive investigation of related works from both academia and industry. Our study includes a detailed summary of the specific challenges in protocol fuzzing, and provides a systematic categorization and overview of existing research efforts. Furthermore, we explore and discuss potential future research directions in protocol fuzzing. This survey serves as a foundational guideline for researchers and practitioners in the field.

Demystify the Fuzzing Methods: A Comprehensive Survey (ACM CSUR 2023)

Link
Paper

Abstract: Massive software applications possess complex data structures or parse complex data structures; in such cases, vulnerabilities in the software become inevitable. The vulnerabilities are the source of cyber-security threats, and discovering this before the software deployment is challenging. Fuzzing is a vulnerability discovery solution that resonates with random-mutation, feedback-driven, coverage-guided, constraint-guided, seed-scheduling, and target-oriented strategies. Each technique is wrapped beneath the black-, white-, and grey-box fuzzers to uncover diverse vulnerabilities. It consists of methods such as identifying structural information about the test cases to detect security vulnerabilities, symbolic and concrete program states to explore the unexplored locations, and full semantics of code coverage to create new test cases. We methodically examine each kind of fuzzers and contemporary fuzzers with a profound observation that addresses various research questions and systematically reviews and analyze the gaps and their solutions. Our survey comprised the recent related works on fuzzing techniques to demystify the fuzzing methods concerning the application domains and the target that, in turn, achieves higher code coverage and sound vulnerability detection.

Fuzzing: Challenges and Reflections

Paper

Abstract: Fuzzing is a method to discover software bugs and vulnerabilities by automatic test input generation which has found tremendous recent interest in both academia and industry. Fuzzing comes in the form of several techniques. On one hand, we have symbolic execution, which enables a particularly effective approach to fuzzing by systematically enumerating the paths of a program. On the other hand, we have random input generation, which generates large amounts of inputs per second with none or minimal program analysis overhead. In this article, we summarize the open challenges and opportunities for fuzzing and symbolic execution as they emerged in discussions among researchers and practitioners in a Shonan Meeting, and were validated in a subsequent survey. We take a forward-looking view of the software vulnerability discovery technologies and provide concrete directions for future research.

The Progress, Challenges, and Perspectives of Directed Greybox Fuzzing (STVR)

Abstract: Greybox fuzzing is a scalable and practical approach for software testing. Most greybox fuzzing tools are coverage-guided as reaching high code coverage is more likely to find bugs. However, since most covered codes may not contain bugs, blindly extending code coverage is less efficient, especially for corner cases. Unlike coverage-guided greybox fuzzing which increases code coverage in an undirected manner, directed greybox fuzzing (DGF) spends most of its time allocation on reaching specific targets (e.g. the bug-prone zone) without wasting resources stressing unrelated parts. Thus, DGF is particularly suitable for scenarios such as patch testing, bug reproduction, and special bug detection. For now, DGF has become an active research area. However, DGF has general limitations and challenges that are worth further studying. Based on the investigation of 42 state-of-the-art fuzzers that are closely related to DGF, we conducted the first in-depth study to summarize the empirical evidence on the research progress of DGF. This paper studies DGF from a broader view, which takes into account not only the location-directed type that targets specific code parts but also the behavior-directed type that aims to expose abnormal program behaviors. By analyzing the benefits and limitations of DGF research, we try to identify gaps in current research, meanwhile, reveal new research opportunities and suggest areas for further investigation.

Fuzzing: Hack, Art, and Science (CACM 2020)

Paper

Abstract: Fuzzing, or fuzz testing, is the process of finding security vulnerabilities in input-parsing code by repeatedly testing the parser with modified, or fuzzed, inputs.35 Since the early 2000s, fuzzing has become a mainstream practice in assessing software security. Thousands of security vulnerabilities have been found while fuzzing all kinds of software applications for processing documents, images, sounds, videos, network packets, Web pages, among others. These applications must deal with untrusted inputs encoded in complex data formats. For example, the Microsoft Windows operating system supports over 360 file formats and includes millions of lines of code just to handle all of these.

Survey of Directed Fuzzy Technology

Paper

Abstract: The fuzzy testing technology can effectively detect vulnerabilities. Based on Directed Symbolic Execution (DSE) fuzzing and Directed Grey Box Fuzzing (DGF), which can reach the specified target locations and scan the vulnerability quickly and efficiently. This paper introduces the theoretical knowledge of directed fuzzy testing technology, and several state-of-the-art fuzzy testing tools to elaborate the principle, advantages, disadvantages and the prospect of directed fuzzy technology.

A Review of Machine Learning Applications in Fuzzing

Paper

Abstract: Fuzzing has played an important role in improving software development and testing over the course of several decades. Recent research in fuzzing has focused on applications of machine learning (ML), offering useful tools to overcome challenges in the fuzzing process. This review surveys the current research in applying ML to fuzzing. Specifically, this review discusses successful applications of ML to fuzzing, briefly explores challenges encountered, and motivates future research to address fuzzing bottlenecks.

A systematic review of fuzzing based on machine learning techniques

Paper

Abstract: Security vulnerabilities play a vital role in network security system. Fuzzing technology is widely used as a vulnerability discovery technology to reduce damage in advance. However, traditional fuzzing techniques have many challenges, such as how to mutate input seed files, how to increase code coverage, and how to effectively bypass verification. Machine learning technology has been introduced as a new method into fuzzing test to alleviate these challenges. This paper reviews the research progress of using machine learning technology for fuzzing test in recent years, analyzes how machine learning improve the fuzz process and results, and sheds light on future work in fuzzing. Firstly, this paper discusses the reasons why machine learning techniques can be used for fuzzing scenarios and identifies six different stages in which machine learning have been used. Then this paper systematically study the machine learning based fuzzing models from selection of machine learning algorithm, pre-processing methods, datasets, evaluation metrics, and hyperparameters setting. Next, this paper assesses the performance of the machine learning models based on the frequently used evaluation metrics. The results of the evaluation prove that machine learning technology has an acceptable capability of categorize predictive for fuzzing. Finally, the comparison on capability of discovering vulnerabilities between traditional fuzzing tools and machine learning based fuzzing tools is analyzed. The results depict that the introduction of machine learning technology can improve the performance of fuzzing. However, there are still some limitations, such as unbalanced training samples and difficult to extract the characteristics related to vulnerabilities.

The Art, Science, and Engineering of Fuzzing: A Survey

Abstract: Among the many software testing techniques available today, fuzzing has remained highly popular due to its conceptual simplicity, its low barrier to deployment, and its vast amount of empirical evidence in discovering real-world software vulnerabilities. At a high level, fuzzing refers to a process of repeatedly running a program with generated inputs that may be syntactically or semantically malformed. While researchers and practitioners alike have invested a large and diverse effort towards improving fuzzing in recent years, this surge of work has also made it difficult to gain a comprehensive and coherent view of fuzzing. To help preserve and bring coherence to the vast literature of fuzzing, this paper presents a unified, general-purpose model of fuzzing together with a taxonomy of the current fuzzing literature. We methodically explore the design decisions at every stage of our model fuzzer by surveying the related literature and innovations in the art, science, and engineering that make modern-day fuzzers effective.

Fuzzing: Art, Science, and Engineering

Paper

Abstract: Among the many software vulnerability discovery techniques available today, fuzzing has remained highly popular due to its conceptual simplicity, its low barrier to deployment, and its vast amount of empirical evidence in discovering real-world software vulnerabilities. At a high level, fuzzing refers to a process of repeatedly running a program with generated inputs that may be syntactically or semantically malformed. While researchers and practitioners alike have invested a large and diverse effort towards improving fuzzing in recent years, this surge of work has also made it difficult to gain a comprehensive and coherent view of fuzzing. To help preserve and bring coherence to the vast literature of fuzzing, this paper presents a unified, general-purpose model of fuzzing together with a taxonomy of the current fuzzing literature. We methodically explore the design decisions at every stage of our model fuzzer by surveying the related literature and innovations in the art, science, and engineering that make modern-day fuzzers effective.

Fuzzing: a survey

Abstract: Security vulnerability is one of the root causes of cyber-security threats. To discover vulnerabilities and fix them in advance, researchers have proposed several techniques, among which fuzzing is the most widely used one. In recent years, fuzzing solutions, like AFL, have made great improvements in vulnerability discovery. This paper presents a summary of the recent advances, analyzes how they improve the fuzzing process, and sheds light on future work in fuzzing. Firstly, we discuss the reason why fuzzing is popular, by comparing different commonly used vulnerability discovery techniques. Then we present an overview of fuzzing solutions, and discuss in detail one of the most popular type of fuzzing, i.e., coverage-based fuzzing. Then we present other techniques that could make fuzzing process smarter and more efficient. Finally, we show some applications of fuzzing, and discuss new trends of fuzzing and potential future directions.

Fuzzing: State of the art

Abstract: As one of the most popular software testing techniques, fuzzing can find a variety of weaknesses in a program, such as software bugs and vulnerabilities, by generating numerous test inputs. Due to its effectiveness, fuzzing is regarded as a valuable bug hunting method. In this paper, we present an overview of fuzzing that concentrates on its general process, as well as classifications, followed by detailed discussion of the key obstacles and some state-of-the-art technologies which aim to overcome or mitigate these obstacles. We further investigate and classify several widely used fuzzing tools. Our primary goal is to equip the stakeholder with a better understanding of fuzzing and the potential solutions for improving fuzzing methods in the spectrum of software testing and security. To inspire future research, we also predict some future directions with regard to fuzzing.

Fuzzing: A Survey for Roadmap

Paper

Abstract: Fuzz testing (fuzzing) has witnessed its prosperity in detecting security flaws recently. It generates a large number of test cases and monitors the executions for defects. Fuzzing has detected thousands of bugs and vulnerabilities in various applications. Although effective, there lacks systematic analysis of gaps faced by fuzzing. As a technique of defect detection, fuzzing is required to narrow down the gaps between the entire input space and the defect space. Without limitation on the generated inputs, the input space is infinite. However, defects are sparse in an application, which indicates that the defect space is much smaller than the entire input space. Besides, because fuzzing generates numerous test cases to repeatedly examine targets, it requires fuzzing to perform in an automatic manner. Due to the complexity of applications and defects, it is challenging to automatize the execution of diverse applications. In this paper, we systematically review and analyze the gaps as well as their solutions, considering both breadth and depth. This survey can be a roadmap for both beginners and advanced developers to better understand fuzzing.

Survey of Software Fuzzing Techniques

Paper

Abstract: As cybersecurity becomes more than an afterthought and receives the attention it deserves, it is critical to use tools that examine and understand the flaws within programs. One such tool is fuzzing, a testing process which subjects a system or program to a stream of input data, where the goal of the test is to check for exploitable edge cases. This testing is vital, since a hacker may bombard a system with a variety of inputs and scan the system for weaknesses after causing the system to fail. As the security landscape is constantly shifting, there is a need for an up to date review of the state of literature so readers can make informed decisions about the current state of their systems and software. This survey does that by summarizing current state-of-the art fuzzing approaches, classifying these approaches, and highlighting key insights into the current state of research. The paper also identifies current challenges and suggests future research directions in this area.

A Review of Fuzzing Tools and Methods

Paper

Abstract: This paper reviewed some of the most noteworthy academic literature and practical work that has been produced in the field of fuzzing. We first examined how vulnerabilities come to exist in software and how security researchers find them. After a brief overview of common vulnerability types and methods of static analysis, we looked in depth at the field of fuzzing. Competing approaches to fuzzing were examined, from simple random inputs all the way to using genetic algorithms and taint analysis. The importance of measuring code coverage to evaluate the completeness of a fuzzing campaign was examined. Finally, the focus was placed on the fuzz testing of web browsers and the specific tools and techniques related to that.

Embedded fuzzing: a review of challenges, tools, and solutions

Paper

Abstract: Fuzzing has become one of the best-established methods to uncover software bugs. Meanwhile, the market of embedded systems, which binds the software execution tightly to the very hardware architecture, has grown at a steady pace, and that pace is anticipated to become yet more sustained in the near future. Embedded systems also benefit from fuzzing, but the innumerable existing architectures and hardware peripherals complicate the development of general and usable approaches, hence a plethora of tools have recently appeared. Here comes a stringent need for a systematic review in the area of fuzzing approaches for embedded systems, which we term “embedded fuzzing” for brevity. The inclusion criteria chosen in this article are semi-objective in their coverage of the most relevant publication venues as well as of our personal judgement. The review rests on a formal definition we develop to represent the realm of embedded fuzzing. It continues by discussing the approaches that satisfy the inclusion criteria, then defines the relevant elements of comparison and groups the approaches according to how the execution environment is served to the system under test. The resulting review produces a table with 42 entries, which in turn supports discussion suggesting vast room for future research due to the limitations noted.

Differential Fuzzing

JIT-Picking: Differential Fuzzing of JavaScript Engines (CCS 2022)

Paper

Abstract: Modern JavaScript engines that power websites and even full applications on the Web are driven by the need for an increasingly fast and snappy user experience. These engines use several complex and potentially error-prone mechanisms to optimize their performance. Unsurprisingly, the inevitable complexity results in a huge attack surface and varioustypes of software vulnerabilities. On the defender's side, fuzz testing has proven to be an invaluable tool for uncovering different kinds of memory safety violations. Although it is difficult to test interpreters and JIT compilers in an automated way, recent proposals for input generation based on grammars or target-specific intermediate representations helped uncovering many software faults. However, subtle logic bugs and miscomputations that arise from optimization passes in JIT engines continue to elude state-of-the-art testing methods. While such flaws might seem unremarkable at first glance, they are often still exploitable in practice. In this paper, we propose a novel technique for effectively uncovering this class of subtle bugs during fuzzing. The key idea is to take advantage of the tight coupling between a JavaScript engine's interpreter and its corresponding JIT compiler as a domain-specific and generic bug oracle, which in turn yields a highly sensitive fault detection mechanism. We have designed and implemented a prototype of the proposed approach in a tool called JIT-Picker. In an empirical evaluation, we show that our method enables us to detect subtle software faults that prior work missed. In total, we uncovered 32 bugs that were not publicly known and received a $10.000 bug bounty from Mozilla as a reward for our contributions to JIT engine security.

SpecDoctor: Differential Fuzz Testing to Find Transient Execution Vulnerabilities

Paper

Abstract: Transient execution vulnerabilities have critical security impacts to software systems since those break the fundamental security assumptions guaranteed by the CPU. Detecting these critical vulnerabilities in the RTL development stage is particularly important, as it offers a chance to fix the vulnerability early before reaching the chip manufacturing stage.

This paper proposes SpecDoctor, an automated RTL fuzzer to discover transient execution vulnerabilities in the CPU. To be specific, SpecDoctor designs a fuzzing template, allowing it to test all different scenarios of transient execution vulnerabilities (e.g., Meltdown, Spectre, ForeShadow, etc.) with a single template. Then SpecDoctor performs a multi-phased fuzzing, where each phase is dedicated to solve an individual vulnerability constraint in the RTL context, thereby effectively finding the vulnerabilities.

We implemented and evaluated SpecDoctor on two out-of-order RISC-V CPUs, Boom and NutShell-Argo. During the evaluation, SpecDoctor found transient-execution vulnerabilities which share the similar attack vectors as the previous works. Furthermore, SpecDoctor found two interesting variants which abuse unique attack vectors: Boombard, and Birgus. Boombard exploits an unknown implementation bug in RISC-V Boom, exacerbating it into a critical transient execution vulnerability. Birgus launches a Spectre-type attack with a port contention side channel in NutShell CPU, which is constructed using a unique combination of instructions. We reported the vulnerabilities, and both are confirmed by the developers, illustrating the strong practical impact of SpecDoctor.

SEDiff: Scope-Aware Differential Fuzzing to Test Internal Function Models in Symbolic Execution (FSE 2022)

Paper

Abstract: Symbolic execution has become a foundational program analysis technique. Performing symbolic execution unavoidably encounters internal functions (e.g., library functions) that provide basic operations such as string processing. Many symbolic execution engines construct internal function models that abstract function behaviors for scalability and compatibility concerns. Due to the high complexity of constructing the models, developers intentionally summarize only partial behaviors of a function, namely modeled functionalities, in the models. The correctness of the internal function models is critical because it would impact all applications of symbolic execution, e.g., bug detection and model checking.

A naive solution to testing the correctness of internal function models is to cross-check whether the behaviors of the models comply with their corresponding original function implementations. However, such a solution would mostly detect overwhelming inconsistencies concerning the unmodeled functionalities, which are out of the scope of models and thus considered false reports. We argue that a reasonable testing approach should target only the functionalities that developers intend to model. While being necessary, automatically identifying the modeled functionalities, i.e., the scope, is a significant challenge.

In this paper, we propose a scope-aware differential testing framework, SEDiff, to tackle this problem. We design a novel algorithm to automatically map the modeled functionalities to the code in the original implementations. SEDiff then applies scope-aware grey-box differential fuzzing to relevant code in the original implementations. It also equips a new scope-aware input generator and a tailored bug checker that efficiently and correctly detect erroneous inconsistencies. We extensively evaluated SEDiff on several popular real-world symbolic execution engines targeting binary, web and kernel. Our manual investigation shows that SEDiff precisely identifies the modeled functionalities and detects 46 new bugs in the internal function models used in the symbolic execution engines.

T-Reqs: HTTP Request Smuggling with Differential Fuzzing (CCS 2021)

Paper

Abstract: HTTP Request Smuggling (HRS) is an attack that exploits the HTTP processing discrepancies between two servers deployed in a proxy-origin configuration, allowing attackers to smuggle hidden requests through the proxy. While this idea is not new, HRS is soaring in popularity due to recently revealed novel exploitation techniques and real-life abuse scenarios. In this work, we step back from the highly-specific exploits hogging the spotlight, and present the first work that systematically explores HRS within a scientific framework. We design an experiment infrastructure powered by a novel grammar-based differential fuzzer, test 10 popular server/proxy/CDN technologies in combinations, identify pairs that result in processing discrepancies, and discover exploits that lead to HRS. Our experiment reveals previously unknown ways to manipulate HTTP requests for exploitation, and for the first time documents the server pairs prone to HRS.

CatchBackdoor: Backdoor Testing by Critical Trojan Neural Path Identification via Differential Fuzzing (2021)

Paper

Abstract: The success of deep neural networks (DNNs) in real-world applications has benefited from abundant pre-trained models. However, the backdoored pre-trained models can pose a significant trojan threat to the deployment of downstream DNNs. Existing DNN testing methods are mainly designed to find incorrect corner case behaviors in adversarial settings but fail to discover the backdoors crafted by strong trojan attacks. Observing the trojan network behaviors shows that they are not just reflected by a single compromised neuron as proposed by previous work but attributed to the critical neural paths in the activation intensity and frequency of multiple neurons. This work formulates the DNN backdoor testing and proposes the CatchBackdoor framework. Via differential fuzzing of critical neurons from a small number of benign examples, we identify the trojan paths and particularly the critical ones, and generate backdoor testing examples by simulating the critical neurons in the identified paths. Extensive experiments demonstrate the superiority of CatchBackdoor, with higher detection performance than existing methods. CatchBackdoor works better on detecting backdoors by stealthy blending and adaptive attacks, which existing methods fail to detect. Moreover, our experiments show that CatchBackdoor may reveal the potential backdoors of models in Model Zoo.

Duo: Differential Fuzzing for Deep Learning Operators (IEEE Transactions on Reliability 2021)

Paper

Abstract: Deep learning (DL) libraries reduce the barriers to the DL model construction. In DL libraries, various building blocks are DL operators with different functionality, responsible for processing high-dimensional tensors during training and inference. Thus, the quality of operators could directly impact the quality of models. However, existing DL testing techniques mainly focus on robustness testing of trained neural network models and cannot locate DL operators' defects. The insufficient test input and undetermined test output in operator testing have become challenging for DL library developers. In this article, we propose an approach, namely Duo, which combines fuzzing techniques and differential testing techniques to generate input and evaluate corresponding output. It implements mutation-based fuzzing to produce tensor inputs by employing nine mutation operators derived from genetic algorithms and differential testing to evaluate outputs' correctness from multiple operator instances. Duo is implemented in a tool and used to evaluate seven operators from TensorFlow, PyTorch, MNN, and MXNet in an experiment. The result shows that Duo can expose defects of DL operators and realize multidimension evaluation for DL operators from different DL libraries.

DiFuzzRTL: Differential Fuzz Testing to Find CPU Bug (S&P 2021)

Code

Abstract: DifuzzRTL is a differential fuzz testing approach for CPU verification. We introduce new coverage metric, register-coverage, which comprehensively captures the states of an RTL design and correctly guides the input generation. DifuzzRTL automatically instruments register-coverage, randomly generates and mutates instructions defined in ISA, then cross-check against an ISA simulator to detect bugs.

DPIFuzz: A Differential Fuzzing Framework to Detect DPI Elusion Strategies for QUIC (ACSAC 2020)

Paper

Abstract: QUIC is an emerging transport protocol that has the potential to replace TCP in the near future. As such, QUIC will become an important target for Deep Packet Inspection (DPI). Reliable DPI is essential, e.g., for corporate environments, to monitor traffic entering and leaving their networks. However, elusion strategies threaten the validity of DPI systems, as they allow attackers to carefully design traffic to fool and thus evade on-path DPI systems. While such elusion strategies for TCP are well documented, it is unclear if attackers will be able to elude QUIC-based DPI systems. In this paper, we systematically explore elusion methodologies for QUIC. To this end, we present DPIFuzz: a differential fuzzing framework which can automatically detect strategies to elude stateful DPI systems for QUIC. We use DPIFuzz to generate and mutate QUIC streams in order to compare (and find differences in) the server-side interpretations of five popular open-source QUIC implementations. We show that DPIFuzz successfully reveals DPI elusion strategies, such as using packets with duplicate packet numbers or exploiting the diverging handling of overlapping stream offsets by QUIC implementations. DPIFuzz additionally finds four security-critical vulnerabilities in these QUIC implementations.

DifFuzz: Differential Fuzzing for Side-Channel Analysis (ICSE 2019)

Abstract: Side-channel attacks allow an adversary to uncover secret program data by observing the behavior of a program with respect to a resource, such as execution time, consumed memory or response size. Side-channel vulnerabilities are difficult to reason about as they involve analyzing the correlations between resource usage over multiple program paths. We present DifFuzz, a fuzzing-based approach for detecting side-channel vulnerabilities related to time and space. DifFuzz automatically detects these vulnerabilities by analyzing two versions of the program and using resource-guided heuristics to find inputs that maximize the difference in resource consumption between secret-dependent paths. The methodology of DifFuzz is general and can be applied to programs written in any language. For this paper, we present an implementation that targets analysis of Java programs, and uses and extends the Kelinci and AFL fuzzers. We evaluate DifFuzz on a large number of Java programs and demonstrate that it can reveal unknown side-channel vulnerabilities in popular applications. We also show that DifFuzz compares favorably against Blazer and Themis, two state-of-the-art analysis tools for finding side-channels in Java programs.

Deep Differential Testing of JVM Implementations (ICSE 2019)

Paper

Abstract: The Java Virtual Machine (JVM) is the cornerstone of the widely-used Java platform. Thus, it is critical to ensure the reliability and robustness of popular JVM implementations. However, little research exists on validating production JVMs. One notable effort is classfuzz, which mutates Java bytecode syntactically to stress-test different JVMs. It is shown that classfuzz mainly produces illegal bytecode files and uncovers defects in JVMs' startup processes. It remains a challenge to effectively test JVMs' bytecode verifiers and execution engines to expose deeper bugs.

This paper tackles this challenge by introducing classming, a novel, effective approach to performing deep, differential JVM testing. The key of classming is a technique, live bytecode mutation, to generate, from a seed bytecode file f, likely valid, executable (live) bytecode files: (1) capture the seed f's live bytecode, the sequence of its executed bytecode instructions; (2) repeatedly manipulate the control- and data-flow in f's live bytecode to generate semantically different mutants; and (3) selectively accept the generated mutants to steer the mutation process toward live, diverse mutants. The generated mutants are then employed to differentially test JVMs.

We have evaluated classming on mainstream JVM implementations, including OpenJDK's HotSpot and IBM's J9, by mutating the DaCapo benchmarks. Our results show that classming is very effective in uncovering deep JVM differences. More than 1,800 of the generated classes exposed JVM differences, and more than 30 triggered JVM crashes. We analyzed and reported the JVM runtime differences and crashes, of which 14 have already been confirmed/fixed, including a highly critical security vulnerability in J9 that allowed untrusted code to disable the security manager and elevate its privileges (CVE-2017-1376).

Hunting for bugs in code coverage tools via randomized differential testing (ICSE 2019)

Abstract: Reliable code coverage tools are critically important as it is heavily used to facilitate many quality assurance activities, such as software testing, fuzzing, and debugging. However, little attention has been devoted to assessing the reliability of code coverage tools. In this study, we propose a randomized differential testing approach to hunting for bugs in the most widely used C code coverage tools. Specifically, by generating random input programs, our approach seeks for inconsistencies in code coverage reports produced by different code coverage tools, and then identifies inconsistencies as potential code coverage bugs. To effectively report code coverage bugs, we addressed three specific challenges: (1) How to filter out duplicate test programs as many of them triggering the same bugs in code coverage tools; (2) how to automatically reduce large test programs to much smaller ones that have the same properties; and (3) how to determine which code coverage tools have bugs? The extensive evaluations validate the effectiveness of our approach, resulting in 42 and 28 confirmed/fixed bugs for gcov and llvm-cov, respectively. This case study indicates that code coverage tools are not as reliable as it might have been envisaged. It not only demonstrates the effectiveness of our approach, but also highlights the need to continue improving the reliability of code coverage tools. This work opens up a new direction in code coverage validation which calls for more attention in this area.

Different is Good: Detecting the Use of Uninitialized Variables through Differential Replay (CCS 2019)

Abstract: The use of uninitialized variables is a common issue. It could cause kernel information leak, which defeats the widely deployed security defense, i.e., kernel address space layout randomization (KASLR). Though a recent system called Bochspwn Reloaded reported multiple memory leaks in Windows kernels, how to effectively detect this issue is still largely behind.

In this paper, we propose a new technique, i.e., differential replay, that could effectively detect the use of uninitialized variables. Specifically, it records and replays a program's execution in multiple instances. One instance is with the vanilla memory, the other one changes (or poisons) values of variables allocated from the stack and the heap. Then it compares program states to find references to uninitialized variables. The idea is that if a variable is properly initialized, it will overwrite the poisoned value and program states in two running instances should be the same. After detecting the differences, our system leverages the symbolic taint analysis to further identify the location where the variable was allocated. This helps us to identify the root cause and facilitate the development of real exploits. We have implemented a prototype called TimePlayer. After applying it to both Windows 7 and Windows 10 kernels (x86/x64), it successfully identified 34 new issues and another 85 ones that had been patched (some of them were publicly unknown.) Among 34 new issues, 17 of them have been confirmed as zero-day vulnerabilities by Microsoft.

Differential Program Analysis with Fuzzing and Symbolic Execution (ASE 2018)

Paper

Abstract: Differential program analysis means to identify the behavioral divergences in one or multiple programs, and it can be classified into two categories: identify the behavioral divergences (1) between two program versions for the same input (aka regression analysis), and (2) for the same program with two different inputs (e.g, side-channel analysis). Most of the existent approaches for both subproblems try to solve it with single techniques, which suffer from its weaknesses like scalability issues or imprecision. This research proposes to combine two very strong techniques, namely fuzzing and symbolic execution to tackle these problems and provide scalable solutions for real-world applications. The proposed approaches will be implemented on top of state-of-the-art tools like AFL and Symbolic PathFinder to evaluate them against existent work.

NEZHA: Efficient Domain-Independent Differential Testing (S&P 2017)

Abstract: Differential testing uses similar programs as cross-referencing oracles to find semantic bugs that do not exhibit explicit erroneous behaviors like crashes or assertion failures. Unfortunately, existing differential testing tools are domain-specific and inefficient, requiring large numbers of test inputs to find a single bug. In this paper, we address these issues by designing and implementing NEZHA, an efficient input-format-agnostic differential testing framework. The key insight behind NEZHA's design is that current tools generate inputs by simply borrowing techniques designed for finding crash or memory corruption bugs in individual programs (e.g., maximizing code coverage). By contrast, NEZHA exploits the behavioral asymmetries between multiple test programs to focus on inputs that are more likely to trigger semantic bugs. We introduce the notion of δ-diversity, which summarizes the observed asymmetries between the behaviors of multiple test applications. Based on δ-diversity, we design two efficient domain-independent input generation mechanisms for differential testing, one gray-box and one black-box. We demonstrate that both of these input generation schemes are significantly more efficient than existing tools at finding semantic bugs in real-world, complex software.

Coverage-Directed Differential Testing of JVM Implementations (PLDI 2016)

Paper
Slides

Java virtual machine (JVM) is a core technology, whose reliability is critical. Testing JVM implementations requires painstaking effort in designing test classfiles (*.class) along with their test oracles. An alternative is to employ binary fuzzing to differentially test JVMs by blindly mutating seeding classfiles and executing the resulting mutants on different JVMs for revealing inconsistent behaviors. However, this blind approach is not cost effective in practice because (1) most of the mutants are invalid and redundant, and (2) the discovered JVM discrepancies, if any, mostly only concern compatibility issues, rather than actual defects. This paper tackles this challenge by introducing classfuzz, a coverage-directed fuzzing approach that focuses on representative classfiles for differential JVM testing. Our core insight is to (1) mutate seeding classfiles using a set of predefined mutation operators and employ Markov Chain Monte Carlo (MCMC) sampling to guide mutator selection, and (2) execute the mutants on a reference JVM implementation and use coverage uniqueness as a discipline for accepting representative ones. The accepted classfiles are used as inputs to differentially test JVMs and find defects. We have implemented classfuzz and conducted an extensive evaluation of it against existing fuzz testing algorithms. Our evaluation results show that classfuzz can enhance the ratio of discrepancy-triggering classfiles from 1.7% to 11.9%. We have also reported 62 defect-indicative discrepancies, along with the test classfiles, to JVM developers. A number of our reported issues have already been confirmed as JVM defects, and some even match recent clarifications and changes to the JVM specification, Java SE 8 Edition.

Evaluate Fuzzing

FIXREVERTER: A Realistic Bug Injection Methodology for Benchmarking Fuzz Testing (USENIX Security2022)

Paper
Code

Abstract: Fuzz testing is an active area of research with proposed improvements published at a rapid pace. Such proposals are assessed empirically: Can they be shown to perform better than the status quo? Such an assessment requires a benchmark of target programs with well-identified, realistic bugs. To ease the construction of such a benchmark, this paper presents FIXREVERTER, a tool that automatically injects realistic bugs in a program. FIXREVERTER takes as input a bugfix pattern which contains both code syntax and semantic conditions. Any code site that matches the specified syntax is undone if the semantic conditions are satisfied, as checked by static analysis, thus (re)introducing a likely bug. This paper focuses on three bugfix patterns, which we call conditional-abort, conditional-execute, and conditional-assign, based on a study of fixes in a corpus of Common Vulnerabilities and Exposures (CVEs). Using FIXREVERTER we have built REVBUGBENCH, which consists of 10 programs into which we have injected nearly 8,000 bugs; the programs are taken from FuzzBench and Binutils, and represent common targets of fuzzing evaluations. We have integrated REVBUGBENCH into the FuzzBench service, and used it to evaluate five fuzzers. Fuzzing performance varies by fuzzer and program, as desired/expected. Overall, 219 unique bugs were reported, 19% of which were detected by just one fuzzer.

On the Reliability of Coverage-Based Fuzzer Benchmarking (ICSE 2022)

Paper

Abstract: In one of the largest studies of measures of fuzzer effectiveness, involving over 13 million lines of code, 10 fuzzers, and 24 CPU years worth of fuzzing campaigns, we identify a \emph{very strong correlation} between the coverage achieved and the number of bugs found by a fuzzer: A fuzzer that covers more code also finds more bugs. Because bug-based benchmarking is expensive and subject to several threats to validity, it might seem reasonable to compare fuzzers in terms of the coverage achieved, and from that derive empirical claims about a fuzzer's superiority at finding bugs.

Curiously enough, however, we find \emph{no strong agreement} on which fuzzer is superior if we compared multiple fuzzers in terms of coverage achieved instead of the number of bugs found. The fuzzer best at achieving coverage, may \emph{not} be best at finding bugs.

Mutation Analysis: Answering the Fuzzing Challenge (2022)

Paper

Abstract: Fuzzing is one of the fastest growing fields in software testing. The idea behind fuzzing is to check the behavior of software against a large number of randomly generated inputs, trying to cover all interesting parts of the input space, while observing the tested software for anomalous behaviour. One of the biggest challenges facing fuzzer users is how to validate software behavior, and how to improve the quality of oracles used. While mutation analysis is the premier technique for evaluating the quality of software test oracles, mutation score is rarely used as a metric for evaluating fuzzer quality. Unless mutation analysis researchers can solve multiple problems that make applying mutation analysis to fuzzing challenging, mutation analysis may be permanently sidelined in one of the most important areas of testing and security research. This paper attempts to understand the main challenges in applying mutation analysis for evaluating fuzzers, so that researchers can focus on solving these challenges.

Evaluating Code Coverage for Kernel Fuzzers via Function Call Graph (Access 2021)

Paper

Abstract: The OS kernel, which has full system privileges, is an attractive attack surface. A kernel fuzzer that targets system calls in fuzzing is a popular tool for discovering kernel bugs that can induce kernel privilege escalation attacks. To the best of our knowledge, the relevance of code coverage, which is obtained by fuzzing, to the system call has not been studied yet. For instance, modern coverage-guided kernel fuzzers, such as Syzkaller, estimate code coverage by comparing the entire set of executed basic blocks (or edges) regardless of the system call relevancy. Our insight is that the system call relevancy could be an essential performance indicator for realizing kernel fuzzing. In this regard, this study aims to assess the system call-related code coverage of kernel fuzzers. For this purpose, we have developed a practical assessment system that leverages the Intel PT and KCOV and assessed the Linux kernel fuzzers, such as Syzkaller, Trinity, and ext4 fuzzer. The experiments on different kernel versions demonstrated that approximately 32,000-47,000 functions are implemented in the Linux kernel, and approximately 9.7-15.2% are related to the system call. Our finding is that fuzzers that achieve higher code coverage in conventional metrics do not execute more basic blocks related to system calls. Thus, we recommend that kernel fuzzers use both system call-related functions and regular basic blocks in coverage metrics to assess fuzzing performance or to improve coverage feedback.

FuzzBench: An Open Fuzzer Benchmarking Platform and Service (FSE 2021)

Paper

Abstract: Fuzzing is a key tool used to reduce bugs in production software. At Google, fuzzing has uncovered tens of thousands of bugs. Fuzzing is also a popular subject of academic research. In 2020 alone, over 120 papers were published on the topic of improving, developing, and evaluating fuzzers and fuzzing techniques. Yet, proper evaluation of fuzzing techniques remains elusive. The community has struggled to converge on methodology and standard tools for fuzzer evaluation.

To address this problem, we introduce FuzzBench as an opensource turnkey platform and free service for evaluating fuzzers. It aims to be easy to use, fast, reliable, and provides reproducible experiments. Since its release in March 2020, FuzzBench has been widely used both in industry and academia, carrying out more than 150 experiments for external users. It has been used by several published and in-the-work papers from academic groups, and has had real impact on the most widely used fuzzing tools in industry. The presented case studies suggest that FuzzBench is on its way to becoming a standard fuzzer benchmarking platform.

An Empirical Study of OSS-Fuzz Bugs (MSR 2021)

Paper

Abstract: Continuous fuzzing is an increasingly popular technique for automated quality and security assurance. Google maintains OSS-Fuzz: a continuous fuzzing service for open source software. We conduct the first empirical study of OSS-Fuzz, analyzing 23,907 bugs found in 316 projects. We examine the characteristics of fuzzer-found faults, the lifecycles of such faults, and the evolution of fuzzing campaigns over time. We find that OSS-Fuzz is often effective at quickly finding bugs, and developers are often quick to patch them. However, flaky bugs, timeouts, and out of memory errors are problematic, people rarely file CVEs for security vulnerabilities, and fuzzing campaigns often exhibit punctuated equilibria, where developers might be surprised by large spikes in bugs found. Our findings have implications on future fuzzing research and practice.

Industrial Oriented Evaluation of Fuzzing Techniques (ICST 2021)

Paper

Abstract: Fuzzing is a promising method for discovering vulnerabilities. Recently, various techniques are developed to improve the efficiency of fuzzing, and impressive gains are observed in evaluation results. However, evaluation is complex, as many factors affect the results, for example, test suites, baseline and metrics. Even more, most experiment setups are lab-oriented, lacking industrial settings such as large code-base and parallel runs. The correlation between the academic evaluation results and the bug-finding ability in real industrial settings has not been sufficiently studied.

In this paper, we test representative fuzzing techniques to reveal their efficiency in industrial settings. First, we apply typical fuzzers on academic widely used small projects from LAVAM suite. We also apply the same fuzzers on large practical projects from Google's fuzzer-test-suite, which is rarely used in academic settings. Both experiments are performed in both single and parallel run. By analyzing the results, we found that most optimizations working well on LAVA-M suite fail to achieve satisfying results on Google's fuzzer-test-suite (e.g. compared to AFL, QSYM detects 82x more synthesized bugs in LAVA-M, but only detects 26% real bugs in Google's fuzzer-test-suite), and the original AFL even outperforms most academic optimization variants in industry widely used parallel runs (e.g. AFL covers 13% more paths than AFLFast). Then, we summarize common pitfalls of those optimizations, analyze the corresponding root causes, and propose potential directions such as orchestrations and synchronization to overcome the problems. For example, when running in parallel on those large practical projects, the proposed horizontal orchestration could cover 36%-82% more paths, and discover 46%-150% more unique crashes or bugs, compared to fuzzers such as AFL, FairFuzz and QSYM.

UNIFUZZ: A Holistic and Pragmatic Metrics-Driven Platform for Evaluating Fuzzers (USENIX Security2021)

Paper

Abstract: A flurry of fuzzing tools (fuzzers) have been proposed in the literature, aiming at detecting software vulnerabilities effectively and efficiently. To date, it is however still challenging to compare fuzzers due to the inconsistency of the benchmarks, performance metrics, and/or environments for evaluation, which buries the useful insights and thus impedes the discovery of promising fuzzing primitives. In this paper, we design and develop UNIFUZZ, an open-source and metrics-driven platform for assessing fuzzers in a comprehensive and quantitative manner. Specifically, UNIFUZZ to date has incorporated 35 usable fuzzers, a benchmark of 20 real-world programs, and six categories of performance metrics. We first systematically study the usability of existing fuzzers, find and fix a number of flaws, and integrate them into UNIFUZZ. Based on the study, we propose a collection of pragmatic performance metrics to evaluate fuzzers from six complementary perspectives. Using UNIFUZZ, we conduct in-depth evaluations of several prominent fuzzers including AFL [1], AFLFast [2], Angora [3], Honggfuzz [4], MOPT [5], QSYM [6], T-Fuzz [7] and VUzzer64 [8]. We find that none of them outperforms the others across all the target programs, and that using a single metric to assess the performance of a fuzzer may lead to unilateral conclusions, which demonstrates the significance of comprehensive metrics. Moreover, we identify and investigate previously overlooked factors that may significantly affect a fuzzer's performance, including instrumentation methods and crash analysis tools. Our empirical results show that they are critical to the evaluation of a fuzzer. We hope that our findings can shed light on reliable fuzzing evaluation, so that we can discover promising fuzzing primitives to effectively facilitate fuzzer designs in the future.

My Fuzzer Beats Them All! Developing a Framework for Fair Evaluation and Comparison of Fuzzers (2021)

Paper

Abstract: Fuzzing has become one of the most popular techniques to identify bugs in software. To improve the fuzzing process, a plethora of techniques have recently appeared in academic literature. However, evaluating and comparing these techniques is challenging as fuzzers depend on randomness when generating test inputs. Commonly, existing evaluations only partially follow best practices for fuzzing evaluations. We argue that the reason for this are twofold. First, it is unclear if the proposed guidelines are necessary due to the lack of comprehensive empirical data in the case of fuzz testing. Second, there does not yet exist a framework that integrates statistical evaluation techniques to enable fair comparison of fuzzers. To address these limitations, we introduce a novel fuzzing evaluation framework called SENF (Statistical EvaluatioN of Fuzzers). We demonstrate the practical applicability of our framework by utilizing the most wide-spread fuzzer AFL as our baseline fuzzer and exploring the impact of different evaluation parameters (e.g., the number of repetitions or run-time), compilers, seeds, and fuzzing strategies. Using our evaluation framework, we show that supposedly small changes of the parameters can have a major influence on the measured performance of a fuzzer.

A Quantitative Comparison of Covera (AST 2020)

Paper

Abstract: In recent years, many tools have been developed for fuzz testing that generates and executes test cases repeatedly. However, many studies use different fuzzing targets and evaluation criteria and then it is difficult to compare the performance of the existing tools for fuzz testing. Therefore, we prepared a unified collection of fuzzing targets and then compared 8 fuzzers with the benchmark. In comparison, we compared the fuzzers based on the number of execution paths and branch coverage. The result shows that the number of execution paths is significantly different between the fuzzers. On the other hand, the statistical difference is not confirmed between the branch converges of the fuzzers.

Fuzzing: On the Exponential Cost of Vulnerability Discovery (FSE 2020)

Paper

Abstract: We present counterintuitive results for the scalability of fuzzing. Given the same non-deterministic fuzzer, finding the same bugs linearly faster requires linearly more machines. Yet, finding linearly more bugs in the same time requires exponentially more machines. Similarly, with exponentially more machines, we can cover the same code exponentially faster, but uncovered code only linearly faster. In other words, re-discovering the same vulnerabilities (or achieving the same coverage) is cheap but finding new vulnerabilities (or achieving more coverage) is expensive. This holds even under the simplifying assumption of no parallelization overhead.

We derive these observations from over four CPU years worth of fuzzing campaigns involving almost three hundred open source programs, two state-of-the-art greybox fuzzers, four measures of code coverage, and two measures of vulnerability discovery. We provide a probabilistic analysis and conduct simulation experiments to explain this phenomenon.

A Feature-Oriented Corpus for understanding, Evaluating and Improving Fuzz Testing (ASIACCS 2019)

Paper

Abstract: Fuzzing is a promising technique for detecting security vulnerabilities. Newly developed fuzzers are typically evaluated in terms of the number of bugs found on vulnerable programs/binaries. However,existing corpora usually do not capture the features that prevent fuzzers from finding bugs, leading to ambiguous conclusions on the pros and cons of the fuzzers evaluated. A typical example is that Driller detects more bugs than AFL, but its evaluation cannot establish if the advancement of Driller stems from the concolic execution or not, since, for example, its ability in resolving a dataset's magic values is unclear. In this paper, we propose to address the above problem by generating corpora based on search-hampering features. As a proof-of-concept, we have designed FEData, a prototype corpus that currently focuses on four search-hampering features to generate vulnerable programs for fuzz testing. Unlike existing corpora that can only answer "how", FEData can also further answer "why" by exposing (or understanding) the reasons for the identified weaknesses in a fuzzer. The "why" information serves as the key to the improvement of fuzzers.

Be Sensitive and Collaborative: Analyzing Impact of Coverage Metrics in Greybox Fuzzing (RAID 2019)

Paper

Abstract: Coverage-guided greybox fuzzing has become one of the most common techniques for finding software bugs. Coverage metric, which decides how a fuzzer selects new seeds, is an essential parameter of fuzzing and can significantly affect the results. While there are many existing works on the effectiveness of different coverage metrics on software testing, little is known about how different coverage metrics could actually affect the fuzzing results in practice. More importantly, it is unclear whether there exists one coverage metric that is superior to all the other metrics. In this paper, we report the first systematic study on the impact of different coverage metrics in fuzzing. To this end, we formally define and discuss the concept of sensitivity, which can be used to theoretically compare different coverage metrics. We then present several coverage metrics with their variants. We conduct a study on these metrics with the DARPA CGC dataset, the LAVA-M dataset, and a set of real-world applications (a total of 221 binaries). We find that because each fuzzing instance has limited resources (time and computation power), (1) each metric has its unique merit in terms of flipping certain types of branches (thus vulnerability finding) and (2) there is no grand slam coverage metric that defeats all the others. We also explore combining different coverage metrics through cross-seeding, and the result is very encouraging: this pure fuzzing based approach can crash at least the same numbers of binaries in the CGC dataset as a previous approach (Driller) that combines fuzzing and concolic execution. At the same time, our approach uses fewer computing resources.

Study and Comparison of General Purpose Fuzzers

Paper

Abstract: Fuzz testing is a widely used technique for the detection of vulnerabilities whose popularity has led to the development of various tools that do fuzz testing. General-purpose fuzzers work in all domains while some other fuzzers are targeted towards some specific domain. Evaluation of these tools is not an easy task since different fuzzing tools excel in di�erent domains. In this paper, we evaluate 3 such general-purpose fuzzing tools namely libFuzzer, American Fuzzy Lop(AFL) and honggfuzz on 2 metrics, i.e. their bug finding capability and their code coverage. We use the google fuzzer-test-suite which has 24 applications spanning several domains. libFuzzer performs best out of the three in finding memory leaks and out-of-memory related bugs but for other kinds of bugs, all three perform at par. honggfuzz seems to be the best in terms of coverage, though libFuzzer is not far behind, which we believe is because of our runs being of short duration.

Evaluating Fuzz Testing (CCS 2018)

Paper

Abstract: Fuzz testing has enjoyed great success at discovering security critical bugs in real software. Recently, researchers have devoted significant effort to devising new fuzzing techniques, strategies, and algorithms. Such new ideas are primarily evaluated experimentally so an important question is: What experimental setup is needed to produce trustworthy results? We surveyed the recent research literature and assessed the experimental evaluations carried out by 32 fuzzing papers. We found problems in every evaluation we considered. We then performed our own extensive experimental evaluation using an existing fuzzer. Our results showed that the general problems we found in existing experimental evaluations can indeed translate to actual wrong or misleading assessments. We conclude with some guidelines that we hope will help improve experimental evaluations of fuzz testing algorithms, making reported results more robust.

Instrumentation

InstruGuard: Find and Fix Instrumentation Errors for Coverage-based Greybox Fuzzing (ASE 2021)

Paper
Code

Abstract: As one of the most successful methods at vulnerability discovery, coverage-based greybox fuzzing relies on the lightweight compiler-level instrumentation to achieve the finegrained coverage feedback of the target program. Researchers improve it by optimizing the coverage metrics without questioning the correctness of the instrumentation. However, instrumentation errors, including missed instrumentation locations and redundant instrumentation locations, harm the ability of fuzzers. According to our experiments, it is a common and severe problem in various coverage-based greybox fuzzers and at different compiler optimization levels.

In this paper, we design and implement InstruGuard, an open-source and pragmatic platform to find and fix instrumentation errors. It detects instrumentation errors by static analysis on target binaries, and fixes them with a general solution based on binary rewriting. To study the impact of instrumentation errors and test our solutions, we built a dataset of 15 real-world rograms and selected 6 representative fuzzers as targets. We used InstruGuard to check and repair the instrumented binaries with different fuzzers and different compiler optimization options. To evaluate the effectiveness of the repair, we ran the fuzzers with original instrumented programs and the repaired ones, and compared the fuzzing results from aspects of execution paths, line coverage, and real bug findings. The results showed that InstruGuard had corrected the instrumentation errors of different fuzzers and helped to find more bugs in the dataset. Moreover, we discovered one new zero-day vulnerability missed by other fuzzers with fixed instrumentation but without any changes to the fuzzers.

RIFF: Reduced Instruction Footprint for Coverage-Guided Fuzzing (USENIX ATC 2021)

Abstract: Coverage-guided fuzzers use program coverage measurements to explore different program paths efficiently. The coverage pipeline consists of runtime collection and post-execution processing procedures. First, the target program executes instrumentation code to collect coverage information. Then the fuzzer performs an expensive analysis on the collected data, yet most program executions lead to no increases in coverage. Inefficient implementations of these steps significantly reduce the fuzzer's overall throughput.

In this paper, we propose RIFF, a highly efficient program coverage measurement mechanism to reduce fuzzing overhead. For the target program, RIFF moves computations originally done at runtime to instrumentation-time through static program analysis, thus reducing instrumentation code to a bare minimum. For the fuzzer, RIFF processes coverage with different levels of granularity and utilizes vector instructions to improve throughput.

We implement RIFF in state-of-the-art fuzzers such as AFL and MOpt and evaluate its performance on real-world programs in Google's FuzzBench and fuzzer-test-suite. The results show that RIFF improves coverage measurement efficiency of fuzzers by 23× and 6× during runtime collection and post-execution processing, respectively. As a result, the fuzzers complete 147% more executions, and use only 6.53 hours to reach the 24-hour coverage of baseline fuzzers on average.

Hashing Fuzzing: Introducing Input Diversity to Improve Crash Detection (TSE 2021)

Paper

Abstract: The utility of a test set of program inputs is strongly influenced by its diversity and its size. Syntax coverage has become a standard proxy for diversity. Although more sophisticated measures exist, such as proximity of a sample to a uniform distribution, methods to use them tend to be type dependent. We use r-wise hash functions to create a novel, semantics preserving, testability transformation for C programs that we call HashFuzz. Use of HashFuzz improves the diversity of test sets produced by instrumentation-based fuzzers. We evaluate the effect of the HashFuzz transformation on eight programs from the Google Fuzzer Test Suite using four state-of-the-art fuzzers that have been widely used in previous research. We demonstrate pronounced improvements in the performance of the test sets for the transformed programs across all the fuzzers that we used. These include strong improvements in diversity in every case, maintenance or small improvement in branch coverage -- up to 4.8% improvement in the best case, and significant improvement in unique crash detection numbers -- between 28% to 97% increases compared to test sets for untransformed programs.

RetroWrite: Statically Instrumenting COTS Binaries for Fuzzing and Sanitization (S&P 2020)

Abstract: Analyzing the security of closed source binaries is currently impractical for end-users, or even developers who rely on third-party libraries. Such analysis relies on automatic vulnerability discovery techniques, most notably fuzzing with sanitizers enabled. The current state of the art for applying fuzzing or sanitization to binaries is a dynamic binary translation, which has a prohibitive performance overhead. The alternate technique, static binary rewriting, cannot fully recover symbolization information and hence has difficulty modifying binaries to track code coverage for fuzzing or to add security checks for sanitizers.

The ideal solution for binary security analysis would be a static rewriter that can intelligently add the required instrumentation as if it were inserted at compile time. Such instrumentation requires an analysis to statically disambiguate between references and scalars, a problem known to be undecidable in the general case. We show that recovering this information is possible in practice for the most common class of software and libraries: 64-bit, position-independent code. Based on this observation, we develop RetroWrite, binary-rewriting instrumentation to support American Fuzzy Lop (AFL) and Address Sanitizer (ASan), and show that it can achieve compiler-level performance while retaining precision. Binaries rewritten for coverage guided fuzzing using RetroWrite are identical in performance to compiler-instrumented binaries and outperform the default QEMU-based instrumentation by 4.5x while triggering more bugs. Our implementation of binary-only Address Sanitizer is 3x faster than Valgrind's memcheck, the state-of-the-art binary-only memory checker, and detects 80% more bugs in our evaluation.

INSTRCR: Lightweight instrumentation optimization based on coverage-guided fuzz testing (CCET 2019)

Paper

Abstract: In Fuzzing facing binary coverage, the main role of instrumentation is feedback code coverage (in the case of Fuzz for binary, instrumentation can provide coverage information, which plays an important role in guiding the operation of seeds in Fuzz) . The current instrumentation optimization technique mainly relies on the control flow graph (CFG) to select key basic blocks at the basic block level, but the accuracy of this method is not high enough. Considering that the actual path in the actual operation of the binary may be different from the CFG generated in advance, this paper is based on the indirect jump that cannot be accurately analyzed in the CFG, and some of the basic blocks that can be optimized for high-frequency interpolation. According to the algorithm proposed in this paper, The combination of static analysis and dynamic analysis is used to continuously adjust and select key basic block nodes for instrumentation. It is verified by experiments that this kind of instrumentation method can effectively improve the coverage rate and reduce the overhead, and provide effective guidance for Fuzzing, which can effectively reduce the Fuzzer's false negatives.

Full-speed Fuzzing: Reducing Fuzzing Overhead through Coverage-guided Tracing (S&P 2019)

Abstract: Coverage-guided fuzzing is one of the most successful approaches for discovering software bugs and security vulnerabilities. Of its three main components: (1) test case generation, (2) code coverage tracing, and (3) crash triage, code coverage tracing is a dominant source of overhead. Coverage-guided fuzzers trace every test case's code coverage through either static or dynamic binary instrumentation, or more recently, using hardware support. Unfortunately, tracing all test cases incurs significant performance penalties--even when the overwhelming majority of test cases and their coverage information are discarded because they do not increase code coverage. To eliminate needless tracing by coverage-guided fuzzers, we introduce the notion of coverage-guided tracing. Coverage-guided tracing leverages two observations: (1) only a fraction of generated test cases increase coverage, and thus require tracing; and (2) coverage-increasing test cases become less frequent over time. Coverage-guided tracing encodes the current frontier of coverage in the target binary so that it self-reports when a test case produces new coverage--without tracing. This acts as a filter for tracing; restricting the expense of tracing to only coverage-increasing test cases. Thus, coverage-guided tracing trades increased time handling coverage-increasing test cases for decreased time handling non-coverage-increasing test cases. To show the potential of coverage-guided tracing, we create an implementation based on the static binary instrumentor Dyninst called UnTracer. We evaluate UnTracer using eight real-world binaries commonly used by the fuzzing community. Experiments show that after only an hour of fuzzing, UnTracer's average overhead is below 1%, and after 24-hours of fuzzing, UnTracer approaches 0% overhead, while tracing every test case with popular white- and black-box-binary tracers AFL-Clang, AFL-QEMU, and AFL-Dyninst incurs overheads of 36%, 612%, and 518%, respectively. We further integrate UnTracer with the state-of-the-art hybrid fuzzer QSYM and show that in 24-hours of fuzzing, QSYM-UnTracer executes 79% and 616% more test cases than QSYM-Clang and QSYM-QEMU, respectively.

INSTRIM Lightweight Instrumentation for Coverage-guided Fuzzing (NDSS 2018 workshop)

Paper
Code

Abstract: Empowered by instrumentation, coverage-guided fuzzing monitors the program execution path taken by an input, and prioritizes inputs based on their contribution to code coverage. Although instrumenting every basic block ensures full visibility, it slows down the fuzzer and thus the speed of vulnerability discovery. This paper shows that thanks to common program structures (e.g., directed acyclic subgraphs and simple loops) and compiler optimization (e.g., knowledge of incoming edges), it is possible to accurately reconstruct coverage information by instrumenting only a small fraction of basic blocks. Specifically, we formulate the problem as a path differentiation problem on the control flow graph, and propose an efficient algorithm to select basic blocks that need to be instrumented so that different execution paths remain differentiable. We extend AFL to support such CFG-aware instrumentation. Our experiment results confirm that, compared with full instrumentation, our CFG-aware instrumentation only needs to instrument about 20% of basic blocks while offering 1.04-1.78x speedup during fuzzing. Finally, we highlight several technical challenges and promising research directions to further improve instrumentation for fuzzing.

SyzGen: Automated Generation of Syscall Specification of Closed-Source macOS Drivers (CCS 2021)

Paper

Abstract: Kernel drivers are a critical part of the attack surface since they constitute a large fraction of kernel codebase and oftentimes lack proper vetting, especially for those closed-source ones. Unfortunately, the complex input structure and unknown relationships/dependencies among interfaces make them very challenging to understand. Thus, security analysts primarily rely on manual audit for interface recovery to generate meaningful fuzzing test cases. In this paper, we present SyzGen, a first attempt to automate the generation ofbsyscall specifications for closed-source macOS drivers and facilitate interface-aware fuzzing. We leverage two insights to overcome the challenges of binary analysis: (1) iterative refinement of syscall knowledge and (2) extraction and extrapolation of ependencies from a small number of execution traces. We evaluated our approach on 25 targets. The results show that SyzGen can effectively produce high-quality specifications, leading to 34 bugs, including one that attackers can exploit to escalate privilege, and 2 CVEs to date.

IoT or protocols fuzzing

Fuzzing Matter(s): A White Paper for Fuzzing the Matter Protocol

Paper

Abstract: IoT and smart home devices have transformed daily life, consequently raising more and more concerns about security vulnerabilities. Robust security testing methods are essential to fortify devices against potential threats. While dynamic analysis techniques, such as fuzzing, help identify vulnerabilities, some challenges arise due to diverse architectures, communication channels and protocols. Testing directly on devices overcomes difficulties in firmware emulation, but lack of protocol standardisation still poses hurdles. The recently released Matter protocol aims to unify smart home ecosystems, thus also simplifying security testing. In particular, Matter inherits the concept of Cluster from Zigbee in its Data Model. The Data Model clearly defines attributes, commands, status codes and events that could be leveraged to design automated security testing techniques such as fuzzing. This paper proposes the design of a fuzzing framework for Matter-enabled smart home devices. The framew ork employs stateful fuzzing to cover the inherent state-fullness of IoT devices. Such a framework would bestow benefits upon manufacturers, researchers, and end-users.

Crash Report Accumulation During Continuous Fuzzing with CASR (IVMEM 2024)

Abstract: Crash report accumulation is a necessary step during continuous fuzzing. Dynamic software analysis techniques like fuzzing and dynamic symbolic execution generate a large number of crashes for analysis. However, the time and resource constraints often lead to the postponement of fixing some less critical issues, potentially introducing new errors in future releases. Thus, there is a need to distinguish new errors from old ones. We propose a crash accumulation method and implemented it as part of the CASR toolset. We evaluated our approach on crash reports collected from fuzzing results.

No More Companion Apps Hacking but One Dongle: Hub-Based Blackbox Fuzzing of IoT Firmware (MobiSys 2023)

Paper Abstract: Given the massive difficulty in emulating IoT firmware, blackbox fuzzing of IoT devices for vulnerability discovery has become an attractive option. However, existing blackbox IoT fuzzers need much time and tedious effort to reverse engineer the IoT companion app (or manually collect test scripts) of each IoT device, which is unscalable when analyzing many devices. Moreover, fuzzing through a companion app is impeded by the input sanitization inside the app and limited to the manually revealed functions. We notice that IoT devices are typically able to connect a hub using standard wireless protocols (such as ZigBee, Z-Wave, and WiFi). We thus propose a uniform hub-based architecture for fuzzing various IoT devices, without reverse engineering any companion apps. It exploits the messages exchanged between a hub and an IoT device to automatically discover all the functions, and then launches systematic function-oriented message-semantics-guided fuzzing. It avoids sanitization imposed by a companion app. In addition, it conducts device state-sensitive fuzzing, which we find very effective in finding IoT bugs. We implement the system named HubFuzzer. The evaluation shows that HubFuzzer leads to much higher coverage than prior state of the art. We test 21 IoT devices and find 23 zero-day vulnerabilities. Four CVEs have been assigned.

PrIntFuzz: Fuzzing Linux Drivers via Automated Virtual Device Simulation (ISSTA 2022)

Paper

Abstract: Linux drivers share the same address space and privilege with the core of the kernel but have a much larger code base and attack surface. The Linux drivers are not well tested and have weaker security guarantees than the kernel. Missing support from hardware devices, existing fuzzing solutions fail to cover a large portion of the driver code, e.g., the initialization code and interrupt handlers. In this paper, we present PrIntFuzz, an efficient and universal fuzzing framework that can test the overlooked driver code, including the PRobing code and INTerrupt handlers. PrIntFuzz first extracts knowledge from the driver through inter-procedural field-sensitive, path-sensitive, and flow-sensitive static analysis. Then it utilizes the information to build a flexible and efficient simulator, which supports device probing, hardware interrupts emulation and device I/O interception. Lastly, PrIntFuzz applies a multi-dimension fuzzing strategy to explore the overlooked code. We have developed a prototype of PrIntFuzz and successfully simulated 311 virtual PCI (Peripheral Component Interconnect) devices, 472 virtual I2C (Inter-Integrated Circuit) devices, 169 virtual USB (Universal Serial Bus) devices, and found 150 bugs in the corresponding device drivers. We have submitted patches for these bugs to the Linux kernel community, and 59 patches have been merged so far. In a control experiment of Linux 5.10-rc6, PrIntFuzz found 99 bugs, while the state-of-the-art fuzzer only found 50. PrIntFuzz covers 11,968 basic blocks on the latest Linux kernel, while the state-of-the-art fuzzer Syzkaller only covers 2,353 basic blocks.

SnapFuzz: High-Throughput Fuzzing of Network Applications (ISSTA 2022)

Paper

Abstract: In recent years, fuzz testing has benefited from increased computational power and important algorithmic advances, leading to systems that have discovered many critical bugs and vulnerabilities in production software. Despite these successes, not all applications can be fuzzed efficiently. In particular, stateful applications such as network protocol implementations are constrained by their low fuzzing throughput and the need to develop fuzzing harnesses that reset their state and isolate their side effects.

In this paper, we present SnapFuzz, a novel fuzzing framework for network applications. SnapFuzz offers a robust architecture that transforms slow asynchronous network communication into fast synchronous communication, snapshots the target at the latest point at which it is safe to do so, speeds up all file operations by redirecting them to a custom in-memory filesystem, and removes the need for many fragile modifications, such as configuring time delays or writing clean-up scripts, together with several other improvements.

Using SnapFuzz, we fuzzed five popular networking applications: LightFTP, TinyDTLS, Dnsmasq, LIVE555 and Dcmqrscp. We report impressive performance speedups of 62.8x, 41.2x, 30.6x, 24.6x, and 8.4x, respectively, with significantly simpler fuzzing harnesses in all cases. Through its performance advantage, SnapFuzz has also found 12 extra crashes compared to AFLNet in these applications.

Efficient Greybox Fuzzing of Applications in Linux-based IoT Devices via Enhanced User-mode Emulation (ISSTA 2022)

Abstract: Greybox fuzzing has become one of the most effective vulnerability discovery techniques. However, greybox fuzzing techniques cannot be directly applied to applications in IoT devices. The main reason is that executing these applications highly relies on specific system environments and hardware. To execute the applications in Linux based IoT devices, most existing fuzzing techniques use full-system emulation for the purpose of maximizing compatibility. However, compared with user-mode emulation, full-system emulation suffers from great overhead. Therefore, some previous works, such as Firm-AFL, propose to combine full-system emulation and user-mode emulation to speed up the fuzzing process. Despite the attempts of trying to shift the application towards user-mode emulation, no existing technique supports to execute these applications fully in the user-mode emulation.

To address this issue, we propose EQUAFL, which can automatically setup the execution environment to execute embedded application under user-mode emulation. EQUAFL first executes the application under full-system emulation and observe for the key points where the program may get stuck or even crash during user-mode emulation. With the observed information, EQUAFL can migrate the needed environment for user-mode emulation. Then, EQUAFL uses an enhanced user-mode emulation to replay system calls of network, and resource management behaviors to fulfill the needs of the embedded application during its execution.

We evaluate EQUAFL on 70 network applications from different series of IoT devices. The result shows EQUAFL outperforms the state-of-the-arts in fuzzing efficiency (on average, 26 times faster than AFL-QEMU with full-system emulation, 14 times than Firm-AFL). We have also discovered ten vulnerabilities including six CVEs from the tested firmware images.

FuzzUSB: Hybrid Stateful Fuzzing of USB Gadget Stacks (FSE 2022)

Video
Paper

Abstract: Universal Serial Bus (USB) is the de facto protocol supported by peripherals and mobile devices, such as USB thumb drives and smart phones. For many devices, USB Type-C ports are the primary interface for charging, file transfer, audio, video, etc. Accordingly, attackers have exploited different vulnerabilities within USB stacks, compromising host machines via BadUSB attacks or jailbreaking iPhones from USB connections. While there exist fuzzing frameworks dedicated to USB vulnerability discovery, all of them focus on USB host stacks and ignore USB gadget stacks, which enable all the features within modern peripherals and smart devices. In this paper, we propose FuzzUSB, the first fuzzing framework for the USB gadget stack within commodity OS kernels, leveraging static analysis, symbolic execution, and stateful fuzzing. FuzzUSB combines static analysis and symbolic execution to extract internal state machines from USB gadget drivers, and uses them to achieve state-guided fuzzing through multi-channel inputs. We have implemented FuzzUSB upon the syzkaller kernel fuzzer and applied it to the most recent mainline Linux, Android, and FreeBSD kernels. As a result, we have found 34 previously unknown bugs within the Linux and Android kernels, and opened 8 CVEs. Furthermore, compared to the baseline, FuzzUSB has also demonstrated different improvements, including 3× higher code coverage, 50× improved bug-finding efficiency for Linux USB gadget stacks, 2× higher code coverage for FreeBSD USB gadget stacks, and reproducing known bugs that could not be detected by the baseline fuzzers. We believe FuzzUSB provides developers a powerful tool to thwart USB-related vulnerabilities within modern devices and complete the current USB fuzzing scope.

BrakTooth: Causing Havoc on Bluetooth Link Manager via Directed Fuzzing (USENIX SEC 2022)

Slides
Paper

Abstract: In this paper we propose, design and evaluate a systematic directed fuzzing framework to automatically discover implementation bugs in arbitrary Bluetooth Classic (BT) devices. The core of our fuzzer is the first over-the-air approach that takes full control of the BT controller baseband from the host. This enables us to intercept and modify arbitrary packets, as well as to inject packets out-of-order in lower layers of closed-source BT stack, i.e., Link Manager Protocol (LMP) and Baseband. To systematically guide our fuzzing process, we propose an extensible and novel rule-based approach to automatically construct the protocol state machine during normal over-the-air communication. In particular, by writing a simple set of rules to identify protocol messages, we can dynamically construct an abstracted protocol state machine, fuzz packets resulting from a state and validate responses from target devices. As of today, we have fuzzed 13 BT devices from 11 vendors and we have discovered a total of 18 unknown implementation flaws, with 24 common vulnerability exposures (CVEs) assigned. Furthermore, our discoveries were awarded with six bug bounties from certain vendors. Finally, to show the broader applicability of our framework beyond BT, we have extended our approach to fuzz other wireless protocols, which additionally revealed 6 unknown bugs in certain Wi-Fi and BLE Host stacks.

Drifuzz: Harvesting Bugs in Device Drivers from Golden Seeds (USENIX SEC 2022)

Slides
Paper

Abstract: Peripheral hardware in modern computers is typically assumed to be secure and not malicious, and device drivers are implemented in a way that trusts inputs from hardware. However, recent vulnerabilities such as Broadpwn have demonstrated that attackers can exploit hosts through vulnerable peripherals, highlighting the importance of securing the OS-peripheral boundary. In this paper, we propose a hardware-free concolic-augmented fuzzer targeting WiFi and Ethernet drivers, and a technique for generating high-quality initial seeds, which we call golden seeds, that allow fuzzing to bypass difficult code constructs during driver initialization. Compared to prior work using symbolic execution or greybox fuzzing, Drifuzz is more successful at automatically finding inputs that allow network interfaces to be fully initialized, and improves fuzzing coverage by 214% (3.1x) in WiFi drivers and 60% (1.6x) for Ethernet drivers. During our experiments with fourteen PCI and USB network drivers, we find twelve previously unknown bugs, two of which were assigned CVEs.

StateFuzz: System Call-Based State-Aware Linux Driver Fuzzing (USENIX SEC 2022)

Slides
Paper

Abstract: Coverage-guided fuzzing has achieved great success in finding software vulnerabilities. Existing coverage-guided fuzzers generally favor test cases that hit new code, and discard ones that exercise the same code. However, such a strategy is not optimum. A new test case exercising the same code could be better than a previous test case, as it may trigger new program states useful for code exploration and bug discovery.

In this paper, we assessed the limitation of coverage-guided fuzzing solutions and proposed a state-aware fuzzing solution StateFuzz to address this issue. First, we model program states with values of state-variables and utilize static analysis to recognize such variables. Then, we instrument target programs to track such variables' values and infer program state transition at runtime. Lastly, we utilize state information to prioritize test cases that can trigger new states, and apply a three-dimension feedback mechanism to fine-tune the evolutionary direction of coverage-guided fuzzers. We have implemented a prototype of StateFuzz, and evaluated it on Linux upstream drivers and Android drivers. Evaluation results show that StateFuzz is effective at discovering both new code and vulnerabilities. It finds 18 unknown vulnerabilities and 2 known but unpatched vulnerabilities, and reaches 19% higher code coverage and 32% higher state coverage than the state-of-the-art fuzzer Syzkaller.

SNPSFuzzer: A Fast Greybox Fuzzer for Stateful Network Protocols using Snapshots (2022)

Paper

Abstract: Greybox fuzzing has been widely used in stateless programs and has achieved great success. However, most state-of-the-art greybox fuzzers generally have the problems of slow speed and shallow state depth coverage in the process of fuzzing stateful network protocol programs which are able to remember and store details of the interactions. The existing greybox fuzzers for network protocol programs send a series of well-defined prefix sequences of input messages first and then send mutated messages to test the target state of a stateful network protocol. The process mentioned above causes a high time cost. In this paper, we propose SNPSFuzzer, a fast greybox fuzzer for stateful network protocol using snapshots. SNPSFuzzer dumps the context information when the network protocol program is under a specific state and restores it when the state needs to be fuzzed. Furthermore, we design a message chain analysis algorithm to explore more and deeper network protocol states. Our evaluation shows that, compared with the state-of-the-art network protocol greybox fuzzer AFLNET, SNPSFuzzer increases the speed of network protocol fuzzing by 112.0%-168.9% and improves path coverage by 21.4%-27.5% within 24 hours. Moreover, SNPSFuzzer exposes a previously unreported vulnerability in program Tinydtls.

SnapFuzz: An Efficient Fuzzing Framework for Network Applications (2022)

Paper

Abstract: In recent years, fuzz testing has benefited from increased computational power and important algorithmic advances, leading to systems that have discovered many critical bugs and vulnerabilities in production software. Despite these successes, not all applications can be fuzzed efficiently. In particular, stateful applications such as network protocol implementations are constrained by their low fuzzing throughput and the need to develop fuzzing harnesses that reset their state and isolate their side effects. In this paper, we present SnapFuzz, a novel fuzzing framework for network applications. SnapFuzz offers a robust architecture that transforms slow asynchronous network communication into fast synchronous communication based on UNIX domain sockets, speeds up all file operations by redirecting them to an in-memory filesystem, and removes the need for many fragile modifications, such as configuring time delays or writing cleanup scripts, together with several other improvements. Using SnapFuzz, we fuzzed five popular networking applications: LightFTP, Dnsmasq, LIVE555, TinyDTLS and Dcmqrscp. We report impressive performance speedups of 72.4x, 49.7x, 24.8x, 23.9x, and 8.5x, respectively, with significantly simpler fuzzing harnesses in all cases. Through its performance advantage, SnapFuzz has also found 12 previously-unknown crashes in these applications.

State Selection Algorithms and Their Impact on The Performance of Stateful Network Protocol Fuzzing (2021)

Paper

Abstract: The statefulness property of network protocol implementations poses a unique challenge for testing and verification techniques, including Fuzzing. Stateful fuzzers tackle this challenge by leveraging state models to partition the state space and assist the test generation process. Since not all states are equally important and fuzzing campaigns have time limits, fuzzers need effective state selection algorithms to prioritize progressive states over others. Several state selection algorithms have been proposed but they were implemented and evaluated separately on different platforms, making it hard to achieve conclusive findings. In this work, we evaluate an extensive set of state selection algorithms on the same fuzzing platform that is AFLNet, a state-of-the-art fuzzer for network servers. The algorithm set includes existing ones supported by AFLNet and our novel and principled algorithm called AFLNetLegion. The experimental results on the ProFuzzBench benchmark show that (i) the existing state selection algorithms of AFLNet achieve very similar code coverage, (ii) AFLNetLegion clearly outperforms these algorithms in selected case studies, but (iii) the overall improvement appears insignificant. These are unexpected yet interesting findings. We identify problems and share insights that could open opportunities for future research on this topic.

ICS3Fuzzer: A Framework for Discovering Protocol Implementation Bugs in ICS Supervisory Software by Fuzzing (ACSAC 2021)

Paper

Abstract: The supervisory software is widely used in industrial control systems (ICSs) to manage field devices such as PLC controllers. Once compromised, it could be misused to control or manipulate these physical devices maliciously, endangering manufacturing process or even human lives. Therefore, extensive security testing of supervisory software is crucial for the safe operation of ICS. However, fuzzing ICS supervisory software is challenging due to the prevalent use of proprietary protocols. Without the knowledge of the program states and packet formats, it is difficult to enter the deep states for effective fuzzing.

In this work, we present a fuzzing framework to automatically discover implementation bugs residing in the communication protocols between the supervisory software and the field devices. To avoid heavy human efforts in reverse-engineering the proprietary protocols, the proposed approach constructs a state-book based on the readily-available execution trace of the supervisory software and the corresponding inputs. Then, we propose a state selection algorithm to find the protocol states that are more likely to have bugs. Our fuzzer distributes more budget on those interesting states. To quickly reach the interesting states, traditional snapshot-based method does not work since the communication protocols are time sensitive. We address this issue by synchronously managing external events (GUI operations and network traffic) during the fuzzing loop. We have implemented a prototype and used it to fuzz the supervisory software of four popular ICS platforms. We have found 13 bugs and received 3 CVEs, 2 are classified as critical (CVSS3.x score CRITICAL 9.8) and affected 40 different products.

Westworld: Fuzzing-Assisted Remote Dynamic Symbolic Execution of Smart Apps on IoT Cloud Platforms (ACSAC 2021)

Paper

Abstract: Existing symbolic execution typically assumes the analyzer can control the I/O environment and/or access the library code, which, however, is not the case when programs run on a remote proprietary execution environment managed by another party. For example, SmartThings, one of the most popular IoT platforms, is such a cloud-based execution environment. For programmers who write automation applications to be deployed on IoT cloud platforms, it raises significant challenges when they want to systematically test their code and find bugs. We propose fuzzing-assisted remote dynamic symbolic execution, which uses dynamic symbolic execution as backbone and utilizes fuzzing when necessary to automatically test programs running in a remote proprietary execution environment over which the analyzer has little control. As a case study, we enable it for analyzing smart apps running on SmartThings. We have developed a prototype and the evaluation shows that it is effective in testing smart apps and finding bugs.

ProFuzzBench - A Benchmark for Stateful Protocol Fuzzing (ISSTA 2021)

Paper
Code

Abstract: We present a new benchmark (ProFuzzBench) for stateful fuzzing of network protocols. The benchmark includes a suite of representative open-source network servers for popular protocols, and tools to automate experimentation. We discuss challenges and potential directions for future research based on this benchmark.

TCP-Fuzz: Detecting Memory and Semantic Bugs in TCP Stacks with Fuzzing (USENIX ATC 2021)

Paper
Slides

Abstract: TCP stacks provide reliable data transmission in network, and thus they should be correctly implemented and well tested to ensure reliability and security. However, testing TCP stacks is difficult. First, a TCP stack accepts packets and system calls that have dependencies between each other, and thus generating effective test cases is challenging. Second, a TCP stack has various complex state transitions, but existing testing approaches target covering states instead of covering state transitions, and thus their testing coverage is limited. Finally, our study of TCP stack commits shows that 87% of bug-fixing commits are related to semantic bugs (such as RFC violations), but existing bug sanitizers can detect only memory bugs not semantic bugs.

In this paper, we design a novel fuzzing framework named TCP-Fuzz, to effectively test TCP stacks and detect bugs. TCP-Fuzz consists of three key techniques: (1) a dependency-based strategy that considers dependencies between packets and system calls, to generate effective test cases; (2) a transition-guided fuzzing approach that uses a new coverage metric named branch transition as program feedback, to improve the coverage of state transitions; (3) a differential checker that compares the outputs of multiple TCP stacks for the same inputs, to detect semantic bugs. We have evaluated TCP-Fuzz on five widely-used TCP stacks (TLDK, F-Stack, mTCP, FreeBSD TCP and Linux TCP), and find 56 real bugs (including 8 memory bugs and 48 semantic bugs). 40 of these bugs have been confirmed by related developers.

ICPFuzzer: proprietary communication protocol fuzzing by using machine learning and feedback strategies (Cybersecurity 2021)

Paper

Abstract: The fuzzing test is able to discover various vulnerabilities and has more chances to hit the zero-day targets. And ICS(Industrial control system) is currently facing huge security threats and requires security standards, like ISO 62443, to ensure the quality of the device. However, some industrial proprietary communication protocols can be customized and have complicated structures, the fuzzing system cannot quickly generate test data that adapt to various protocols. It also struggles to define the mutation field without having prior knowledge of the protocols. Therefore, we propose a fuzzing system named ICPFuzzer that uses LSTM(Long short-term memory) to learn the features of a protocol and generates mutated test data automatically. We also use the responses of testing and adjust the weight strategies to further test the device under testing (DUT) to find more data that cause unusual connection status. We verified the effectiveness of the approach by comparing with the open-source and commercial fuzzers. Furthermore, in a real case, we experimented with the DLMS/COSEM for a smart meter and found that the test data can cause a unusual response. In summary, ICPFuzzer is a black-box fuzzing system that can automatically execute the testing process and reveal vulnerabilities that interrupt and crash industrial control communication. Not only improves the quality of ICS but also improves safety.

Fuzzing With Optimized Grammar-Aware Mutation Strategies (Access 2021)

Paper

Abstract: Fuzzing is a widely used technique to discover vulnerabilities in software. However, for programs requiring highly structured inputs, the byte-based mutation strategies in existing fuzzers have difficulties in generating valid inputs. To resolve this challenge, Grammar-Based Fuzzing (GBF) utilizes existing grammar specifications to generate new inputs. Some GBFs perform mutation based on Abstract Syntax Trees (ASTs), which can generate inputs conforming to grammars. However, the existing GBFs neglect using feedback to optimize mutation strategies, and blindly generate inputs without considering the effectiveness of those inputs. In this paper, we use the power schedule and the subtree pool to optimize mutation strategies. Specifically, we first translate input files into ASTs, and extract subtrees from ASTs into a subtree pool. Then, we optimize the power schedule on AST nodes based on a probabilistic model. That is, we adaptively determine the time budget for mutating an AST node. Finally, we replace AST nodes along with their subtrees using the ones we select from the subtree pool. We implement a fuzzing tool to demonstrate our strategies. The experiment results show that our method outperforms the state-of-the-art methods in fuzzing efficiency.

FIRM-COV: High-Coverage Greybox Fuzzing for IoT Firmware via Optimized Process Emulation (Access 2021)

Paper

Abstract: With the growing prevalence of the Internet of Things (IoT), related security threats have kept pace. The need to dynamically detect vulnerabilities in IoT devices cannot be overstated. In this work, we present FIRM-COV, the first high coverage-oriented greybox fuzzer for IoT firmware. FIRM-COV leverages newly optimized process emulation by targeting IoT programs and mining real-world vulnerabilities. FIRM-COV focuses on solving problems of IoT fuzzing based on empirical analyses, using the required structured input, the inaccuracy and instability of emulation, and the required high code coverage. By optimizing the existing emulation technique, FIRM-COV always maintains a stable state and achieves high accuracy when detecting vulnerabilities. We also implement a dictionary generation algorithm to provide structured input values and synergy scheduling to achieve high coverage and throughput. We compare FIRM-COV with other IoT fuzzing frameworks for eight real-world IoT devices. As a result, FIRM-COV achieves the highest coverage and throughput, finding the fastest and most 1-day vulnerabilities with almost no false-positives. It also found two 0-day vulnerabilities in real-world IoT devices within 24 h.

DIANE: Identifying Fuzzing Triggers in Apps to Generate Under-constrained Inputs for IoT Devices (S&P 2020)

Abstract: Internet of Things (IoT) devices have rooted themselves in the everyday life of billions of people. Thus, researchers have applied automated bug finding techniques to improve their overall security. However, due to the difficulties in extracting and emulating custom firmware, black-box fuzzing is often the only viable analysis option. Unfortunately, this solution mostly produces invalid inputs, which are quickly discarded by the targeted IoT device and do not penetrate its code. Another proposed approach is to leverage the companion app (i.e., the mobile app typically used to control an IoT device) to generate well-structured fuzzing inputs. Unfortunately, the existing solutions produce fuzzing inputs that are constrained by app-side validation code, thus significantly limiting the range of discovered vulnerabilities. In this paper, we propose a novel approach that overcomes these limitations. Our key observation is that there exist functions inside the companion app that can be used to generate optimal (i.e., valid yet under-constrained) fuzzing inputs. Such functions, which we call fuzzing triggers, are executed before any data-transforming functions (e.g., network serialization), but after the input validation code. Consequently, they generate inputs that are not constrained by app-side sanitization code, and, at the same time, are not discarded by the analyzed IoT device due to their invalid format. We design and develop Diane, a tool that combines static and dynamic analysis to find fuzzing triggers in Android companion apps, and then uses them to fuzz IoT devices automatically. We use Diane to analyze 11 popular IoT devices, and identify 11 bugs, 9 of which are zero days. Our results also show that without using fuzzing triggers, it is not possible to generate bug-triggering inputs for many devices.

Snipuzz: Black-box Fuzzing of IoT Firmware via Message Snippet Inference (CCS 2021)

Paper

Abstract: The proliferation of Internet of Things (IoT) devices has made people's lives more convenient, but it has also raised many security concerns. Due to the difficulty of obtaining and emulating IoT firmware, the black-box fuzzing of IoT devices has become a viable option. However, existing black-box fuzzers cannot form effective mutation optimization mechanisms to guide their testing processes, mainly due to the lack of feedback. It is difficult or even impossible to apply existing grammar-based fuzzing strategies. Therefore, an efficient fuzzing approach with syntax inference is required in the IoT fuzzing domain. To address these critical problems, we propose a novel automatic black-box fuzzing for IoT firmware, termed Snipuzz. Snipuzz runs as a client communicating with the devices and infers message snippets for mutation based on the responses. Each snippet refers to a block of consecutive bytes that reflect the approximate code coverage in fuzzing. This mutation strategy based on message snippets considerably narrows down the search space to change the probing messages. We compared Snipuzz with four state-of-the-art IoT fuzzing approaches, i.e., IoTFuzzer, BooFuzz, Doona, and Nemesys. Snipuzz not only inherits the advantages of app-based fuzzing (e.g., IoTFuzzer, but also utilizes communication responses to perform efficient mutation. Furthermore, Snipuzz is lightweight as its execution does not rely on any prerequisite operations, such as reverse engineering of apps. We also evaluated Snipuzz on 20 popular real-world IoT devices. Our results show that Snipuzz could identify 5 zero-day vulnerabilities, and 3 of them could be exposed only by Snipuzz. All the newly discovered vulnerabilities have been confirmed by their vendors.

Learning-Based Fuzzing of IoT Message Brokers (ICST 2021)

Paper

Abstract: The number of devices in the Internet of Things (IoT) immensely grew in recent years. A frequent challenge in the assurance of the dependability of IoT systems is that components of the system appear as a black box. This paper presents a semi-automatic testing methodology for black-box systems that combines automata learning and fuzz testing. Our testing technique uses stateful fuzzing based on a model that is automatically inferred by automata learning. Applying this technique, we can simultaneously test multiple implementations for unexpected behavior and possible security vulnerabilities.We show the effectiveness of our learning-based fuzzing technique in a case study on the MQTT protocol. MQTT is a widely used publish/subscribe protocol in the IoT. Our case study reveals several inconsistencies between five different MQTT brokers. The found inconsistencies expose possible security vulnerabilities and violations of the MQTT specification.

RiverFuzzRL - an open-source tool to experiment with reinforcement learning for fuzzing (ICST 2021)

Paper

Abstract: Combining fuzzing techniques and reinforcement learning could be an important direction in software testing. However, there is a gap in support for experimentation in this field, as there are no open-source tools to let academia and industry to perform experiments easily. The purpose of this paper is to fill this gap by introducing a new framework, named RiverFuzzRL, on top of our already mature framework for AI-guided fuzzing, River. We provide out-of-the-box implementations for users to choose from or customize for their test target. The work presented here is performed on testing binaries and does not require access to the source code, but it can be easily adapted to other types of software testing as well. We also discuss the challenges faced, opportunities, and factors that are important for performance, as seen in the evaluation.

Vulnerability Detection in SIoT Applications: A Fuzzing Method on their Binaries (IEEE Transactions on Network Science and Engineering 2020)

Paper

Abstract: SIoT enables devices to communicate with each other automatically, which is not reliable when applications in SIoT are vulnerable. To improve the security of SIoT, different techniques have been employed so far, mainly to detect vulnerabilities in SIoT applications. Among the detection techniques, fuzzing is one of the most effective ones that can significantly improve the security of SIoT applications. However, the existing fuzzing methods have three problems. First of all, the schemes to instrument target binaries cause high memory overhead because they instrument at all edges to obtain the coverage information. Moreover, they introduce a severe problem called edge collision, i.e., two different edges are deemed the same during fuzzing. Thirdly, none of the existing fuzzers conduct fuzzing using path coverage because path coverage has high memory overhead. In this paper, we propose BECFuzz to resolve the above three problems. BECFuzz instruments at specific edges, and conducts fuzzing based on both edge coverage and path coverage, which greatly improves its effectiveness. We implement our BECFuzz based on two typical fuzzers which are widely recognised as baselines, AFL and AFLFast, and run experiments on 18 real-world programs. The results demonstrate that our method suppresses the state-of-art fuzzers in performance.

Analysis of DTLS Implementations Using Protocol State Fuzzing (USENIX Security2020)

Paper

Abstract: Recent years have witnessed an increasing number of protocols relying on UDP. Compared to TCP, UDP offers performance advantages such as simplicity and lower latency. This has motivated its adoption in Voice over IP, tunneling technologies, IoT, and novel Web protocols. To protect sensitive data exchange in these scenarios, the DTLS protocol has been developed as a cryptographic variation of TLS. DTLS's main challenge is to support the stateless and unreliable transport of UDP. This has forced protocol designers to make choices that affect the complexity of DTLS, and to incorporate features that need not be addressed in the numerous TLS analyses.

We present the first comprehensive analysis of DTLS implementations using protocol state fuzzing. To that end, we extend TLS-Attacker, an open source framework for analyzing TLS implementations, with support for DTLS tailored to the stateless and unreliable nature of the underlying UDP layer. We build a framework for applying protocol state fuzzing on DTLS servers, and use it to learn state machine models for thirteen DTLS implementations. Analysis of the learned state models reveals four serious security vulnerabilities, including a full client authentication bypass in the latest JSSE version, as well as several functional bugs and non-conformance issues. It also uncovers considerable differences between the models, confirming the complexity of DTLS state machines.

Frankenstein: Advanced Wireless Fuzzing to Exploit New Bluetooth Escalation Targets (USENIX Security2020)

Paper

Abstract: Wireless communication standards and implementations have a troubled history regarding security. Since most implementations and firmwares are closed-source, fuzzing remains one of the main methods to uncover Remote Code Execution (RCE) vulnerabilities in deployed systems. Generic over-the-air fuzzing suffers from several shortcomings, such as constrained speed, limited repeatability, and restricted ability to debug. In this paper, we present Frankenstein, a fuzzing framework based on advanced firmware emulation, which addresses these shortcomings. Frankenstein brings firmware dumps "back to life", and provides fuzzed input to the chip's virtual modem. The speed-up of our new fuzzing method is sufficient to maintain interoperability with the attached operating system, hence triggering realistic full-stack behavior. We demonstrate the potential of Frankenstein by finding three zero-click vulnerabilities in the Broadcom and Cypress Bluetooth stack, which is used in most Apple devices, many Samsung smartphones, the Raspberry Pis, and many others. Given RCE on a Bluetooth chip, attackers may escalate their privileges beyond the chip's boundary. We uncover a Wi-Fi/Bluetooth coexistence issue that crashes multiple operating system kernels and a design flaw in the Bluetooth 5.2 specification that allows link key extraction from the host. Turning off Bluetooth will not fully disable the chip, making it hard to defend against RCE attacks. Moreover, when testing our chip-based vulnerabilities on those devices, we find BlueFrag, a chip-independent Android RCE.

A deep convolution generative adversarial networks based fuzzing framework for industry control protocols

Paper

Abstract: A growing awareness is brought that the safety and security of industrial control systems cannot be dealt with in isolation, and the safety and security of industrial control protocols (ICPs) should be considered jointly. Fuzz testing (fuzzing) for the ICP is a common way to discover whether the ICP itself is designed and implemented with flaws and network security vulnerability. Traditional fuzzing methods promote the safety and security testing of ICPs, and many of them have practical applications. However, most traditional fuzzing methods rely heavily on the specification of ICPs, which makes the test process a costly, time-consuming, troublesome and boring task. And the task is hard to repeat if the specification does not exist. In this study, we propose a smart and automated protocol fuzzing methodology based on improved deep convolution generative adversarial network and give a series of performance metrics. An automated and intelligent fuzzing framework BLSTM-DCNNFuzz for application is designed. Several typical ICPs, including Modbus and EtherCAT, are applied to test the effectiveness and efficiency of our framework. Experiment results show that our methodology outperforms the existing ones like General Purpose Fuzzer and other deep learning based fuzzing methods in convenience, effectiveness, and efficiency.

ICS Protocol Fuzzing: Coverage Guided Packet Crack and Generation (DAC 2020)

Paper

Abstract: Industrial Control System (ICS) protocols play an essential role in building communications among system components. Recently, many severe vulnerabilities, such as Stuxnet and DragonFly, exposed in ICS protocols have affected a wide distribution of devices. Therefore, it is of vital importance to ensure their correctness. However, the vulnerability detection efficiency of traditional techniques such as fuzzing is challenged by the complexity and diversity of the protocols.

In this paper, we propose to equip the traditional protocol fuzzing with coverage-guided packet crack and generation. We collect the coverage information during testing procedure, save those valuable packets that trigger new path coverage and crack them into pieces, based on which, we can construct higher quality new packets for further testing. For evaluation, we build Peach* on top of Peach, which is one of the most widely used protocol fuzzers, and conduct experiments on several ICS protocols such as Modbus and DNP3. Results show that, compared with the original Peach, Peach* achieves the same code coverage and bug detection numbers at the speed of 1.2X-25X. It also gains final increase with 8.35%-36.84% more paths within 24 hours, and has exposed 9 previously unknown vulnerabilities.

AFLNET: A Greybox Fuzzer for Network Protocols (ICST 2020)

Paper
Code

Abstract: Server fuzzing is difficult. Unlike simple command-line tools, servers feature a massive state space that can be traversed effectively only with well-defined sequences of input messages. Valid sequences are specified in a protocol. In this paper, we present AFLNET, the first grey-box fuzzer for protocol implementations. Unlike existing protocol fuzzers, AFLNET takes a mutational approach and uses state-feedback to guide the fuzzing process. AFLNET is seeded with a corpus of recorded message exchanges between the server and an actual client. No protocol specification or message grammars are required. AFLNET acts as a client and replays variations of the original sequence of messages sent to the server and retains those variations that were effective at increasing the coverage of the code or state space. To identify the server states that are exercised by a message sequence, AFLNET uses the server's response codes. From this feedback, AFLNET identifies progressive regions in the state space, and systematically steers towards such regions. The case studies with AFLNET on two popular protocol implementations demonstrate a substantial performance boost over the state-of-the-art. AFLNET discovered two new CVEs that are classified as critical (CVSS score CRITICAL 9.8).

Finding Security Vulnerabilities in Network Protocol Implementations (Arxiv 2020)

Paper

Abstract: Implementations of network protocols are often prone to vulnerabilities caused by developers' mistakes when accessing memory regions and dealing with arithmetic operations. Finding practical approaches for checking the security of network protocol implementations has proven to be a challenging problem. The main reason is that the protocol software state-space is too large to be explored. Here we propose a novel verification approach that combines fuzzing with symbolic execution to verify intricate properties in network protocol implementations. We use fuzzing for an initial exploration of the network protocol, while symbolic execution explores both the program paths and protocol states, which were uncovered by fuzzing. From this combination, we automatically generate high-coverage test input packets for a network protocol implementation.We surveyed various approaches based on fuzzing and symbolic execution to understand how these techniques can be effectively combined and then choose a suitable tool to develop further our model on top of it. In our preliminary evaluation, we used ESBMC, Map2Check, and KLEE as software verifiers and SPIKE as fuzzer to check their suitability to verify our network protocol implementations. Our experimental results show that ESBMC can be further developed within our verification framework called FuSeBMC, to efficiently and effectively detect intricate security vulnerabilities in network protocol implementations.

Smart seed selection-based effective black box fuzzing for IIoT protocol (2020)

Paper

Abstract: Connections of cyber-physical system (CPS) components are gradually increasing owing to the introduction of the Industrial Internet of Things (IIoT). IIoT vulnerability analysis has become a major issue because complex skillful cyber-attacks on CPS systems exploit their zero-day vulnerabilities. However, current white box techniques for vulnerability analysis are difficult to use in real heterogeneous environments, where devices supplied by various manufacturers and diverse firmware versions are used. Therefore, we herein propose a novel protocol fuzzing test technique that can be applied in a heterogeneous environment. As seed configuration can significantly influence the test result in a black box test, we update the seed pool using test cases that travel different program paths compared to the seed. The input, output, and Delta times are used to determine if a new program area has been searched in the black box environment. We experimentally verified the effectiveness of the proposed.

Fw-fuzz: A code coverage-guided fuzzing framework for network protocols on firmware (2020)

Paper

Abstract: Fuzzing is an effective approach to detect software vulnerabilities utilizing changeable generated inputs. However, fuzzing the network protocol on the firmware of IoT devices is limited by inefficiency of test case generation, cross-architecture instrumentation, and fault detection. In this article, we propose the Fw-fuzz, a coverage-guided and crossplatform framework for fuzzing network services running in the context of firmware on embedded architectures, which can generate more valuable test cases by introspecting program runtime information and using a genetic algorithm model. Specifically, we propose novel dynamic instrumentation in Fw-fuzz to collect the running state of the firmware program. Then Fw-fuzz adopts a genetic algorithm model to guide the generation of inputs with high code coverage. We fully implement the prototype system of Fw-fuzz and conduct evaluations on network service programs of various architectures in MIPS, ARM, and PPC. By comparing with the protocol fuzzers Boofuzz and Peach in metrics of edge coverage, our prototype system achieves an average growth of 33.7% and 38.4%, respectively. We further verify six known vulnerabilities and discover 5 0-day vulnerabilities with the Fw-fuzz, which prove the validity and utility of our framework. The overhead of our system expressed as an additional 5% of memory growth.

BaseSAFE: Baseband SAnitized Fuzzing through Emulation (WiSec 2020)

Paper
Code

Abstract: Rogue base stations are an effective attack vector. Cellular basebands represent a critical part of the smartphone's security: they parse large amounts of data even before authentication. They can, therefore, grant an attacker a very stealthy way to gather information about calls placed and even to escalate to the main operating system, over-the-air. In this paper, we discuss a novel cellular fuzzing framework that aims to help security researchers find critical bugs in cellular basebands and similar embedded systems. BaseSAFE allows partial rehosting of cellular basebands for fast instrumented fuzzing off-device, even for closed-source firmware blobs. BaseSAFE's sanitizing drop-in allocator, enables spotting heap-based buffer-overflows quickly. Using our proof-of-concept harness, we fuzzed various parsers of the Nucleus RTOS-based MediaTek cellular baseband that are accessible from rogue base stations. The emulator instrumentation is highly optimized, reaching hundreds of executions per second on each core for our complex test case, around 15k test-cases per second in total. Furthermore, we discuss attack vectors for baseband modems. To the best of our knowledge, this is the first use of emulation-based fuzzing for security testing of commercial cellular basebands. Most of the tooling and approaches of BaseSAFE are also applicable for other low-level kernels and firmware. Using BaseSAFE, we were able to find memory corruptions including heap out-of-bounds writes using our proof-of-concept fuzzing harness in the MediaTek cellular baseband. BaseSAFE, the harness, and a large collection of LTE signaling message test cases will be released open-source upon publication of this paper.

Poster: Fuzzing IoT Firmware via Multi-stage Message Generation (CCS 2019)

Paper

Abstract: In this work, we present IoTHunter, the first grey-box fuzzer for fuzzing stateful protocols in IoT firmware. IoTHunter addresses the state scheduling problem based on a multi-stage message generation mechanism on runtime monitoring of IoT firmware. We evaluate IoTHunter with a set of real-world programs, and the result shows that IoTHunter outperforms black-box fuzzer boofuzz, which has a 2.2x, 2.0x, and 2.5x increase for function coverage, block coverage, and edge coverage, respectively. IoTHunter also found five new vulnerabilities in the firmware of home router Mikrotik, which have been reported to the vendor.

SeqFuzzer: An Industrial Protocol Fuzzing Framework in Deep Learning Perspective (ICST 2019)

Paper

Abstract: Industrial networks are the cornerstone of modern industrial control systems. Performing security checks of industrial communication processes help detect unknown risks and vulnerabilities. Fuzz testing is a widely used method for performing security checks that takes advantage of automation. However, there is a big challenge to carry out security checks on the industrial networks due to the increasing variety and complexity of industrial communication protocols. In this case, existing approaches usually take a long time to model the protocol for generating test cases, which is labor-intensive and timeconsuming. This becomes even worse when the target protocol is stateful. To help in addressing this problem, we employed a deep learning model to learn the structures of protocol frames and deal with the temporal features of stateful protocols. We propose a fuzzing framework named SeqFuzzer which automatically learns the protocol frame structures from communication traffic and generates fake but plausible messages as test cases. For proving the usability of our approach, we applied SeqFuzzer to widelyused Ethernet for Control Automation Technology (EtherCAT) devices and successfully detected several security vulnerabilities

SPFuzz: A Hierarchical Scheduling Framework for Stateful Network Protocol Fuzzing (IEEE Access 2019)

Paper

Abstract: In recent years, the fuzzing technology is widely used to detect the software vulnerabilities owing to the coverage improvement in the target program and the easiness of use. However, it is less efficient to fuzz the stateful protocols due to the difficulties like maintaining states and dependencies of messages. To address these challenges, we present SPFuzz, a framework for building flexible, coverage guided stateful protocol fuzzing. We define a language in SPFuzz to describe the protocol specifications, protocol states transitions and dependencies for generating valuable test cases, maintaining correct messages in session states and handling protocol dependencies by updating message data in time. The SPFuzz adopts a three-level mutation strategy, namely head, content, and sequence mutation strategy to drive the fuzzing process to cover more paths, in conjunction with the method to randomly assign weights to messages and strategies. We use the following metrics to evaluate the performance of SPFuzz and other frameworks upon three protocol implementations, i.e., Proftpd, Oftpd, and OpenSSL, which are three-granularity coverages specifically function, basic block, and edge. In experiments, the SPFuzz framework outperforms the existing stateful protocol fuzzing tool Boofuzz by an average of 69.12% in three granularities coverage tests. This demonstrates that the SPFuzz has the ability to explore more and deeper paths of the target program. We further triggered CVE-2015-0291 in OpenSSL 1.0.2 with the SPFuzz, which proves the validity and utility of our framework.

HFuzz: Towards automatic fuzzing testing of NB-IoT core network protocols implementations (FGCS 2019)

Paper

Abstract: Narrowband Internet of Things (NB-loT) is widely deployed in the cellular network of operators, yet implementations of its core network protocols are suffering from bugs. Due to the complexity of the frame structure of NB-IoT core network protocols, testing the protocols in this field is notoriously difficult. In this paper, we propose a novel fuzzing framework, named HFuzz, to generate a great many high-quality test inputs automatically. HFuzz is an automatic hierarchy-aware fuzzing framework and can allocate computing resources efficiently. We put forward the concept of Message Structure Tree to transform the seed file and generate mutated data of the tested protocols and optimize the resource allocation for each hierarchy of the transformed structure by a novel scheduling algorithm. Therefore HFuzz can get a balance between breadth and depth in finding new paths. Compared to traditional fuzzing tools, HFuzz can easily pass the early verification and induce a better coverage of the target implementations by taking full advantage of format information of NB-IoT core network protocols. Our framework applies to various protocols, and we evaluate the performance of HFuzz on GPRS Tunnelling Protocol version 2(GTPv2) in this paper and conduct experiments with two protocol implementations, Open Air Interface (OAI) and B*(a development system). The experimental results show HFuzz yields higher coverage than American Fuzzy Lop (AFL) and Peach, and we further find a real implementation bug in OAI.

FIRM-AFL: High-Throughput Greybox Fuzzing of IoT Firmware via Augmented Process Emulation (USENIX Security2019)

Abstract: Cyber attacks against IoT devices are a severe threat. These attacks exploit software vulnerabilities in IoT firmware. Fuzzing is an effective software testing technique for vulnerability discovery. In this work, we present FIRM-AFL, the first high-throughput grey box fuzzer for IoT firmware. FIRMAFL addresses two fundamental problems in IoT fuzzing. First, it addresses compatibility issues by enabling fuzzing for POSIX-compatible firmware that can be emulated in a system emulator. Second, it addresses the performance bottleneck caused by system-mode emulation with a novel technique called augmented process emulation. By combining system mode emulation and user-mode emulation in a novel way, augmented process emulation provides high compatibility as system-mode emulation and high throughput as user-mode emulation. Our evaluation results show that (1) FIRM-AFL is fully functional and capable of finding real-world vulnerabilities in IoT programs; (2) the throughput of FIRM-AFL is on average 8.2 times higher than system-mode emulation based fuzzing; and (3) FIRM-AFL is able to find 1-day vulnerabilities much faster than system-mode emulation based fuzzing, and is able to find 0-day vulnerabilities.

Exploring Effective Fuzzing Strategies to Analyze Communication Protocols (FEAST 2019)

Paper

Abstract: In recent years, coverage-based greybox fuzzing has become popular forvulnerability detection due to its simplicity and efficiency. However, it is less powerful when applied directly to protocol fuzzing due to the unique challenges involved in fuzzing communication protocols. In particular, the communication among multiple ends contains more than one packet, which are not necessarily dependent upon each other, i.e., fuzzing single (usually the first) packet can only achieve extremely limited code coverage. In this paper, we study such challenges and demonstrate the limitation of current non-stateful greybox fuzzer. In order to achieve higher code coverage, we design stateful protocol fuzzing strategies for communication protocols to explore the code related to different protocol states. Our approach contains a state switching engine, together with a multi-state forkserver to consistently and flexibly fuzz different states of an compiler-instrumented protocol program. Our experimental results on OpenSSL show that our approach achieves an improvement of 73% more code coverage and 2× unique crashes when comparing against fuzzing the first packet during a protocol handshake.

Leveraging Textual Specifications for Grammar-Based Fuzzing of Network Protocols (AAAI 2019)

Paper

Abstract: Grammar-based fuzzing is a technique used to find software vulnerabilities by injecting well-formed inputs generated following rules that encode application semantics. Most grammar-based fuzzers for network protocols rely on human experts to manually specify these rules. In this work we study automated learning of protocol rules from textual specifications (i.e. RFCs). We evaluate the automatically extracted protocol rules by applying them to a state-of-the-art fuzzer for transport protocols and show that it leads to a smaller number of test cases while finding the same attacks as the system that uses manually specified rules.

MTF-Storm: a high performance fuzzer for Modbus-TCP (ETFA 2018)

Paper
Code

Abstract: MTF-Storm is a highly effective fuzzer for industrial systems employing Modbus/TCP connectivity. It achieves high fault coverage, while offering high performance and quick testing of the System-Under-Test (SUT). Analogously to its predecessor MTF, MTF-Storm operates in 3 phases: a) reconnaissance b) fuzz testing and failure detection. Reconnaissance identifies the memory organization of the SUT and the supported functionality, enabling selection and synthesis of fuzz testing sequences that are effective for the specific SUT. MTF-Storm develops its test sequences systematically, starting with single field tests and proceeding with combined field tests, adopting techniques for automated combinatorial software testing and reducing the test space through partitioning field value ranges. MTF-Storm has been used to evaluate 9 different Modbus/TCP implementations and has identified issues with all of them, ranging from out-of-spec responses to successful denial-of-service attacks and crashes.

Advancing Protocol Fuzzing for Industrial Automation and Control Systems (ICISSP 2018)

Paper

Abstract: Testing for security vulnerabilities is playing an important role in the changing domain of industrial automation and control systems. These systems are increasingly connected to each other via networking technology and are faced with new cyber threats. To improve the security properties of such systems, their robustness must be ensured. Security testing frameworks aim at enabling the assurance of robustness even at the time of development and can play a key role in bringing security into the industrial domain.

Fuzzing describes a technique to discover vulnerabilities in technical systems and is best known from its usage in IT security testing. It uses randomly altered data to provoke unexpected behaviour and can be used in combination with regular unit testing. Combined with the power of fuzzing, the effectiveness of security testing frameworks can be increased.

In this work, different fuzzing tools were evaluated for their properties and then compared with the requirements for an application in the industrial domain. As no fuzzer was fully satisfying these requirements, a new fuzzer, combining the strength of different others, was designed and implemented, and then evaluated. The evaluation includes a real-world application where multiple vulnerabilities in industrial automation components could be identified.

IoTFuzzer: Discovering Memory Corruptions in IoT Through App-based Fuzzing (NDSS 2018)

Abstract: With more IoT devices entering the consumer market, it becomes imperative to detect their security vulnerabilities before an attacker does. Existing binary analysis based approaches only work on firmware, which is less accessible except for those equipped with special tools for extracting the code from the device. To address this challenge in IoT security analysis, we present in this paper a novel automatic fuzzing framework, called IOTFUZZER, which aims at finding memory corruption vulnerabilities in IoT devices without access to their firmware images. The key idea is based upon the observation that most IoT devices are controlled through their official mobile apps, and such an app often contains rich information about the protocol it uses to communicate with its device. Therefore, by identifying and reusing program-specific logic (e.g., encryption) to mutate the test case (particularly message fields), we are able to effectively probe IoT targets without relying on any knowledge about its protocol specifications. In our research, we implemented IOTFUZZER and evaluated 17 real-world IoT devices running on different protocols, and our approach successfully identified 15 memory corruption vulnerabilities (including 8 previously unknown ones).

Bbuzz: A Bit-aware Fuzzing Framework for Network Protocol Systematic Reverse Engineering and Analysis (MCC 2017)

Paper

Abstract: Fuzzing is a critical part of secure software development life-cycle, for finding vulnerabilities, developing exploits, and reverse engineering. This relies on appropriate approaches, tools and frameworks. File and protocol fuzzing is well covered, multiple approaches and implementations exist. Unfortunately, assessed tools do not posses the required capabilities for working with protocols, where constructing bit groups are not byte aligned. In this paper, a systematic approach is proposed and tool prototype developed for the cyber red teaming purposes. In a case study, the developed Bbuzz tool is used to reverse engineer a proprietary NATO Link-1 network protocol allowing to inject rogue airplane tracks into air operations command and control system.

Test Data Generation for Stateful Network Protocol Fuzzing Using a Rule-Based State Machine (2016)

Paper

Abstract: To improve the efficiency and coverage of stateful network protocol fuzzing, this paper proposes a new method, using a rule-based state machine and a stateful rule tree to guide the generation of fuzz testing data. The method first builds a rule-based state machine model as a formal description of the states of a network protocol. This removes safety paths, to cut down the scale of the state space. Then it uses a stateful rule tree to describe the relationship between states and messages, and then remove useless items from it. According to the message sequence obtained by the analysis of paths using the stateful rule tree and the protocol specification, an abstract data model of test case generation is defined. The fuzz testing data is produced by various generation algorithms through filling data in the fields of the data model. Using the rule-based state machine and the stateful rule tree, the quantity of test data can be reduced. Experimental results indicate that our method can discover the same vulnerabilities as traditional approaches, using less test data, while optimizing test data generation and improving test efficiency.

Protocol State Fuzzing of TLS Implementations (USENIX Security2015)

Abstract: We describe a largely automated and systematic analysis of TLS implementations by what we call 'protocol state fuzzing': we use state machine learning to infer state machines from protocol implementations, using only blackbox testing, and then inspect the inferred state machines to look for spurious behaviour which might be an indication of flaws in the program logic. For detecting the presence of spurious behaviour the approach is almost fully automatic: we automatically obtain state machines and any spurious behaviour is then trivial to see. Detecting whether the spurious behaviour introduces exploitable security weaknesses does require manual investigation. Still, we take the point of view that any spurious functionality in a security protocol implementation is dangerous and should be removed.

We analysed both server- and client-side implementations with a test harness that supports several key exchange algorithms and the option of client certificate authentication. We show that this approach can catch an interesting class of implementation flaws that is apparently common in security protocol implementations: in three of the TLS implementations analysed new security flaws were found (in GnuTLS, the Java Secure Socket Extension, and OpenSSL). This shows that protocol state fuzzing is a useful technique to systematically analyse security protocol implementations. As our analysis of different TLS implementations resulted in different and unique state machines for each one, the technique can also be used for fingerprinting TLS implementations.

A Modbus-TCP Fuzzer for testing internetworked industrial systems (ETFA 2015)

Paper
Code

Abstract: Modbus/TCP is a network protocol for industrial communications encapsulated in TCP/IP network packets. There is an increasing need to test existing Modbus protocol implementations for security vulnerabilities, as devices become accessible even from the Internet. Fuzz testing can be used to discover implementation bugs in a fast and economical way. We present the design and implementation of MTF, a Modbus/TCP Fuzzer. The MTF incorporates a reconnaissance phase in the testing procedure so as to assist mapping the capabilities of the tested device and to adjust the attack vectors towards a more guided and informed testing rather than plain random testing. The MTF was used to test eight implementations of the Modbus protocol and revealed bugs and vulnerabilities that crash the execution, effectively resulting in denial of service attacks using only a few network packets.

PULSAR: Stateful Black-Box Fuzzing of Proprietary Network Protocols (Springer, Cham, 2015)

Paper

Abstract: The security of network services and their protocols critically depends on minimizing their attack surface. A single flaw in an implementation can suffice to compromise a service and expose sensitive data to an attacker. The discovery of vulnerabilities in protocol implementations, however, is a challenging task: While for standard protocols this process can be conducted with regular techniques for auditing, the situation becomes difficult for proprietary protocols if neither the program code nor the specification of the protocol are easily accessible. As a result, vulnerabilities in closed-source implementations can often remain undiscovered for a longer period of time. In this paper, we present PULSAR, a method for stateful black-box fuzzing of proprietary network protocols. Our method combines concepts from fuzz testing with techniques for automatic protocol reverse engineering and simulation. It proceeds by observing the traffic of a proprietary protocol and inferring a generative model for message formats and protocol states that can not only analyze but also simulate communication. During fuzzing this simulation can effectively explore the protocol state space and thereby enables uncovering vulnerabilities deep inside the protocol implementation. We demonstrate the efficacy of PULSAR in two case studies, where it identifies known as well as unknown vulnerabilities.

SECFUZZ: Fuzz-testing Security Protocols (AST 2012)

Paper

Abstract: We propose a light-weight, yet effective, technique for fuzz-testing security protocols. Our technique is modular, it exercises (stateful) protocol implementations in depth, and handles encrypted traffic. We use a concrete implementation of the protocol to generate valid inputs, and mutate the inputs using a set of fuzz operators. A dynamic memory analysis tool monitors the execution as an oracle to detect he vulnerabilities exposed by fuzz-testing. We provide the fuzzer with the necessary keys and cryptographic algorithms in order to properly mutate encrypted messages. We present a case study on two widely used, mature implementations of the Internet Key Exchange (IKE) protocol and report on two new vulnerabilities discovered by our fuzz-testing tool. We also compare the effectiveness of our technique to two existing model-based fuzz-testing tools for IKE.

Extension of SPIKE for Encrypted Protocol Fuzzing (2011)

Paper

Abstract: A fuzzer is a program that attempts to find security vulnerabilities in an application by sending random or semi-random input. Fuzzers have been widely used to find vulnerabilities in protocol implementations. The implementations may conform to the design of the protocol, but most of the times some glitches might remain. As a result vulnerabilities might remain unnoticed. Consequently, different implementations of the same protocol may be vulnerable to different kind of attacks. Fuzzers help us discover such implementation flaws. Among the currently available and popular ones, SPIKE is one recognized open-source fuzzing framework. However, SPIKE has a limitation of fuzzing only non-encrypted protocols. This paper presents the extension of SPIKE, called ESPIKE, for fuzzing of encrypted protocols. ESPIKE will facilitate testing implementations of SSL encrypted protocols. As a proof of concept for efficiency of ESPIKE we demonstrate its usage on sftp and https protocol.

AutoFuzz: Automated Network Protocol Fuzzing Framework (IJCSNS 2010)

Paper

Abstract: Assessing software security involves steps such as code review, risk analysis, penetration testing and fuzzing. During the fuzzing phase, the tester's goal is to find flaws in software by sending unexpected input to the target application and monitoring its behavior. In this paper we introduce the AutoFuzz [1] - extendable, open source framework used for testing network protocol implementations. AutoFuzz is a 'smart', man-in-the-middle, semi deterministic network protocol fuzzing framework. AutoFuzz learns a protocol implementation by constructing a Finite State Automaton (FSA) which captures the observed communications between a client and a server [5]. In addition, AutoFuzz learns individual message syntax, including fields and probable types, by applying the bioinformatics techniques of [2]. Finally, AutoFuzz can fuzz client or server protocol implementations by intelligently modifying the communication sessions between them using the FSA as a guide. AutoFuzz was applied to a variety of File Transfer Protocol (FTP) server implementations, confirming old and discovering new vulnerabilities.

Fuzzing with LLMs

Large Language Model guided Protocol Fuzzing (NDSS 2024)

Paper

Abstract: How to find security flaws in a protocol implementation without a machine-readable specification of the protocol? Facing the internet, protocol implementations are particularly security-critical software systems where inputs must adhere to a specific structure and order that is often informally specified in hundreds of pages in natural language (RFC). Without some machine-readable version of that protocol, it is difficult to automatically generate valid test inputs for its implementation that follow the required structure and order. It is possible to partially alleviate this challenge using mutational fuzzing on a set of recorded message sequences as seed inputs. However, the set of available seeds is often quite limited and will hardly cover the great diversity of protocol states and input structures. In this paper, we explore the opportunities of systematic interaction with pre-trained large language models (LLMs), which have ingested millions of pages of human-readable protocol specifications, to draw out machine-readable information about the protocol that can be used during protocol fuzzing. We use the knowledge of the LLMs about protocol message types for well-known protocols. We also checked the LLM’s capability in detecting “states” for stateful protocol implementations by generating sequences of messages and predicting response codes. Based on these observations, we have developed an LLM-guided protocol implementation fuzzing engine. Our protocol fuzzer CHATAFL constructs grammars for each message type in a protocol, and then mutates messages or predicts the next messages in a message sequence via interactions with LLMs. Experiments on a wide range of real-world protocols from PROFUZZBENCH show significant efficacy in state and code coverage. Our LLM- guided stateful fuzzer was compared with state-of-the-art fuzzers AFLNET and NSFUZZ. CHATAFL covers 47.60% and 42.69% more state transitions, 29.55% and 25.75% more states, and 5.81% and 6.74% more code, respectively. Apart from enhanced coverage, CHATAFL discovered nine distinct and previously unknown vulnerabilities in widely-used and extensively-tested protocol implementations while AFLNET and NSFUZZ only discovered three and four of them, respectively.

Large Language Models are Zero-Shot Fuzzers:Fuzzing Deep-Learning Libraries via Large Language Models (ISSTA 2023)

Paper

Abstract: Detecting bugs in Deep Learning (DL) libraries (e.g., TensorFlow/PyTorch) is critical for almost all downstream DL systems in ensuring effectiveness/safety for end users. Meanwhile, traditional fuzzing techniques can be hardly effective for such a challenging domain since the input DL programs need to satisfy both the input language (e.g., Python) syntax/semantics and the DL API input/shape constraints for tensor computations. To address these limitations, we propose TitanFuzz - the first approach to directly leveraging Large Language Models (LLMs) to generate input programs for fuzzing DL libraries. LLMs are titanic models trained on billions of code snippets and can auto-regressively generate human-like code snippets. Our key insight is that modern LLMs can also include numerous code snippets invoking DL library APIs in their training corpora, and thus can implicitly learn both language syntax/semantics and intricate DL API constraints for valid DL program generation. More specifically, we use both generative and infilling LLMs (e.g., Codex/InCoder) to generate and mutate valid/diverse input DL programs for fuzzing. Our experimental results demonstrate that TitanFuzz can achieve 30.38%/50.84% higher code coverage than state-of-the-art fuzzers on TensorFlow/PyTorch. Furthermore, TitanFuzz is able to detect 65 bugs, with 41 already confirmed as previously unknown bugs. This paper demonstrates that modern titanic LLMs can be leveraged to directly perform both generation-based and mutation-based fuzzing studied for decades, while being fully automated, generalizable, and applicable to domains challenging for traditional approaches (such as DL systems). We hope TitanFuzz can stimulate more work in this promising direction of LLMs for fuzzing.

Large Language Models are Few-shot Testers: Exploring LLM-based General Bug Reproduction (ICSE 2023)

Paper

Abstract: Many automated test generation techniques have been developed to aid developers with writing tests. To facilitate full automation, most existing techniques aim to either increase coverage, or generate exploratory inputs. However, existing test generation techniques largely fall short of achieving more semantic objectives, such as generating tests to reproduce a given bug report. Reproducing bugs is nonetheless important, as our empirical study shows that the number of tests added in open source repositories due to issues was about 28% of the corresponding project test suite size. Meanwhile, due to the difficulties of transforming the expected program semantics in bug reports into test oracles, existing failure reproduction techniques tend to deal exclusively with program crashes, a small subset of all bug reports. To automate test generation from general bug reports, we propose LIBRO, a framework that uses Large Language Models (LLMs), which have been shown to be capable of performing code-related tasks. Since LLMs themselves cannot execute the target buggy code, we focus on post-processing steps that help us discern when LLMs are effective, and rank the produced tests according to their validity. Our evaluation of LIBRO shows that, on the widely studied Defects4J benchmark, LIBRO can generate failure reproducing test cases for 33% of all studied cases (251 out of 750), while suggesting a bug reproducing test in first place for 149 bugs. To mitigate data contamination, we also evaluate LIBRO against 31 bug reports submitted after the collection of the LLM training data terminated: LIBRO produces bug reproducing tests for 32% of the studied bug reports. Overall, our results show LIBRO has the potential to significantly enhance developer efficiency by automatically generating tests from bug reports.

SMT Fuzzing

Validating SMT Solvers via Skeleton Enumeration Empowered by Historical Bug-Triggering Inputs (ICSE 2023)

Paper

Abstract: SMT solvers check the satisfiability of logic formulas over first-order theories, which have been utilized in a rich number of critical applications, such as software verification, test case generation, and program synthesis. Bugs hidden in SMT solvers would severely mislead those applications and further cause severe consequences. Therefore, ensuring the reliability and robustness of SMT solvers is of critical importance. Although many approaches have been proposed to test SMT solvers, it is still a challenge to discover bugs effectively. To tackle such a challenge, we conduct an empirical study on the historical bug-triggering formulas in SMT solvers' bug tracking systems. We observe that the historical bug-triggering formulas contain valuable skeletons (i.e., core structures of formulas) as well as associated atomic formulas which can cast significant impacts on formulas' ability in triggering bugs. Therefore, we propose a novel approach that utilizes the skeletons extracted from the historical bug-triggering formulas and enumerates atomic formulas under the guidance of association rules derived from historical formulas. In this study, we realized our approach as a practical fuzzing tool HistFuzz and conducted extensive testing on the well-known SMT solvers Z3 and cvc5. To date, HistFuzz has found 111 confirmed new bugs for Z3 and cvc5, of which 108 have been fixed by the developers. More notably, out of the confirmed bugs, 23 are soundness bugs and invalid model bugs found in the solvers' default mode, which are essential for SMT solvers. In addition, our experiments also demonstrate that HistFuzz outperforms the state-of-the-art SMT solver fuzzers in terms of achieved code coverage and effectiveness.

Alt-Ergo-Fuzz: A fuzzer for the Alt-Ergo SMT solver (JFLA 2022)

Paper
Code

Abstract: Alt-Ergo is an open source Satisfiability Modulo Theories (SMT) solver programmed in OCaml. It was designed for program verification and it's used as a back end by other software verification tools such as Frama-C, SPARK, Why3, Atelier-B and Caveat, the reliability of which depends on the soundness of Alt-Ergo's answers and the absence of bugs in it. Fuzzing is an efficient technique to test programs and find bugs. It works by quickly and automatically generating input data with which to test the software. American Fuzzy Lop (AFL) is one of the most well-known and most used fuzzers in both academia and the industry. It has managed to find many bugs in various programs thanks to its grey box fuzzing technique that uses genetic algorithms and program instrumentation to generate test data that maximizes code and execution path coverage in the targeted software. In this paper we present Alt-Ergo-Fuzz, a fuzzer for Alt-Ergo that we developed with the aim of finding faults and unsoundness bugs to solve and improve its reliability. By using AFL as a back end, the Crowbar OCaml library for test case generation and the CVC5 SMT solver as a reference solver of which the answers will be used to determine whether or not Alt-Ergo's answers are correct, we managed to develop Alt-Ergo-Fuzz, which even as a work in progress and in only twenty days of testing managed to find four never found before bugs in Alt-Ergo.

BanditFuzz: Fuzzing SMT Solvers with Multi-agent Reinforcement Learning (FM 2021)

Paper

Abstract: We present BanditFuzz, a multi-agent reinforcement learning (RL) guided performance fuzzer for state-of-the-art Satisfiability Modulo Theories (SMT) solvers. BanditFuzz constructs inputs that expose performance issues in a set of target solvers relative to a set of reference solvers, and is the first performance fuzzer that supports the entirety of the theories in the SMT-LIB initiative. Another useful feature of BanditFuzz is that users can specify the size of inputs they want, thus enabling developers to construct very small inputs that zero-in on a performance problem in their SMT solver relative to other competitive solvers. We evaluate BanditFuzz across 52 logics from SMT-COMP '20 targeting competition-winning solvers against runner-ups. We baseline BanditFuzz against random fuzzing and a single agent algorithm and observe a significant improvement, with up to a 82.6% improvement in the margin of PAR-2 scores across baselines on their respective benchmarks. Furthermore, we reached out to developers and contributors of the CVC4, Z3, and Bitwuzla solvers and provide case studies of how BanditFuzz was able to expose surprising performance deficiencies in each of these tools.

Skeletal Approximation Enumeration for SMT Solver Testing (FSE 2021)

Paper

Abstract: Ensuring the equality of SMT solvers is critical due to its broad spectrum of applications in academia and industry, such as symbolic execution and program verification. Existing approaches to testing SMT solvers are either too costly or find difficulties generalizing to different solvers and theories, due to the test oracle problem. To complement existing approaches and overcome their weaknesses, this paper introduces skeletal approximation enumeration (SAE), a novel lightweight and general testing technique for all first-order theories. To demonstrate its practical utility, we have applied the SAE technique to test Z3 and CVC4, two comprehensively tested, state-of-the-art SMT solvers. By the time of writing, our approach had found 71 confirmed bugs in Z3 and CVC4,55 of which had already been fixed.

Fuzzing SMT Solvers via Two-Dimensional Input Space Exploration (ISSTA 2021)

Paper

Abstract: Satisfiability Modulo Theories (SMT) solvers serve as the core engine of many techniques, such as symbolic execution. Therefore, ensuring the robustness and correctness of SMT solvers is critical. While fuzzing is an efficient and effective method for validating the quality of SMT solvers, we observe that prior fuzzing work only focused on generating various first-order formulas as the inputs but neglected the algorithmic configuration space of an SMT solver, which leads to under-reporting many deeply-hidden bugs. In this paper, we present Falcon, a fuzzing technique that explores both the formula space and the configuration space. Combining the two spaces significantly enlarges the search space and makes it challenging to detect bugs efficiently. We solve this problem by utilizing the correlations between the two spaces to reduce the search space, and introducing an adaptive mutation strategy to boost the search efficiency. During six months of extensive testing, Falcon finds 518 confirmed bugs in CVC4 and Z3, two state-of-the-art SMT solvers, 469 of which have already been fixed. Compared to two state-of-the-art fuzzers, Falcon detects 38 and 44 more bugs and improves the coverage by a large margin in 24 hours of testing.

Detecting Critical Bugs in SMT Solvers Using Blackbox Mutational Fuzzing (FSE 2020)

Paper

Abstract: Formal methods use SMT solvers extensively for deciding formula satisfiability, for instance, in software verification, systematic test generation, and program synthesis. However, due to their complex implementations, solvers may contain critical bugs that lead to unsound results. Given the wide applicability of solvers in software reliability, relying on such unsound results may have detrimental consequences. In this paper, we present STORM, a novel blackbox mutational fuzzing technique for detecting critical bugs in SMT solvers. We run our fuzzer on seven mature solvers and find 29 previously unknown critical bugs. STORM is already being used in testing new features of popular solvers before deployment.

On the Unusual Effectiveness of Type-aware Mutations for Testing SMT Solvers

Paper

Abstract: We propose type-aware operator mutation, a simple, but unusually effective approach for testing SMT solvers. The key idea is to mutate operators of conforming types within the seed formulas to generate well-typed mutant formulas. These mutant formulas are then used as the test cases for SMT solvers. We realized typeaware operator mutation within the OpFuzz tool and used it to stress-test Z3 and CVC4, two state-of-the-art SMT solvers. Type-aware operator mutations are unusually effective: During nine months of extensive testing with OpFuzz, we reported 909 bugs in Z3 and CVC4,1 out of which 632 bugs were confirmed and 531 of the confirmed bugs were fixed by the developers. The detected bugs are highly diverse - we found bugs of many different types (soundness bugs, invalid model bugs, crashes, etc.), logics and solver configurations. We have further conducted an in-depth study on the bugs found by OpFuzz. The study results show that the bugs found by OpFuzz are of high quality. Many of them affect core components of the SMT solvers' codebases, and some required major changes for the developers to fix. Among the 909 bugs found by OpFuzz, 130 were soundness bugs, the most critical bugs in SMT solvers, and 501 were in the default modes of the solvers. Notably, OpFuzz found 16 critical soundness bugs in CVC4, which has proved to be a very stable SMT solver

BanditFuzz: Fuzzing SMT Solvers with Reinforcement Learning (2020)

Paper

Satisfiability Modulo Theories (SMT) solvers are fundamental tools in the broad context of software engineering and security research. If SMT solvers are to continue to have an impact, it is imperative we develop efficient and systematic testing methods for them. To this end, we present a reinforcement learning driven fuzzing system BanditFuzz that zeroes in on the grammatical constructs of well-formed solver inputs that are the root cause of performance or correctness issues in solvers-under-test. To the best of our knowledge, BanditFuzz is the first machine-learning based fuzzer for SMT solvers. BanditFuzz takes as input a grammar G describing the well-formed inputs to a set of distinct solvers (say, P_1 and P_2) that implement the same specification and a fuzzing objective (e.g., maximize the relative performance difference between P_1 and P_2), and outputs a ranked list of grammatical constructs that are likely to maximize performance differences between P_1 and P_2 or are root causes of errors in these solvers. Typically, mutation fuzzing is implemented as a set of random mutations applied to a given input. By contrast, the key innovation behind BanditFuzz is the modeling of a grammar-preserving fuzzing mutator as a reinforcement learning (RL) agent that, via blackbox interactions with programs-under-test, learns which grammatical constructs are most likely the cause of an error or performance issue. Using BanditFuzz, we discovered 1700 syntactically unique inputs resulting in inconsistent answers across state-of-the-art SMT solvers Z3, CVC4, Colibri, MathSAT, and Z3str3 over the floating-point and string SMT theories. Further, using BanditFuzz, we constructed two benchmark suites (with 400 floating-point and 110 string instances) that expose performance issues in all considered solvers. We also performed a comparison of BanditFuzz against random, mutation, and evolutionary fuzzing methods. We observed up to a 31% improvement in performance fuzzing and up to 81% improvement in the number of bugs found by BanditFuzz relative to these other methods for the same amount of time provided to all methods.

Validating SMT Solvers via Semantic Fusion (PLDI 2020)

Abstract: We introduce Semantic Fusion, a general, effective methodology for validating Satisfiability Modulo Theory (SMT) solvers. Our key idea is to fuse two existing equisatisfiable (i.e., both satisfiable or unsatisfiable) formulas into a new formula that combines the structures of its ancestors in a novel manner and preserves the satisfiability by construction. This fused formula is then used for validating SMT solvers. We realized Semantic Fusion as YinYang, a practical SMT solver testing tool. During four months of extensive testing, YinYang has found 45 confirmed, unique bugs in the default arithmetic and string solvers of Z3 and CVC4, the two state-of-the-art SMT solvers. Among these, 41 have already been fixed by the developers. The majority (29/45) of these bugs expose critical soundness issues. Our bug reports and testing effort have been well-appreciated by SMT solver developers.

Automatically Testing String Solvers (ICSE 2020)

Paper

Abstract: SMT solvers are at the basis of many applications, such as program verification, program synthesis, and test case generation. For all these applications to provide reliable results, SMT solvers must answer queries correctly. However, since they are complex, highly-optimized software systems, ensuring their correctness is challenging. In particular, state-of-the-art testing techniques do not reliably detect when an SMT solver is unsound.

In this paper, we present an automatic approach for generating test cases that reveal soundness errors in the implementations of string solvers, as well as potential completeness and performance issues. We synthesize input formulas that are satisfiable or unsatisfiable by construction and use this ground truth as test oracle. We automatically apply satisfiability-preserving transformations to generate increasingly-complex formulas, which allows us to detect many errors with simple inputs and, thus, facilitates debugging.

The experimental evaluation shows that our technique effectively reveals bugs in the implementation of widely-used SMT solvers and applies also to other types of solvers, such as automata-based solvers. We focus on strings here, but our approach carries over to other theories and their combinations.

StringFuzz: A fuzzer for string solvers (CAV 2018)

Paper

Abstract: In this paper, we introduce StringFuzz: a modular SMT-LIB problem instance transformer and generator for string solvers. We supply a repository of instances generated by StringFuzz in SMT-LIB 2.0/2.5 format. We systematically compare Z3str3, CVC4, Z3str2, and Norn on groups of such instances, and identify those that are particularly challenging for some solvers. We briefly explain our observations and show how StringFuzz helped discover causes of performance degradations in Z3str3.

Anti Fuzzing

Antifuzz: impeding fuzzing audits of binary executables (USENIX Security2019)

Paper
Code

Abstract: A general defense strategy in computer security is to increase the cost of successful attacks in both computational resources as well as human time. In the area of binary security, this is commonly done by using obfuscation methods to hinder reverse engineering and the search for software vulnerabilities. However, recent trends in automated bug finding changed the modus operandi. Nowadays it is very common for bugs to be found by various fuzzing tools. Due to ever-increasing amounts of automation and research on better fuzzing strategies, large-scale, dragnet-style fuzzing of many hundreds of targets becomes viable. As we show, current obfuscation techniques are aimed at increasing the cost of human understanding and do little to slow down fuzzing. In this paper, we introduce several techniques to protect a binary executable against an analysis with automated bug finding approaches that are based on fuzzing, symbolic/concolic execution, and taint-assisted fuzzing (commonly known as hybrid fuzzing). More specifically, we perform a systematic analysis of the fundamental assumptions of bug finding tools and develop general countermeasures for each assumption. Note that these techniques are not designed to target specific implementations of fuzzing tools, but address general assumptions that bug finding tools necessarily depend on. Our evaluation demonstrates that these techniques effectively impede fuzzing audits, while introducing a negligible performance overhead. Just as obfuscation techniques increase the amount of human labor needed to find a vulnerability, our techniques render automated fuzzing-based approaches futile.

FUZZIFICATION: Anti-Fuzzing Technique (USENIX Security2019)

Abstract: Fuzzing is a software testing technique that quickly and automatically explores the input space of a program without knowing its internals. Therefore, developers commonly use fuzzing as part of test integration throughout the software development process. Unfortunately, it also means that such a blackbox and the automatic natures of fuzzing are appealing to adversaries who are looking for zero-day vulnerabilities. To solve this problem, we propose a new mitigation approach, called Fuzzification , that helps developers protect the released, binary-only software from attackers who are capable of applying state-of-the-art fuzzing techniques. Given a performance budget, this approach aims to hinder the fuzzing process from adversaries as much as possible. We propose three Fuzzification techniques: 1) SpeedBump, which amplifies the slowdown in normal executions by hundreds of times to the fuzzed execution, 2) BranchTrap, interfering with feedback logic by hiding paths and polluting coverage maps, and 3) AntiHybrid, hindering taint-analysis and symbolic execution. Each technique is designed with best-effort, defensive measures that attempt to hinder adversaries from bypassing Fuzzification. Our evaluation on popular fuzzers and real-world applications shows that Fuzzification effectively reduces the number of discovered paths by 70.3% and decreases the number of identified crashes by 93.0% from real-world binaries, and decreases the number of detected bugs by 67.5% from LAVA-M dataset while under user-specified overheads for common workloads. We discuss the robustness of Fuzzification techniques against adversarial analysis techniques. We open-source our Fuzzification system to foster future research.

Kernel Fuzzing

No Grammar, No Problem: Towards Fuzzing the Linux Kernel without System-Call Descriptions (NDSS 2023)

Paper

Abstract: The integrity of the entire computing ecosystem depends on the security of our operating systems (OSes). Unfortunately, due to the scale and complexity of OS code, hundreds of security issues are found in OSes, every year. As such, operating systems have constantly been prime use-cases for applying security-analysis tools. In recent years, fuzz-testing has appeared as the dominant technique for automatically finding security issues in software. As such, fuzzing has been adapted to find thousands of bugs in kernels. However, modern OS fuzzers, such as Syzkaller, rely on precise, extensive, manually created harnesses and grammars for each interface fuzzed within the kernel. Due to this reliance on grammars, current OS fuzzers are faced with scaling-issues.

In this paper, we present FuzzNG, our generic approach to fuzzing system-calls on OSes. Unlike Syzkaller, FuzzNG does not require intricate descriptions of system-call interfaces in order to function. Instead FuzzNG leverages fundamental Kernel design features in order to reshape and simplify the fuzzer’s input-space. As such FuzzNG only requires a small config, for each new target: essentially a list of files and system-call numbers the fuzzer should explore.

We implemented FuzzNG for the Linux kernel. Testing FuzzNG over 10 Linux components with extensive descrip tions in Syzkaller showed that, on average, FuzzNG achieves 102.5% of Syzkaller’s coverage. FuzzNG found 9 new bugs (5 in components that Syzkaller had already fuzzed extensively, for years). Additionally, FuzzNG’s lightweight configs are less than 1.7% the size of Syzkaller’s manually-written grammars. Crucially, FuzzNG achieves this without initial seed-inputs, or expert guidance.

SFuzz: Slice-based Fuzzing for Real-Time Operating Systems (CCS 2022)

Paper

Abstract: Real-Time Operating System (RTOS) has become the main category of embedded systems. It is widely used to support tasks requiring real-time response such as printers and switches. The security of RTOS has been long overlooked as it was running in special environments isolated from attackers. However, with the rapid development of IoT devices, tremendous RTOS devices are connected to the public network. Due to the lack of security mechanisms, these devices are extremely vulnerable to a wide spectrum of attacks. Even worse, the monolithic design of RTOS combines various tasks and services into a single binary, which hinders the current program testing and analysis techniques working on RTOS. In this paper, we propose SFuzz, a novel slice-based fuzzer, to detect security vulnerabilities in RTOS. Our insight is that RTOS usually divides a complicated binary into many separated but single-minded tasks. Each task accomplishes a particular event in a deterministic way and its control flow is usually straightforward and independent. Therefore, we identify such code from the monolithic RTOS binary and synthesize a slice for effective testing. Specifically, SFuzz first identifies functions that handle user input, constructs call graphs that start from callers of these functions, and leverages forward slicing to build the execution tree based on the call graphs and pruning the paths independent of external inputs. Then, it detects and handles roadblocks within the coarse-grain scope that hinder effective fuzzing, such as instructions unrelated to the user input. And then, it conducts coverage-guided fuzzing on these code snippets. Finally, SFuzz leverages forward and backward slicing to track and verify each path constraint and determine whether a bug discovered in the fuzzer is a real vulnerability. SFuzz successfully discovered 77 zero-day bugs on 35 RTOS samples, and 67 of them have been assigned CVE or CNVD IDs. Our empirical evaluation shows that SFuzz outperforms the state-of-the-art tools (e.g., UnicornAFL) on testing RTOS.

Demystifying the Dependency Challenge in Kernel Fuzzing (ICSE 2022)

Paper

Abstract: Fuzz testing operating system kernels remains a daunting task to date. One known challenge is that much of the kernel code is locked under specific kernel states and current kernel fuzzers are not effective in exploring such an enormous state space. We refer to this problem as the dependency challenge. Though there are some efforts trying to address the dependency challenge, the prevalence and categorization of dependencies have never been studied. Most prior work simply attempted to recover dependencies opportunistically whenever they are relatively easy to recognize. In this paper, we undertake a substantial measurement study to systematically understand the real challenge behind dependencies. To our surprise, we show that even for well-fuzzed kernel modules, unresolved dependencies still account for 59% - 88% of the uncovered branches. Furthermore, we show that the dependency challenge is only a symptom rather than the root cause of failing to achieve more coverage. By distilling and summarizing our findings, we believe the research provides valuable guidance to future research in kernel fuzzing. Finally, we propose a number of novel research directions directly based on the insights gained from the measurement study.

Semantic-Informed Driver Fuzzing Without Both the Hardware Devices and the Emulators (ICSE 2022)

Paper

Abstract: Device drivers are security-critical. In monolithic kernels like Linux, there are hundreds of thousands of drivers which run in the same privilege as the core kernel. Consequently, a bug in a driver can compromise the whole system. More critically, drivers are particularly buggy. First, drivers receive complex and untrusted inputs from not only the user space but also the hardware. Second, the driver code can be developed by less-experienced third parties, and is less tested because running a driver requires the corresponding hardware device or the emulator. Therefore, existing studies show that drivers tend to have a higher bug density and have become a major security threat. Existing testing techniques have to focus the fuzzing on a limited number of drivers that have the corresponding devices or the emulators, thus cannot scale.

In this paper, we propose a device-free driver fuzzing system, DR. FUZZ, that does not require hardware devices to fuzztest drivers. The core of DR. FUZZ is a semantic-informed mechanism that efficiently generates inputs to properly construct relevant data structures to pass the 'validation chain'? in driving initialization, which enables subsequent device-free driver fuzzing. The elimination of the needs for the hardware devices and the emulators removes the bottleneck in driver testing. The semanticinformed mechanism incorporates multiple new techniques to make device-free driver fuzzing practical: inferring valid input values for passing the validation chain in initialization, inferring the temporal usage order of input bytes to minimize mutation space, and employing error states as a feedback to direct the fuzzing going through the validation chain. Moreover, the semantic-informed mechanism is generic; we can also instruct it to generate semi-malformed inputs for a higher code coverage. We evaluate DR. FUZZ on 214 Linux drivers. With an only 24-hour time budget, DR. FUZZ can successfully initialize and enable most of the drivers without the corresponding devices, whereas existing fuzzers like syzkaller cannot succeed in any case. DR. FUZZ also significantly outperforms existing driver fuzzers that are even equipped with the device or emulator in other aspects: it increases the code coverage by 70% and the throughput by 18%. With DR. FUZZ, we also find 46 new bugs in these Linux drivers.

SyzScope: Revealing High-Risk Security Impacts of Fuzzer-Exposed Bugs in Linux kernel (USENIX SEC'22)

Video
Paper

Abstract: Fuzzing has become one of the most effective bug finding approach for software. In recent years, 24*7 continuous fuzzing platforms have emerged to test critical pieces of software, e.g., Linux kernel. Though capable of discovering many bugs and providing reproducers (e.g., proof-of-concepts), a major problem is that they neglect a critical function that should have been built-in, i.e., evaluation of a bug's security impact. It is well-known that the lack of understanding of security impact can lead to delayed bug fixes as well as patch propagation. In this paper, we develop SyzScope, a system that can automatically uncover new 'high-risk'? impacts given a bug with seemingly 'low-risk'? impacts. From analyzing over a thousand low-risk bugs on syzbot, SyzScope successfully determined that 183 low-risk bugs (more than 15%) in fact contain high-risk impacts, e.g., control flow hijack and arbitrary memory write, some of which still do not have patches available yet.

HEALER: Relation Learning Guided Kernel Fuzzing (SOSP 2021)

Video
Paper

Abstract: Modern operating system kernels are too complex to be free of bugs. Fuzzing is a promising approach for vulnerability detection and has been applied to kernel testing. However, existing work does not consider the influence relations between system calls when generating and mutating inputs, resulting in difficulties when trying to reach into the kernel's deeper logic effectively.

In this paper, we propose HEALER, a kernel fuzzer that improves fuzzing's effectiveness by utilizing system call relation learning. HEALER learns the influence relations between system calls by dynamically analyzing minimized test cases. Then, HEALER utilizes the learned relations to guide input generation and mutation, which improves the quality of test cases and the effectiveness of fuzzing. We implemented HEALER and evaluated its performance on recent versions of the Linux kernel. Compared to state-of-the-art kernel fuzzers such as Syzkaller and Moonshine, HEALER improves branch coverage by 28% and 21%, while achieving a speedup of 2.2× and 1.8×, respectively. In addition, HEALER detected 218 vulnerabilities, 33 of which are previously unknown and have been confirmed by the corresponding kernel maintainers.

NTFUZZ: Enabling Type-Aware Kernel Fuzzing on Windows with Static Binary Analysis(S&P 2021)

Abstract: Although it is common practice for kernel fuzzers to leverage type information of system calls, current Windows kernel fuzzers do not follow the practice as most system calls are private and largely undocumented. In this paper, we present a practical static binary analyzer that automatically infers system call types on Windows at scale. We incorporate our analyzer to NTFUZZ, a type-aware Windows kernel fuzzing framework. To our knowledge, this is the first practical fuzzing system that utilizes scalable binary analysis on a COTS OS. With NTFUZZ, we found 11 previously unknown kernel bugs, and earned $25,000 through the bug bounty program offered by Microsoft. All these results confirm the practicality of our system as a kernel fuzzer.

Finding Bugs in File Systems with an Extensible Fuzzing Framework (TOS 2020)

Paper

Abstract: File systems are too large to be bug free. Although handwritten test suites have been widely used to stress file systems, they can hardly keep up with the rapid increase in file system size and complexity, leading to new bugs being introduced. These bugs come in various flavors: buffer overflows to complicated semantic bugs. Although bug-specific checkers exist, they generally lack a way to explore file system states thoroughly. More importantly, no turnkey solution exists that unifies the checking effort of various aspects of a file system under one umbrella.

In this article, to highlight the potential of applying fuzzing to find any type of file system bugs in a generic way, we propose Hydra, an extensible fuzzing framework. Hydra provides building blocks for file system fuzzing, including input mutators, feedback engines, test executors, and bug post-processors. As a result, developers only need to focus on building the core logic for finding bugs of their interests. We showcase the effectiveness of Hydra with four checkers that hunt crash inconsistency, POSIX violations, logic assertion failures, and memory errors. So far, Hydra has discovered 157 new bugs in Linux file systems, including three in verified file systems (FSCQ and Yxv6).

Finding race conditions in Kernels: from fuzzing to symbolic exection (2020)

Paper

Abstract: The scale and pervasiveness of concurrent software pose challenges for security researchers: race conditions are more prevalent than ever, and the growing software complexity keeps exacerbating the situation - expanding the arms race between security practitioners and attackers beyond memory errors. As a consequence, we need a new generation of bug hunting tools that not only scale well with increasingly larger codebases but also catch up with the growing importance of race conditions.

In this thesis, two complementary race detection frameworks for OS kernels are presented: multi-dimensional fuzz testing and symbolic checking. Fuzz testing turns bug finding into a probabilistic search, but current practices restrict themselves to one dimension only (sequential executions). This thesis illustrates how to explore the concurrency dimension and extend the bug scope beyond memory errors to the broad spectrum of concurrency bugs. On the other hand, conventional symbolic executors face challenges when applied to OS kernels, such as path explosions due to branching and loops. They also lack a systematic way of modeling and tracking constraints in the concurrency dimension (e.g., to enforce a particular schedule for thread interleavings) The gap can be partially filled with novel techniques for symbolic execution in this thesis.

Hydra: An Extensible Fuzzing Framework for Finding Semantic Bugs in File Systems (SOSP 2019)

Abstract: File systems are too large to be bug free. Although handwritten test suites have been widely used to stress file systems, they can hardly keep up with the rapid increase in file system size and complexity, leading to new bugs being introduced and reported regularly. These bugs come in various flavors: simple buffer overflows to sophisticated semantic bugs. Although bug-specific checkers exist, they generally lack a way to explore file system states thoroughly. More importantly, no turnkey solution exists that unifies the checking effort of various aspects of a file system under one umbrella.

In this paper, we highlight the potential of applying fuzzing to find not just memory errors but, in theory, any type of file system bugs with an extensible fuzzing framework: Hydra. Hydra provides building blocks for file system fuzzing, including input mutators, feedback engines, a libOS-based executor, and a bug reproducer with test case minimization. As a result, developers only need to focus on building the core logic for finding bugs of their own interests. We showcase the effectiveness of Hydra with four checkers that hunt crash inconsistency, POSIX violations, logic assertion failures, and memory errors. So far, Hydra has discovered 91 new bugs in Linux file systems, including one in a verified file system (FSCQ), as well as four POSIX violations.

Fuzzing File Systems via Two-Dimensional Input Space Exploration (S&P 2019)

Abstract: File systems, a basic building block of an OS, are too big and too complex to be bug free. Nevertheless, file systems rely on regular stress-testing tools and formal checkers to find bugs, which are limited due to the ever-increasing complexity of both file systems and OSes. Thus, fuzzing, proven to be an effective and a practical approach, becomes a preferable choice, as it does not need much knowledge about a target. However, three main challenges exist in fuzzing file systems: mutating a large image blob that degrades overall performance, generating image-dependent file operations, and reproducing found bugs, which is difficult for existing OS fuzzers. Hence, we present JANUS, the first feedback-driven fuzzer that explores the two-dimensional input space of a file system, i.e., mutating metadata on a large image, while emitting image-directed file operations. In addition, JANUS relies on a library OS rather than on traditional VMs for fuzzing, which enables JANUS to load a fresh copy of the OS, thereby leading to better reproducibility of bugs. We evaluate JANUS on eight file systems and found 90 bugs in the upstream Linux kernel, 62 of which have been acknowledged. Forty-three bugs have been fixed with 32 CVEs assigned. In addition, JANUS achieves higher code coverage on all the file systems after fuzzing 12 hours, when compared with the state-of-the-art fuzzer Syzkaller for fuzzing file systems. JANUS visits 4.19x and 2.01x more code paths in Btrfs and ext4, respectively. Moreover, JANUS is able to reproduce 88-100% of the crashes, while Syzkaller fails on all of them.

Unicorefuzz: On the Viability of Emulation for Kernelspace Fuzzing (USENIX WOOT'19)

Abstract: Fuzzing uncovers an ever-growing number of critical vulnerabilities. Despite the simple concept - execute the target until it crashes - setting up fuzz tests can pose complex challenges. This is especially true for code that cannot run as part of a userland process on desktop operating systems - for example device drivers and kernel components. In this paper, we explore the use of CPU emulation to fuzz arbitrary parsers in kernelspace with coverage-based feedback. We propose and open-source Unicorefuzz and explain merits and pitfalls of emulation-based fuzzing approaches. The viability of the approach is evaluated against artificial Linux kernel modules, the Open vSwitch network virtualization component as well as bugs originally uncovered by syzcaller. Emulator-based fuzzing of kernel code is not very complex to set up and can even be used to fuzz operating systems and devices for which no source code is available.

Razzer: Finding Kernel Race Bugs through Fuzzing (S&P 2019)

Abstract: A data race in a kernel is an important class of bugs, critically impacting the reliability and security of the associated system. As a result of a race, the kernel may become unresponsive. Even worse, an attacker may launch a privilege escalation attack to acquire root privileges. In this paper, we propose Razzer, a tool to find race bugs in kernels. The core of Razzer is in guiding fuzz testing towards potential data race spots in the kernel. Razzer employs two techniques to find races efficiently: a static analysis and a deterministic thread interleaving technique. Using a static analysis, Razzer identifies over-approximated potential data race spots, guiding the fuzzer to search for data races in the kernel more efficiently. Using the deterministic thread interleaving technique implemented at the hypervisor, Razzer tames the non-deterministic behavior of the kernel such that it can deterministically trigger a race. We implemented a prototype of Razzer and ran the latest Linux kernel (from v4.16-rc3 to v4.18-rc3) using Razzer. As a result, Razzer discovered 30 new races in the kernel, with 16 subsequently confirmed and accordingly patched by kernel developers after they were reported.

MoonShine: Optimizing OS Fuzzer Seed Selection with Trace Distillation (USENIX Security2018)

Abstract: OS fuzzers primarily test the system call interface between the OS kernel and user-level applications for security vulnerabilities. The effectiveness of evolutionary OS fuzzers depends heavily on the quality and diversity of their seed system call sequences. However, generating good seeds for OS fuzzing is a hard problem as the behavior of each system call depends heavily on the OS kernel state created by the previously executed system calls. Therefore, popular evolutionary OS fuzzers often rely on hand-coded rules for generating valid seed sequences of system calls that can bootstrap the fuzzing process. Unfortunately, this approach severely restricts the diversity of the seed system call sequences and therefore limits the effectiveness of the fuzzers. In this paper, we develop MoonShine, a novel strategy for distilling seeds for OS fuzzers from system call traces of real-world programs while still maintaining the dependencies across the system calls. MoonShine leverages light-weight static analysis for efficiently detecting dependencies across different system calls. We designed and implemented MoonShine as an extension to Syzkaller, a state-of-the-art evolutionary fuzzer for the Linux kernel. Starting from traces containing 2.8 million system calls gathered from 3,220 real-world programs, MoonShine distilled down to just over 14,000 calls while preserving 86% of the original code coverage. Using these distilled seed system call sequences, MoonShine was able to improve Syzkaller's achieved code coverage for the Linux kernel by 13% on average. MoonShine also found 14 new vulnerabilities in the Linux kernel that were not found by Syzkaller.

FUZE: Towards Facilitating Exploit Generation for Kernel Use-After-Free Vulnerabilities (USENIX Security2018)

Abstract: Software vendors usually prioritize their bug remediation based on ease of their exploitation. However, accurately determining exploitability typically takes tremendous hours and requires significant manual efforts. To address this issue, automated exploit generation techniques can be adopted. In practice, they however exhibit an insufficient ability to evaluate exploitability particularly for the kernel Use-After-Free (UAF) vulnerabilities. This is mainly because of the complexity of UAF exploitation as well as the scalability of an OS kernel.

In this paper, we therefore propose FUZE, a new framework to facilitate the process of kernel UAF exploitation. The design principle behind this technique is that we expect the ease of crafting an exploit could augment a security analyst with the ability to expedite exploitability evaluation. Technically, FUZE utilizes kernel fuzzing along with symbolic execution to identify, analyze and evaluate the system calls valuable and useful for kernel UAF exploitation.

To demonstrate the utility of FUZE, we implement FUZE on a 64-bit Linux system by extending a binary analysis framework and a kernel fuzzer. Using 15 real-world kernel UAF vulnerabilities on Linux systems, we then demonstrate FUZE could not only escalate kernel UAF exploitability and but also diversify working exploits. In addition, we show that FUZE could facilitate security mitigation bypassing, making exploitability evaluation less labor-intensive and more efficient.

kAFL: Hardware-Assisted Feedback Fuzzing for OS Kernels (Usenix Security2017)

Abstract: Many kinds of memory safety vulnerabilities have been endangering software systems for decades. Amongst other approaches, fuzzing is a promising technique to unveil various software faults. Recently, feedback-guided fuzzing demonstrated its power, producing a steady stream of security-critical software bugs. Most fuzzing efforts - especially feedback fuzzing - are limited to user space components of an operating system (OS), although bugs in kernel components are more severe, because they allow an attacker to gain access to a system with full privileges. Unfortunately, kernel components are difficult to fuzz as feedback mechanisms (i.e., guided code coverage) cannot be easily applied. Additionally, non-determinism due to interrupts, kernel threads, statefulness, and similar mechanisms poses problems. Furthermore, if a process fuzzes its own kernel, a kernel crash highly impacts the performance of the fuzzer as the OS needs to reboot.

In this paper, we approach the problem of coverage-guided kernel fuzzing in an OS-independent and hardware-assisted way: We utilize a hypervisor and Intel's Processor Trace (PT) technology. This allows us to remain independent of the target OS as we just require a small user space component that interacts with the targeted OS. As a result, our approach introduces almost no performance overhead, even in cases where the OS crashes, and performs up to 17,000 executions per second on an off-the-shelf laptop. We developed a framework called kernel-AFL (kAFL) to assess the security of Linux, macOS, and Windows kernel components. Among many crashes, we uncovered several flaws in the ext4 driver for Linux, the HFS and APFS file system of macOS, and the NTFS driver of Windows.

DIFUZE: Interface aware fuzzing for kernel drivers (CCS 2017)

Abstract: Device drivers are an essential part in modern Unix-like systems to handle operations on physical devices, from hard disks and printers to digital cameras and Bluetooth speakers. The surge of new hardware, particularly on mobile devices, introduces an explosive growth of device drivers in system kernels. Many such drivers are provided by third-party developers, which are susceptible to security vulnerabilities and lack proper vetting. Unfortunately, the complex input data structures for device drivers render traditional analysis tools, such as fuzz testing, less effective, and so far, research on kernel driver security is comparatively sparse. In this paper, we present DIFUZE, an interface-aware fuzzing tool to automatically generate valid inputs and trigger the execution of the kernel drivers. We leverage static analysis to compose correctly-structured input in the userspace to explore kernel drivers. DIFUZE is fully automatic, ranging from identifying driver handlers, to mapping to device file names, to constructing complex argument instances. We evaluate our approach on seven modern Android smartphones. The results showthat DIFUZE can effectively identify kernel driver bugs, and reports 32 previously unknown vulnerabilities, including flaws that lead to arbitrary code execution.

IMF: Inferred Model-based Fuzzer (CCS 2017)

Paper
Code

Abstract: Kernel vulnerabilities are critical in security because they naturally allow attackers to gain unprivileged root access. Although there has been much research on finding kernel vulnerabilities from source code, there are relatively few research on kernel fuzzing, which is a practical bug finding technique that does not require any source code. Existing kernel fuzzing techniques involve feeding in random input values to kernel API functions. However, such a simple approach does not reveal latent bugs deep in the kernel code, because many API functions are dependent on each other, and they can quickly reject arbitrary parameter values based on their calling context. In this paper, we propose a novel fuzzing technique for commodity OS kernels that leverages inferred dependence model between API function calls to discover deep kernel bugs. We implement our technique on a fuzzing system, called IMF. IMF has already found 32 previously unknown kernel vulnerabilities on the latest macOS version 10.12.3 (16D32) at the time of this writing.

Hybrid Fuzzing

Evaluating and Improving Hybrid Fuzzing (ICSE 2023)

Paper

Abstract To date, various hybrid fuzzers have been proposed for maximal program vulnerability exposure by integrating the power of fuzzing strategies and concolic executors. While the existing hybrid fuzzers have shown their superiority over conventional coverage-guided fuzzers, they seldom follow equivalent evaluation setups, e.g., benchmarks and seed corpora. Thus, there is a pressing need for a comprehensive study on the existing hybrid fuzzers to provide implications and guidance for future research in this area. To this end, in this paper, we conduct the first extensive study on state-of-the-art hybrid fuzzers. Surprisingly, our study shows that the performance of existing hybrid fuzzers may not well generalize to other experimental settings. Meanwhile, their performance advantages over conventional coverage-guided fuzzers are overall limited. In addition, instead of simply updating the fuzzing strategies or concolic executors, updating their coordination modes potentially poses crucial performance impact of hybrid fuzzers. Accordingly, we propose Cohuzz to improve the effectiveness of hybrid fuzzers by upgrading their coordination modes. Specifically, based on the baseline hybrid fuzzer QSYM, Cohuzz adopts \textit{edge-oriented scheduling} to schedule edges for applying concolic execution via an online linear regression model with Stochastic Gradient Descent. It also adopts \textit{sampling-augmenting synchronization} to derive seeds for applying fuzzing strategies via the interval path abstraction and John walk as well as incrementally updating the model. Our evaluation results indicate that Cohuzz can significantly increase the edge coverage (e.g., 16.31% higher than the best existing hybrid fuzzer in our study) and expose around 2X more unique crashes than all studied hybrid fuzzers. Moreover, Cohuzz successfully detects 37 previously unknown bugs where 30 are confirmed with 8 new CVEs and 20 are fixed.

Sydr-Fuzz: Continuous Hybrid Fuzzing and Dynamic Analysis for Security Development Lifecycle (ISPRAS Open 2022)

Paper
Slides

Abstract Nowadays automated dynamic analysis frameworks for continuous testing are in high demand to ensure software safety and satisfy the security development lifecycle (SDL) requirements. The security bug hunting efficiency of cutting-edge hybrid fuzzing techniques outperforms widely utilized coverage-guided fuzzing. We propose an enhanced dynamic analysis pipeline to leverage productivity of automated bug detection based on hybrid fuzzing. We implement the proposed pipeline in the continuous fuzzing toolset Sydr-Fuzz which is powered by hybrid fuzzing orchestrator, integrating our DSE tool Sydr with libFuzzer and AFL++. Sydr-Fuzz also incorporates security predicate checkers, crash triaging tool Casr, and utilities for corpus minimization and coverage gathering. The benchmarking of our hybrid fuzzer against alternative state-of-the-art solutions demonstrates its superiority over coverage-guided fuzzers while remaining on the same level with advanced hybrid fuzzers. Furthermore, we approve the relevance of our approach by discovering 85 new real-world software flaws within the OSS-Sydr-Fuzz project. Finally, we open Casr source code to the community to facilitate examination of the existing crashes.

TensileFuzz: Facilitating Seed Input Generation in Fuzzing via String Constraint Solving (ISSTA 2022)

Paper

Abstract Seed inputs are critical to the performance of mutation based fuzzers. Existing techniques make use of symbolic execution and gradient descent to generate seed inputs. However, these techniques are not particular suitable for input growth (i.e., making input longer and longer), a key step in seed input generation. Symbolic execution models very low level constraints and prefer fix-sized inputs whereas gradient descent only handles cases where path conditions are arithmetic functions of inputs. We observe that growing an input requires considering a number of relations: length, offset, and count, in which a field is the length of another field, the offset of another field, and the count of some pattern in another field, respective. Theory of string solver is particularly suitable for addressing these relations. We hence propose a novel technique called TensileFuzz, in which we identify input fields and denote them as string variables such that a seed input is the concatenation of these string variables. Additional padding string variables are inserted in between field variables. The aforementioned relations are reverse-engineered and lead to string constraints, solving which instantiates the padding variables and hence grows the input. Our technique also integrates linear regression and gradient descent to ensure the grown inputs satisfy path constraints that lead to path exploration. Our comparison with AFL, and a number of state-of-the-art fuzzers that have similar target applications, including Qsym, Angora, and SLF, shows that TensileFuzz substantially outperforms the others, by 39% - 98% in terms of path coverage.

CONFETTI: Amplifying Concolic Guidance for Fuzzers (ICSE 2022)

Paper

Abstract Fuzz testing (fuzzing) allows developers to detect bugs and vulnerabilities in code by automatically generating defect-revealing inputs. Most fuzzers operate by generating inputs for applications and mutating the bytes of those inputs, guiding the fuzzing process with branch coverage feedback via instrumentation. Whitebox guidance (e.g., taint tracking or concolic execution) is sometimes integrated with coverage-guided fuzzing to help cover tricky-to-reach branches that are guarded by complex conditions (so-called magic values). This integration typically takes the form of a targeted input mutation, eg placing particular byte values at a specific offset of some input in order to cover a branch. However, these dynamic analysis techniques are not perfect in practice, which can result in the loss of important relationships between input bytes and branch predicates, thus reducing the effective power of the technique. We introduce a new, surprisingly simple, but effective technique, global hinting, which allows the fuzzer to insert these interesting bytes not only at a targeted position, but in any position of any input. We implemented this idea in Java, creating CONFETTI, which uses both targeted and global hints for fuzzing. In an empirical comparison with two baseline approaches, a state-of-the-art greybox Java fuzzer and a version of CONFETTI without global hinting, we found that CONFETTI covers more branches and finds 15 previously unreported bugs, including 9 that neither baseline could find. By conducting a forensic analysis of CONFETTI's execution, we determined that global hinting was at least as effective at revealing new coverage as traditional, targeted hinting.

FuSeBMC v. 4: Smart Seed Generation for Hybrid Fuzzing (2021)

Paper

Abstract FuSeBMC is a test generator for finding security vulnerabilities in C programs. In earlier work [4], we described a previous version that incrementally injected labels to guide Bounded Model Checking (BMC) and Evolutionary Fuzzing engines to produce test cases for code coverage and bug finding. This paper introduces a new version of FuSeBMC that utilizes both engines to produce smart seeds. First, the engines are run with a short time limit on a lightly instrumented version of the program to produce the seeds. The BMC engine is particularly useful in producing seeds that can pass through complex mathematical guards. Then, FuSeBMC runs its engines with more extended time limits using the smart seeds created in the previous round. FuSeBMC manages this process in two main ways using its Tracer subsystem. Firstly, it uses shared memory to record the labels covered by each test case. Secondly, it evaluates test cases, and those of high impact are turned into seeds for subsequent test fuzzing. As a result, we significantly increased our code coverage score from last year, outperforming all tools that participated in this year's competition in every single category.

A Tight Integration of Symbolic Execution and Fuzzing (short paper 2021)

Paper

Abstract Most bug finding tools rely on either fuzzing or symbolic execution. While they both work well in some situations, fuzzing struggles with complex conditions and symbolic execution suffers from path explosion and high constraint solving costs. In order to enjoy the advantages from both techniques, we propose a new approach called Lightweight Symbolic Execution (LSE) that integrates well with fuzzing. Especially, LSE does not require any call to a constraint solver and allows for quickly enumerating inputs. In this short paper, we present the basic concepts of LSE together with promising preliminary experiments.

Symbolic Security Predicates: Hunt Program Weaknesses (ISPRAS Open 2021)

Paper
Slides

Abstract Dynamic symbolic execution (DSE) is a powerful method for path exploration during hybrid fuzzing and automatic bug detection. We propose security predicates to effectively detect undefined behavior and memory access violation errors. Initially, we symbolically execute program on paths that don't trigger any errors (hybrid fuzzing may explore these paths). Then we construct a symbolic security predicate to verify some error condition. Thus, we may change the program data flow to entail null pointer dereference, division by zero, out-of-bounds access, or integer overflow weaknesses. Unlike static analysis, dynamic symbolic execution does not only report errors but also generates new input data to reproduce them. Furthermore, we introduce function semantics modeling for common C/C++ standard library functions. We aim to model the control flow inside a function with a single symbolic formula. This assists bug detection, speeds up path exploration, and overcomes overconstraints in path predicate. We implement the proposed techniques in our dynamic symbolic execution tool Sydr. Thus, we utilize powerful methods from Sydr such as path predicate slicing that eliminates irrelevant constraints.

We present Juliet Dynamic to measure dynamic bug detection tools accuracy. The testing system also verifies that generated inputs trigger sanitizers. We evaluate Sydr accuracy for 11 CWEs from Juliet test suite. Sydr shows 95.59% overall accuracy. We make Sydr evaluation artifacts publicly available to facilitate results reproducibility.

Towards Symbolic Pointers Reasoning in Dynamic Symbolic Execution (IVMEM 2021)

Paper
Slides

Abstract Dynamic symbolic execution is a widely used technique for automated software testing, designed for execution paths exploration and program errors detection. A hybrid approach has recently become widespread, when the main goal of symbolic execution is helping fuzzer increase program coverage. The more branches symbolic executor can invert, the more useful it is for fuzzer. A program control flow often depends on memory values, which are obtained by computing address indexes from user input. However, most DSE tools don't support such dependencies, so they miss some desired program branches. We implement symbolic addresses reasoning on memory reads in our dynamic symbolic execution tool Sydr. Possible memory access regions are determined by either analyzing memory address symbolic expressions, or binary searching with SMT-solver. We propose an enhanced linearization technique to model memory accesses.

Different memory modeling methods are compared on the set of programs. Our evaluation shows that symbolic addresses handling allows to discover new symbolic branches and increase the program coverage.

FUZZOLIC: Mixing fuzzing and concolic execution (Computers&Security 2021)

Paper

Abstract: In the last few years, a large variety of approaches and methodologies have been explored in the context of software testing, ranging from black-box techniques, such as fuzzing, to white-box techniques, such as concolic execution, with a full spectrum of instances in between. Using these techniques, developers and security researchers have been able to identify in the last decade a large number of critical vulnerabilities in thousands of software projects.

In this article, we investigate how to improve the performance and effectiveness of concolic execution, proposing two main enhancements to the original approach. On one side, we devise a novel concolic executor that can analyze complex binary programs while running under QEMU and efficiently produce symbolic queries, which could generate valuable program inputs when solved. On the other side, we investigate whether techniques borrowed from the fuzzing domain can be applied to solve the symbolic queries generated by concolic execution, providing a viable alternative to accurate but expensive SMT solving techniques. We show that the combination of our concolic engine, Fuzzolic, and our approximate solver, Fuzzy-Sat, can perform better in terms of code coverage than popular state-of-the-art fuzzers on a variety of complex programs and can identify different unknown bugs in several real-world applications.

Concolic-Fuzzing of JavaScript Programs using GraalVM and Truffle (SKILL 2021)

Paper

Abstract: The scripting language JavaScript has established itself as a central component of the modern internet. However, the dynamic execution model of the language limits the support for source-code analysis, which leaves a developer without essential tools to maintain safety and security requirements. This paper describes a concolic-fuzzer based on the GraalVM to automatically test JavaScript programs. The fuzzer shows promising results in both code coverage and runtime evaluations and provides developers with additional features such as special analysis targets.

SHFuzz: A hybrid fuzzing method assisted by static analysis for binary programs (China Communications 2021)

Paper

Abstract: Fuzzing is an effective technique to find security bugs in programs by quickly exploring the input space of programs. To further discover vulnerabilities hidden in deep execution paths, the hybrid fuzzing combines fuzzing and concolic execution for going through complex branch conditions. In general, we observe that the execution path which comes across more and complex basic blocks may have a higher chance of containing a security bug. Based on this observation, we propose a hybrid fuzzing method assisted by static analysis for binary programs. The basic idea of our method is to prioritize seed inputs according to the complexity of their associated execution paths. For this purpose, we utilize static analysis to evaluate the complexity of each basic block and employ the hardware trace mechanism to dynamically extract the execution path for calculating the seed inputs' weights. The key advantage of our method is that our system can test binary programs efficiently by using the hardware trace and hybrid fuzzing. To evaluate the effectiveness of our method, we design and implement a prototype system, namely SHFuzz. The evaluation results show SHFuzz discovers more unique crashes on several real-world applications and the LAVA-M dataset when compared to the previous solutions.

A Priority Based Path Searching Method for Improving Hybrid Fuzzing (Computers & Security 2021)

Paper

Abstract: Hybrid fuzzing which combines classical fuzzing with concolic execution to produce effective test suites is an advanced software vulnerability detection technique. Because fuzzing and concolic execution are complementary in nature, some researchers propose 'optimal strategy' and 'discriminative dispatch strategy' to improve the performance of hybrid fuzzing. Although the ideas are interesting and useful, they have some limitations, such as high time overhead and difficulties in implementation. In this paper, we propose a Priority Based Path Searching method (PBPS) to utilize the capability of concolic execution better. PBPS evaluates each path's solving cost and solving demand, and prioritizes them based on two path characteristics, which are path lengths and sample-hits for concolic execution. The rationale is to keep the pipeline full by readily feeding the concolic engine with paths whose constraints are simpler to solve and are less likely to be explored by fuzz testing. We implement PBPS in Driller, which is a popular hybrid fuzzer and we evaluate our system 'QuickFuzz' with the CQE dataset. Experimental results show that compared with DigFuzz and the original Driller, 'QuickFuzz' discovers more vulnerabilities and achieves higher code coverage on the CQE dataset.

Sydr: Cutting Edge Dynamic Symbolic Execution (ISPRAS Open 2020)

Abstract The security development lifecycle (SDL) is becoming an industry standard. Dynamic symbolic execution (DSE) has enormous amount of applications in computer security (fuzzing, vulnerability discovery, reverse-engineering, etc.). We propose several performance and accuracy improvements for dynamic symbolic execution. Skipping non-symbolic instructions allows to build a path predicate 1.2--3.5 times faster. Symbolic engine simplifies formulas during symbolic execution. Path predicate slicing eliminates irrelevant conjuncts from solver queries. We handle each jump table (switch statement) as multiple branches and describe the method for symbolic execution of multi-threaded programs. The proposed solutions were implemented in Sydr tool. Sydr performs inversion of branches in path predicate. Sydr combines DynamoRIO dynamic binary instrumentation tool with Triton symbolic engine. We evaluated Sydr features on 64-bit Linux executables.

CSEFuzz: Fuzz Testing Based on Symbolic Execution (Access 2020)

Paper

Abstract: Fuzz testing has been successful in finding defects of various software packages. These defects include file parsing, image processing, Internet browsers, and network protocols. However, the quality of the initial seed test cases greatly influences the coverage and defect detection capability of fuzz testing. To address this issue, we propose CSEFuzz, a fuzz testing approach based on symbolic execution for defect detection. First, CSEFuzz generates candidate test cases by symbolic execution and collects coverage information of the test cases. Then, CSEFuzz extracts the test-case templates of the test cases and selects a set of test-case templates according to specific coverage criteria. Finally, CSEFuzz selects test cases according to the selected test-case templates, and the selected test cases are used as initial seed test cases for fuzz testing. Experiments are conducted on 11 open-source programs. The results show that in comparison with afl-cmin, which is the test-case selection command of Kelinci, CSEFuzz with a path coverage criterion reduces the time costs of the initial seed test selection and verification by 94.26%. In addition, compared with afl-cmin, 32 more paths are covered and 16 more defects are detected by CSEFuzz

Sequence directed hybrid fuzzing (SANER 2020)

Paper

Abstract: Existing directed grey-box fuzzers are effective compared with coverage-based fuzzers. However, they fail to achieve a balance between effectiveness and efficiency, and it is difficult to cover complex paths due to random mutation. To mitigate the issue, we propose a novel approach, sequence directed hybrid fuzzing (SDHF), which leverages a sequence-directed strategy and concolic execution technique to enhance the effectiveness of fuzzing. Given a set of target statement sequences of a program, SDHF aims to generate inputs that can reach the statements in each sequence in order and trigger potential bugs in the program. We implement the proposed approach in a tool called Berry and evaluate its capability on crash reproduction, true positive verification, and vulnerability detection. Experimental results demonstrate that Berry outperforms four state-of-the-art fuzzers, including directed fuzzers BugRedux, AFLGo and Lolly, and undirected hybrid fuzzer QSYM. Moreover, Berry found 7 new vulnerabilities in real-world programs such as UPX and GNU Libextractor, and 3 new CVEs were assigned.

HFL: Hybrid Fuzzing on the Linux Kernel (NDSS 2020)

Abstract: Hybrid fuzzing, combining symbolic execution and fuzzing, is a promising approach for vulnerability discovery because each approach can complement the other. However, we observe that applying hybrid fuzzing to kernel testing is challenging because the following unique characteristics of the kernel make a naive adoption of hybrid fuzzing inefficient: 1) having many implicit control transfers determined by syscall arguments, 2) controlling and matching internal system state via system calls, and 3) inferring nested argument type for invoking system calls. Failure to handling such challenges will render both fuzzing and symbolic execution inefficient, and thereby, will result in an inefficient hybrid fuzzing. Although these challenges are essential to both fuzzing and symbolic execution, however, to the best of our knowledge, existing kernel testing approaches either naively use each technique separately without handling such challenges or imprecisely handle a part of challenges only by static analysis.

To this end, this paper proposes HFL, which not only combines fuzzing with symbolic execution for hybrid fuzzing but also addresses kernel-specific fuzzing challenges via three distinct features: 1) converting implicit control transfers to explicit transfers, 2) inferring system call sequence to build a consistent system state, and 3) identifying nested arguments types of system calls. As a result, HFL found 24 previously unknown vulnerabilities in recent Linux kernels. Additionally, HFL achieves 14% higher code coverage than Syzkaller, and over S2E/TriforceAFL, achieving even eight times better coverage, using the same amount of resource (CPU, time, etc.). Regarding vulnerability discovery performance, HFL found 13 known vulnerabilities more than three times faster than Syzkaller.

PANGOLIN: Incremental Hybrid Fuzzing with Polyhedral Path Abstraction (S&P 2020)

Abstract: Hybrid fuzzing, which combines the merits of both fuzzing and concolic execution, has become one of the most important trends in coverage-guided fuzzing techniques. Despite the tremendous research on hybrid fuzzers, we observe that existing techniques are still inefficient. One important reason is that these techniques, which we refer to as non-incremental fuzzers, cache and reuse few computation results and, thus, lose many optimization opportunities. To be incremental, we propose 'polyhedral path abstraction', which preserves the exploration state in the concolic execution stage and allows more effective mutation and constraint solving over existing techniques. We have implemented our idea as a tool, namely PANGOLIN, and evaluated it using LAVA-M as well as nine real-world programs. The evaluation results showed that PANGOLIN outperforms the state-of-the-art fuzzing techniques with the improvement of coverage rate ranging from 10% to 30%. Moreover, PANGOLIN found 400 more bugs in LAVA-M and discovered 41 unseen bugs with 8 of them assigned with the CVE IDs.

SAVIOR: Towards Bug-Driven Hybrid Testing (S&P 2020)

Abstract: Hybrid testing combines fuzz testing and concolic execution. It leverages fuzz testing to test easy-to-reach code regions and uses concolic execution to explore code blocks guarded by complex branch conditions. As a result, hybrid testing is able to reach deeper into program state space than fuzz testing or concolic execution alone. Recently, hybrid testing has seen significant advancement. However, its code coveragecentric design is inefficient in vulnerability detection. First, it blindly selects seeds for concolic execution and aims to explore new code continuously. However, as statistics shows, a large portion of the explored code is often invulnerable. Therefore, giving equal attention to every part of the code during hybrid testing is a non-optimal strategy. It also slows down the detection of real vulnerabilities by over 43%. Second, classic hybrid testing quickly moves on after reaching a chunk of code, rather than examining the hidden defects inside. It may frequently miss subtle yet exploitable vulnerabilities despite that it has already explored the vulnerable code paths.

We propose SAVIOR, a new hybrid testing framework pioneering a bug-driven principle. Unlike the existing hybrid testing tools, SAVIOR prioritizes the concolic execution of the seeds that are likely to uncover more vulnerabilities. Moreover, SAVIOR verifies all vulnerable program locations along the executing program path. By modeling faulty situations using SMT constraints, SAVIOR reasons the feasibility of vulnerabilities and generates concrete test cases as proofs. Our evaluation shows that the bugdriven approach outperforms the mainstream automated testing techniques, including the state-of-the-art hybrid testing driven by code coverage. On average, SAVIOR detects vulnerabilities 43.4% faster than DRILLER and 44.3% faster than QSYM, leading to the discovery of 88 and 76 more security violations, respectively. According to the experimental result on 11 well-fuzzed benchmark programs, SAVIOR triggers 481 unique security violations within the first 24 hours.

Deferred Concretization in Symbolic Execution via Fuzzing (ISSTA 2019)

Paper

Abstract: Concretization is an effective weapon in the armory of symbolic execution engines. However, concretization can lead to loss in coverage, path divergence, and generation of test-cases on which the intended bugs are not reproduced. In this paper, we propose an algorithm, Deferred Concretization, that uses a new category for values within symbolic execution (referred to as the symcrete values) to pend concretization till they are actually needed. Our tool, COLOSSUS, built around these ideas, was able to gain an average coverage improvement of 66.94% and reduce divergence by more than 55% relative to the state-of-the-art symbolic execution engine, KLEE. Moreover, we found that KLEE loses about 38.60% of the states in the symbolic execution tree that COLOSSUS is able to recover, showing that COLOSSUS is capable of covering a much larger coverage space.

Send Hardest Problems My Way: Probabilistic Path Prioritization for Hybrid Fuzzing (NDSS 2019)

Abstract: Hybrid fuzzing which combines fuzzing and concolic execution has become an advanced technique for software vulnerability detection. Based on the observation that fuzzing and concolic execution are complementary in nature, the state-of-the-art hybrid fuzzing systems deploy "demand launch" and "optimal switch" strategies. Although these ideas sound intriguing, we point out several fundamental limitations in them, due to oversimplified assumptions. We then propose a novel "discriminative dispatch" strategy to better utilize the capability of concolic execution. We design a novel Monte Carlo based probabilistic path prioritization model to quantify each path's difficulty and prioritize them for concolic execution. This model treats fuzzing as a random sampling process. It calculates each path's probability based on the sampling information. Finally, our model prioritizes and assigns the most difficult paths to concolic execution. We implement a prototype system DigFuzz and evaluate our system with two representative datasets. Results show that the concolic execution in DigFuzz outperforms than that in a state-of-the-art hybrid fuzzing system Driller in every major aspect. In particular, the concolic execution in DigFuzz contributes to discovering more vulnerabilities (12 vs. 5) and producing more code coverage (18.9% vs. 3.8%) on the CQE dataset than the concolic execution in Driller.

Intriguer: Field-Level Constraint Solving for Hybrid Fuzzing (CCS 2019)

Abstract: Hybrid fuzzing, which combines fuzzing and concolic execution, is promising in light of the recent performance improvements in concolic engines. We have observed that there is room for further improvement: symbolic emulation is still slow, unnecessary constraints dominate solving time, resources are overly allocated, and hard-to-trigger bugs are missed. To address these problems, we present a new hybrid fuzzer named Intriguer. The key idea of Intriguer is field-level constraint solving, which optimizes symbolic execution with field-level knowledge. Intriguer performs instruction-level taint analysis and records execution traces without data transfer instructions like mov. Intriguer then reduces the execution traces for tainted instructions that accessed a wide range of input bytes, and infers input fields to build field transition trees. With these optimizations, Intriguer can efficiently perform symbolic emulation for more relevant instructions and invoke a solver for complicated constraints only. Our evaluation results indicate that Intriguer outperforms the state-of-the-art fuzzers: Intriguer found all the bugs in the LAVA-M(5h) benchmark dataset for ground truth performance, and also discovered 43 new security bugs in seven real-world programs. We reported the bugs and received 23 new CVEs.

DeepFuzzer: Accelerated Deep Greybox Fuzzing (TDSC 2019)

Paper

Abstract: Fuzzing is one of the most effective vulnerability detection techniques, widely used in practice. However, the performance of fuzzers may be limited by their inability to pass complicated checks, inappropriate mutation frequency, arbitrary mutation strategy, or the variability of the environment. In this paper, we present DeepFuzzer, an enhanced greybox fuzzer with qualified seed generation, balanced seed selection, and hybrid seed mutation. First, we use symbolic execution in a lightweight approach to generate qualified initial seeds which then guide the fuzzer through complex checks. Second, we apply a statistical seed selection algorithm to balance the mutation frequency between different seeds. Further, we develop a hybrid mutation strategy. The random and restricted mutation strategies are combined to maintain a dynamic balance between global exploration and deep search. We evaluate DeepFuzzer on the widely used benchmark Google fuzzer-test-suite which consists of real-world programs. Compared with AFL, AFLFast, FairFuzz, QSYM, and MOPT in the 24-hour experiment, DeepFuzzer discovers 30%, 240%, 102%, 147%, and 257% more unique crashes, executes 40%, 36%, 36%, 98%, and 15% more paths, and covers 37%, 34%, 34%, 101%, and 11% more branches, respectively. Furthermore, we present the practice of fuzzing a message middleware from Huawei with DeepFuzzer, and 9 new vulnerabilities are reported.

QSYM: A Practical Concolic Execution Engine Tailored for Hybrid Fuzzing (USENIX Security2018)

Abstract: Recently, hybrid fuzzing has been proposed to address the limitations of fuzzing and concolic execution by combining both approaches. The hybrid approach has shown its effectiveness in various synthetic benchmarks such as DARPA Cyber Grand Challenge (CGC) binaries, but it still suffers from scaling to find bugs in complex, real-world software. We observed that the performance bottleneck of the existing concolic executor is the main limiting factor for its adoption beyond a small-scale study. To overcome this problem, we design a fast concolic execution engine, called QSYM, to support hybrid fuzzing. The key idea is to tightly integrate the symbolic emulation with the native execution using dynamic binary translation, making it possible to implement more fine-grained, so faster, instruction-level symbolic emulation. Additionally, QSYM loosens the strict soundness requirements of conventional concolic executors for better performance, yet takes advantage of a faster fuzzer for validation, providing unprecedented opportunities for performance optimizations, e.g., optimistically solving constraints and pruning uninteresting basic blocks. Our evaluation shows that QSYM does not just outperform state-of-the-art fuzzers (i.e., found 14× more bugs than VUzzer in the LAVA-M dataset, and outperformed Driller in 104 binaries out of 126), but also found 13 previously unknown security bugs in eight real-world programs like Dropbox Lepton, ffmpeg, and OpenJPEG, which have already been intensively tested by the state-of-the-art fuzzers, AFL and OSS-Fuzz.

Angora: Efficient Fuzzing by Principled Search (S&P 2018)

Abstract: Abstract-Fuzzing is a popular technique for finding software bugs. However, the performance of the state-of-the-art fuzzers leaves a lot to be desired. Fuzzers based on symbolic execution produce quality inputs but run slow, while fuzzers based on random mutation run fast but have difficulty producing quality inputs. We propose Angora, a new mutation-based fuzzer that outperforms the state-of-the-art fuzzers by a wide margin. The main goal of Angora is to increase branch coverage by solving path constraints without symbolic execution. To solve path constraints efficiently, we introduce several key techniques: scalable byte-level taint tracking, context-sensitive branch count, search based on gradient descent, and input length exploration. On the LAVA-M data set, Angora found almost all the injected bugs, found more bugs than any other fuzzer that we compared with, and found eight times as many bugs as the second-best fuzzer in the program who. Angora also found 103 bugs that the LAVA authors injected but could not trigger. We also tested Angora on eight popular, mature open source programs. Angora found 6, 52, 29, 40 and 48 new bugs in file, jhead, nm, objdump and size, respectively. We measured the coverage of Angora and evaluated how its key techniques contribute to its impressive performance.

SAFL: increasing and accelerating testing coverage with symbolic execution and guided fuzzing (ICSE 2018)

Paper

Abstract: Mutation-based fuzzing is a widely used software testing technique for bug and vulnerability detection, and the testing performance is greatly affected by the quality of initial seeds and the effectiveness of mutation strategy. In this paper, we present SAFL, an efficient fuzzing testing tool augmented with qualified seed generation and efficient coverage-directed mutation. First, symbolic execution is used in a lightweight approach to generate qualified initial seeds. Valuable explore directions are learned from the seeds, thus the later fuzzing process can reach deep paths in program state space earlier and easier. Moreover, we implement a fair and fast coverage-directed mutation algorithm. It helps the fuzzing process to exercise rare and deep paths with higher probability. We implement SAFL based on KLEE and AFL and conduct thoroughly repeated evaluations on real-world program benchmarks against state-of-the-art versions of AFL. After 24 hours, compared to AFL and AFLFast, it discovers 214% and 133% more unique crashes, covers 109% and 63% more paths and achieves 279% and 180% more covered branches. Video link: https://youtu.be/LkiFLNMBhVE

CAB-Fuzz: Practical Concolic Testing Techniques for COTS Operating Systems (Usenix Security2017)

Paper
Slides

Abstract: Discovering the security vulnerabilities of commercial off-the-shelf (COTS) operating systems (OSes) is challenging because they not only are huge and complex, but also lack detailed debug information. Concolic testing, which generates all feasible inputs of a program by using symbolic execution and tests the program with the generated inputs, is one of the most promising approaches to solve this problem. Unfortunately, the state-of-the-art concolic testing tools do not scale well for testing COTS OSes because of state explosion. Indeed, they often fail to find a single bug (or crash) in COTS OSes despite their long execution time. In this paper, we propose CAB-FUZZ (Context-Aware and Boundary-focused), a practical concolic testing tool to quickly explore interesting paths that are highly likely triggering real bugs without debug information. First, CAB-FUZZ prioritizes the boundary states of arrays and loops, inspired by the fact that many vulnerabilities originate from a lack of proper boundary checks. Second, CAB-FUZZ exploits real programs interacting with COTS OSes to construct proper contexts to explore deep and complex kernel states without debug information. We applied CAB-FUZZ to Windows 7 and Windows Server 2008 and found 21 undisclosed unique crashes, including two local privilege escalation vulnerabilities (CVE2015-6098 and CVE-2016-0040) and one information disclosure vulnerability in a cryptography driver (CVE2016-7219). CAB-FUZZ found vulnerabilities that are non-trivial to discover; five vulnerabilities have existed for 14 years, and we could trigger them even in the initial version of Windows XP (August 2001).

Driller: Argumenting Fuzzing Through Selective Symbolic Execution (NDSS 2016)

Abstract: Memory corruption vulnerabilities are an everpresent risk in software, which attackers can exploit to obtain unauthorized access to confidential information. As products with access to sensitive data are becoming more prevalent, the number of potentially exploitable systems is also increasing, resulting in a greater need for automated software vetting tools. DARPA recently funded a competition, with millions of dollars in prize money, to further research focusing on automated vulnerability finding and patching, showing the importance of research in this area. Current techniques for finding potential bugs include static, dynamic, and concolic analysis systems, which each having their own advantages and disadvantages. A common limitation of systems designed to create inputs which trigger vulnerabilities is that they only find shallow bugs and struggle to exercise deeper paths in executables. We present Driller, a hybrid vulnerability excavation tool which leverages fuzzing and selective concolic execution in a complementary manner, to find deeper bugs. Inexpensive fuzzing is used to exercise compartments of an application, while concolic execution is used to generate inputs which satisfy the complex checks separating the compartments. By combining the strengths of the two techniques, we mitigate their weaknesses, avoiding the path explosion inherent in concolic analysis and the incompleteness of fuzzing. Driller uses selective concolic execution to explore only the paths deemed interesting by the fuzzer and to generate inputs for conditions that the fuzzer cannot satisfy. We evaluate Driller on 126 applications released in the qualifying event of the DARPA Cyber Grand Challenge and show its efficacy by identifying the same number of vulnerabilities, in the same time, as the top-scoring team of the qualifying event.

Hybrid Fuzz Testing - Discovering Software Bugs via Fuzzing and Symbolic Execution (2012)

Paper

Abstract: Random mutational fuzz testing (fuzzing) and symbolic executions are program testing techniques that have been gaining popularity in the security research community. Fuzzing finds bugs in a target program by natively executing it with random inputs while monitoring the execution for abnormal behaviors such as crashes. While fuzzing may have a reputation of being able to explore deep into a program's state space efficiently, naive fuzzers usually have limited code coverage for typical programs since unconstrained random inputs are unlikely to drive the execution down many different paths. In contrast, symbolic execution tests a program by treating the program's input as symbols and interpreting the program over such symbolic inputs. Although in theory symbolic execution is guaranteed to be effective in achieving code coverage if we explore all possible paths, this generally requires exponential resource and is thus not practical for many real-world programs.

This thesis presents our attempt to attain the best of both worlds by combining fuzzing with symbolic execution in a novel manner. Our technique, called hybrid fuzzing, first uses symbolic execution to discover frontier nodes that represent unique paths in the program. After collecting as many frontier nodes as possible under a user-specifiable resource constraint, it transits to fuzz the program with preconditioned random inputs, which are provably random inputs that respect the path predicate leading to each frontier node. Our current implementation supports programs with linear path predicates and can automatically generate preconditioned random inputs from a polytope model of the input space extracted from binaries. These preconditioned random inputs can then be used with any fuzzer. Experiments show that our implementation is efficient in both time and space, and the inputs generated by it are able to gain extra breadth and depth over previous approaches.

Hybrid concolic testing (2007)

Paper

Abstract: We present hybrid concolic testing, an algorithm that interleaves random testing with concolic execution to obtain both a deep and a wide exploration of program state space. Our algorithm generates test inputs automatically by interleaving random testing until saturation with bounded exhaustive symbolic exploration of program points. It thus combines the ability of random search to reach deep program states quickly together with the ability of concolic testing to explore states in a neighborhood exhaustively. We have implemented our algorithm on top of CUTE and applied it to obtain better branch coverage for an editor implementation (VIM 5.7, 150K lines of code) as well as a data structure implementation in C. Our experiments suggest that hybrid concolic testing can handle large programs and provide, for the same testing budget, almost 4× the branch coverage than random testing and almost 2× that of concolic testing.

Mutation\Coverage\Path

GrayC: Greybox Fuzzing of Compilers and Analysers for C (ISSTA 2023)

Paper
Code

Abstract: Fuzzing of compilers and code analysers has led to a large number of bugs being found and fixed in widely-used frameworks such as LLVM, GCC and Frama-C. Most such fuzzing techniques have taken a blackbox approach, with compilers and code analysers starting to become relatively immune to such fuzzers.

We propose a coverage-directed, mutation-based approach for fuzzing C compilers and code analysers, inspired by the success of this type of greybox fuzzing in other application domains. The main challenge of applying mutation-based fuzzing in this context is that naive mutations are likely to generate programs that do not compile. Such programs are not useful for finding deep bugs that affect optimisation, analysis, and code generation routines.

We have designed a novel greybox fuzzer for C compilers and analysers by developing a new set of mutations to target common C constructs, and transforming fuzzed programs so that they produce meaningful output, allowing differential testing to be used as a test oracle, and paving the way for fuzzer-generated programs to be integrated into compiler and code analyser regression test suites.

We have implemented our approach in GrayC, a new open-source LibFuzzer-based tool, and present experiments showing that it provides more coverage on the middle- and back-end stages of compilers and analysers compared to other mutation-based approaches, including Clang-Fuzzer, PolyGlot, and a technique similar to LangFuzz.

We have used GrayC to identify 30 confirmed compiler and code analyser bugs: 25 previously unknown bugs (with 22 of them already fixed in response to our reports) and 5 confirmed bugs reported independently shortly before we found them. A further 3 bug reports are under investigation. Apart from the results above, we have contributed 24 simplified versions of coverage-enhancing test cases produced by GrayC to the Clang/LLVM test suite, targeting 78 previously uncovered functions in the LLVM codebase.

DARWIN: Survival of the Fittest Fuzzing Mutators (NDSS 2023)

Paper

Abstract: Fuzzing is an automated software testing technique broadly adopted by the industry. A popular variant is mutation-based fuzzing, which discovers a large number of bugs in practice. While the research community has studied mutation-based fuzzing for years now, the algorithms' interactions within the fuzzer are highly complex and can, together with the randomness in every instance of a fuzzer, lead to unpredictable effects. Most efforts to improve this fragile interaction focused on optimizing seed scheduling. However, real-world results like Google's FuzzBench highlight that these approaches do not consistently show improvements in practice. Another approach to improve the fuzzing process algorithmically is optimizing mutation scheduling. Unfortunately, existing mutation scheduling approaches also failed to convince because of missing real-world improvements or too many user-controlled parameters whose configuration requires expert knowledge about the target program. This leaves the challenging problem of cleverly processing test cases and achieving a measurable improvement unsolved.

We present DARWIN, a novel mutation scheduler and the first to show fuzzing improvements in a realistic scenario without the need to introduce additional user-configurable parameters, opening this approach to the broad fuzzing community. DARWIN uses an Evolution Strategy to systematically optimize and adapt the probability distribution of the mutation operators during fuzzing. We implemented a prototype based on the popular general-purpose fuzzer AFL. DARWIN significantly outperforms the state-of-the-art mutation scheduler and the AFL baseline in our own coverage experiment, in FuzzBench, and by finding 15 out of 21 bugs the fastest in the MAGMA benchmark. Finally, DARWIN found 20 unique bugs (including one novel bug), 66% more than AFL, in widely-used real-world applications.

Rainfuzz: Reinforcement-Learning Driven Heat-Maps for Boosting Coverage-Guided Fuzzing (ICPRAM 2023)

Paper

Abstract: Fuzzing is a dynamic analysis technique that repeatedly executes the target program with many different inputs to trigger abnormal behavior, such as a crash. One of the most successful techniques consists in generating inputs to increase code-coverage by using a mutational approach: this type of fuzzers maintains a population of inputs, they perform mutations on the inputs in the current population, and they add mutated inputs to the population if they discover new code-coverage in the target program. Researchers are continuously looking for techniques to increment the efficiency of fuzzers; one of these techniques consists in generating heat-maps for targeting specific bytes during the mutation of the input, as not all bytes might be useful for controlling the program’s workflow. We propose the first approach in the literature that uses reinforcement learning for building heat-maps, by formalizing the problem of choosing the position to be mutated within the input as a reinforcement-learning problem. We model the policy by means of a neural network, and we train it by using Proximal Policy Optimization (PPO). We implement our approach in Rainfuzz, and we show the effectiveness of its heat-maps by comparing Rainfuzz against an equivalent fuzzer that performs mutations at random positions. We achieve the best performance by running AFL++ and Rainfuzz in parallel (in a collaborative fuzzing setting), outperforming a setting where we run two AFL++ instances in parallel.

Evaluating the Fork-Awareness of Coverage-Guided Fuzzers (ICISSP 2023)

Paper

Abstract: Fuzz testing (or fuzzing) is an effective technique used to find security vulnerabilities. It consists of feeding a software under test with malformed inputs, waiting for a weird system behaviour (often a crash of the system). Over the years, different approaches have been developed, and among the most popular lies the coverage-based one. It relies on the instrumentation of the system to generate inputs able to cover as much code as possible. The success of this approach is also due to its usability as fuzzing techniques research approaches that do not require (or only partial require) human interactions. Despite the efforts, devising a fully-automated fuzzer still seems to be a challenging task. Target systems may be very complex; they may integrate cryptographic primitives, compute and verify check-sums and employ forks to enhance the system security, achieve better performances or manage different connections at the same time. This paper introduces the fork-awareness property to express the fuzzer ability to manage systems using forks. This property is leveraged to evaluate 14 of the most widely coverage-guided fuzzers and highlight how current fuzzers are ineffective against systems using forks.

One Fuzzing Strategy to Rule Them All (ICSE 2022)

Paper

Abstract: Coverage-guided fuzzing has become mainstream in fuzzing to automatically expose program vulnerabilities. Recently, a group of fuzzers are proposed to adopt a random search mechanism namely Havoc, explicitly or implicitly, to augment their edge exploration. However, they only tend to adopt the default setup of Havoc as an implementation option while none of them attempts to explore its power under diverse setups or inspect its rationale for potential improvement. In this paper, to address such issues, we conduct the first empirical study on Havoc to enhance the understanding of its characteristics. Specifically, we first find that applying the default setup of Havoc to fuzzers can significantly improve their edge coverage performance. Interestingly, we further observe that even simply executing Havoc itself without appending it to any fuzzer can lead to strong edge coverage performance and outperform most of our studied fuzzers. Moreover, we also extend the execution time of Havoc and find that most fuzzers can not only achieve significantly higher edge coverage, but also tend to perform similarly (i.e., their performance gaps get largely bridged). Inspired by the findings, we further propose Havoc𝑀𝐴𝐵, which models the Havoc mutation strategy as a multi-armed bandit problem to be solved by dynamically adjusting the mutation strategy. The evaluation result presents that Havoc𝑀𝐴𝐵 can significantly increase the edge coverage by 11.1% on average for all the benchmark projects compared with Havoc and even slightly outperform state-of-the-art QSYM which augments its computing resource by adopting three parallel threads. We further execute Havoc𝑀𝐴𝐵 with three parallel threads and result in 9% higher average edge coverage over QSYM upon all the benchmark projects.

BeDivFuzz: Integrating Behavioral Diversity into Generator-based Fuzzing (ICSE 2022)

Paper

Abstract: A popular metric to evaluate the performance of fuzzers is branch coverage. However, we argue that focusing solely on covering many different branches (i.e., the richness) is not sufficient, since the majority of the covered branches may have been exercised only once, which does not inspire a high confidence in the reliability of the covered code. Instead, the distribution of the executed branches (i.e., the evenness) should be considered as well. That is, behavioral diversity is only given if the generated inputs not only trigger many different branches, but also trigger them evenly often with diverse inputs. We introduce BeDivFuzz, a feedback-driven fuzzing technique for generator-based fuzzers. BeDivFuzz distinguishes between structure-preserving and structure-changing mutations in the space of syntactically valid inputs, and biases its mutation strategy towards behavioral diversity based on the received program feedback. We have evaluated BeDivFuzz on Ant, Maven, Closure, Rhino, and Nashorn. The results show that BeDivFuzz achieves better behavioral diversity compared to the state of the art, measured by established biodiversity metrics from the field of ecology.

FuzzingDriver: the Missing Dictionary to Increase Code Coverage in Fuzzers (SANER 2022)

Paper

Abstract: We propose a tool, called FuzzingDriver, to generate dictionary tokens for coverage-based greybox fuzzers (CGF) from the codebase of any target program. FuzzingDriver does not add any overhead to the fuzzing job as it is run beforehand. We compared FuzzingDriver to Google dictionaries by fuzzing six open-source targets, and we found that FuzzingDriver consistently achieves higher code coverage in all tests. We also executed eight benchmarks on FuzzBench to demonstrate how utilizing FuzzingDriver's dictionaries can outperform six widely-used CGF fuzzers. In future work, investigating the impact of FuzzingDriver's dictionaries on improving bug coverage might prove important.

EMS: History-Driven Mutation for Coverage-based Fuzzing (NDSS 2022)

Code
Paper

Abstract: Mutation-based fuzzing is one of the most popular approaches to discover vulnerabilities in a program. To alleviate the inefficiency of mutation-based fuzzing incurred by high randomness in the mutation process, multiple solutions are developed in recent years, especially coverage-based fuzzing. They mainly employ adaptive mutation strategies or integrate constraint-solving techniques to make a good exploration of the test cases which trigger unique paths and crashes. However, they lack a fine-grained reusing of fuzzing history to construct these interesting test cases, i.e., they largely fail to properly utilize fuzzing history across different fuzzing trials. In fact, we discover that test cases in fuzzing history contain rich knowledge of the key mutation strategies that lead to the discovery of unique paths and crashes. Specifically, partial path constraint solutions implicitly carried in these mutation strategies can be reused to accelerate the discovery of new paths and crashes that share similar partial path constraints.

Therefore, we first propose a lightweight and efficient Probabilistic Byte Orientation Model (PBOM) that properly captures the byte-level mutation strategies from intra- and inter-trial history and thus can effectively trigger unique paths and crashes. We then present a novel history-driven mutation framework named EMS that employs PBOM as one of the mutation operators to probabilistically provide desired mutation byte values according to the input ones. We evaluate EMS against state-of-the-art fuzzers including AFL, QSYM, MOPT, MOPT-dict, EcoFuzz, and AFL++ on 9 real world programs. The results show that EMS discovers up to 4.91× more unique vulnerabilities than the baseline, and finds more line coverage than other fuzzers on most programs. We report all of the discovered new vulnerabilities to vendors and will open source the prototype of EMS on GitHub.

OTA: An Operation-oriented Time Allocation Strategy for Greybox Fuzzing (SANER 2021)

Paper

Abstract: Coverage-based greybox fuzzing (CGF) has been widely studied and commonly used for software vulnerability detection. Existing CGF fuzzers fairly allocate execution time for each mutation operation to generate test cases. However, the fair-time-allocation strategy is revealed to be inefficient by our significant experimental observation that different operations have heterogeneous effectiveness on coverage. Those ineffective operations with vast test cases thus occupy the majority of limited runtime, reducing the opportunities for effective operations to explore more paths and find potential vulnerabilities.In this paper, we propose a novel operation-oriented time allocation strategy OTA, which dynamically allocates operation execution time in real time to cope with the effectiveness variation per operation. OTA has three distinguishing advantages: (1) the execution time per operation is novelly initialized on demand and program-dependent; (2) the execution time for each operation is dynamically weighted by its real-time effectiveness on exploring new coverage; (3) the determination of the execution time per operation is well controlled to achieve a quick convergence. Extensive experiments based on real-world programs and the LAVA-M dataset have been conducted to evaluate the path discovery and vulnerability detection abilities of OTA, which substantially outperforms 5 state-of-the-art fuzzers. In addition, OTA exposes 18 previously unknown vulnerabilities in 6 well-tested programs with 13 confirmed with new CVE IDs.

MaxAFL: Maximizing Code Coverage with a Gradient-Based Optimization Technique (Electronics 2020)

Paper

Abstract: Evolutionary fuzzers generally work well with typical software programs because of their simple algorithm. However, there is a limitation that some paths with complex constraints cannot be tested even after long execution. Fuzzers based on concolic execution have emerged to address this issue. The concolic execution fuzzers also have limitations in scalability. Recently, the gradient-based fuzzers that use a gradient to mutate inputs have been introduced. Gradient-based fuzzers can be applied to real-world programs and achieve high code coverage. However, there is a problem that the existing gradient-based fuzzers require heavyweight analysis or sufficient learning time. In this paper, we propose a new type of gradient-based fuzzer, MaxAFL, to overcome the limitations of existing gradient-based fuzzers. Our approach constructs an objective function through fine-grained static analysis. After constructing a well-made objective function, we can apply the gradient-based optimization algorithm. We use a modified gradient-descent algorithm to minimize our objective function and propose some probabilistic techniques to escape local optimum. We introduce an adaptive objective function which aims to explore various paths in the program. We implemented MaxAFL based on the original AFL. MaxAFL achieved increase of code coverage per time compared with three other fuzzers in six open-source Linux binaries. We also measured cumulative code coverage per total execution, and MaxAFL outperformed the other fuzzers in this metric. Finally, MaxAFL can also find more bugs than the other fuzzers.

PathAFL: Path-Coverage Assisted Fuzzing (ASIA CCS 2020)

Paper

Abstract: Fuzzing is an effective method to find software bugs and vulnerabilities. One of the most useful techniques is the coverage-guided fuzzing, whose key element is the tracing code coverage information. Existing coverage-guided fuzzers generally use the the number of basic blocks or edges explored to measure code coverage. Path-coverage can provide more accurate coverage information than basic block and edge coverage. However, the number of paths grows exponentially as the size of a program increases. It is almost impossible to trace all the paths of a real-world application. In this paper, we propose a fuzzing solution named PathAFL, which assists a fuzzer by path identification. It can effectively identify and utilize the important h-path, which is a new path but whose edges have all been touched previously. First, PathAFL only inserts one assembly instruction to AFL's original code to calculate the path hash, and uses a selective instrumentation strategy to reduce the tracing granularity of an execution path. Second, we design a fast filtering algorithm to choose higher weight paths from a large number of h-paths and add them to the seed queue. Third, both the seed selection algorithm and the power schedule are implemented based on the path weight. Finally we implemented PathAFL based on the popular fuzzer AFL and evaluated it on 10 well-fuzzed benchmark programs. In 24 hours, PathAFL explored 38% more paths and 9.3% more edges than AFL. Compared with CollAFL-x, the number is 25% and 5.9% correspondingly. Moreover, PathAFL found the more bugs on the LAVA-M dataset, even four unlisted bugs. The results show that PathAFL outperforms the previous fuzzers in terms of both code coverage and bug discovery. In well-tested programs, PathAFL found 8 new security bugs with 6 CVEs assigned.

Zeror: Speed Up Fuzzing with Coverage-sensitive Tracing and Scheduling (ASE 2020)

Paper

Abstract: Coverage-guided fuzzing is one of the most popular software testing techniques for vulnerability detection. While effective, current fuzzing methods suffer from significant performance penalty due to instrumentation overhead, which limits its practical use. Existing solutions improve the fuzzing speed by decreasing instrumentation overheads but sacrificing coverage accuracy, which results in unstable performance of vulnerability detection.

In this paper, we propose a coverage-sensitive tracing and scheduling framework Zeror that can improve the performance of existing fuzzers, especially in their speed and vulnerability detection. The Zeror is mainly made up of two parts: (1) a self-modifying tracing mechanism to provide a zero-overhead instrumentation for more effective coverage collection, and (2) a real-time scheduling mechanism to support adaptive switch between the zero-overhead instrumented binary and the fully instrumented binary for better vulnerability detection. In this way, Zeror is able to decrease collection overhead and preserve fine-grained coverage for guidance.

For evaluation, we implement a prototype of Zeror and evaluate it on Google fuzzer-test-suite, which consists of 24 widely-used applications. The results show that Zeror performs better than existing fuzzing speed-up frameworks such as Untracer and INSTRIM, improves the execution speed of the state-of-the-art fuzzers such as AFL and MOPT by 159.80%, helps them achieve better coverage (averagely 10.14% for AFL, 6.91% for MOPT) and detect vulnerabilities faster (averagely 29.00% for AFL, 46.99% for MOPT)

Not All Coverage Measurements Are Equal: Fuzzing by Coverage Accounting for Input Prioritization (NDSS 2020)

Abstract: Coverage-based fuzzing has been actively studied and widely adopted for finding vulnerabilities in real-world software applications. With code coverage, such as statement coverage and transition coverage, as the guidance of input mutation, coverage-based fuzzing can generate inputs that cover more code and thus find more vulnerabilities without prerequisite information such as input format. Current coverage-based fuzzing tools treat covered code equally. All inputs that contribute to new statements or transitions are kept for future mutation no matter what the statements or transitions are and how much they impact security. Although this design is reasonable from the perspective of software testing, which aims to full code coverage, it is inefficient for vulnerability discovery since that 1) current techniques are still inadequate to reach full coverage within a reasonable amount of time, and that 2) we always want to discover vulnerabilities early so that it can be patched promptly. Even worse, due to the non-discriminative code coverage treatment, current fuzzing tools suffer from recent anti-fuzzing techniques and become much less effective in finding real-world vulnerabilities.

To resolve the issue, we propose coverage accounting, an innovative approach that evaluates code coverage by security impacts. Based on the proposed metrics, we design a new scheme to prioritize fuzzing inputs and develop TortoiseFuzz, a greybox fuzzer for memory corruption vulnerabilities. We evaluated TortoiseFuzz on 30 real-world applications and compared it with 5 state-of-the-art greybox and hybrid fuzzers (AFL, AFLFast, FairFuzz, QSYM, and Angora). TortoiseFuzz outperformed all greybox fuzzers and most hybrid fuzzers. It also had comparative results for other hybrid fuzzers yet consumed much fewer resources. Additionally, TortoiseFuzz found 18 new real-world vulnerabilities and has got 8 new CVEs so far. We will open source TortoiseFuzz to foster future research.

Matryoshka: fuzzing deeply nested branches (CCS 2019)

Abstract: Greybox fuzzing has made impressive progress in recent years, evolving from heuristics-based random mutation to approaches for solving individual path constraints. However, they have difficulty solving path constraints that involve deeply nested conditional statements, which are common in image and video decoders, network packet analyzers, and checksum tools. We propose an approach for addressing this problem. First, we identify all the control flow-dependent conditional statements of the target conditional statement. Next, we select the data flow-dependent conditional statements. Finally, we use three strategies to find an input that satisfies all conditional statements simultaneously. We implemented this approach in a tool called Matryoshka and compared its effectiveness on 13 open source programs against other state-of-the-art fuzzers. Matryoshka found significantly more unique crashes than AFL, QSYM, and Angora. We manually classified those crashes into 41 unique new bugs, and obtained 12 CVEs. Our evaluation also uncovered the key technique contributing to Matryoshka's impressive performance: it collects only the nesting constraints that may cause the target conditional statements unreachable, which greatly simplifies the constraints that it has to solve.

REDQUEEN: Fuzzing with Input-to-State Correspondence (NDSS2019)

Abstract: Automated software testing based on fuzzing has experienced a revival in recent years. Especially feedback-driven fuzzing has become well-known for its ability to efficiently perform randomized testing with limited input corpora. Despite a lot of progress, two common problems are magic numbers and (nested) checksums. Computationally expensive methods such as taint tracking and symbolic execution are typically used to overcome such roadblocks. Unfortunately, such methods often require access to source code, a rather precise description of the environment (e.g., behavior of library calls or the underlying OS), or the exact semantics of the platform's instruction set. In this paper, we introduce a lightweight, yet very effective alternative to taint tracking and symbolic execution to facilitate and optimize state-of-the-art feedback fuzzing that easily scales to large binary applications and unknown environments. We observe that during the execution of a given program, parts of the input often end up directly (i.e., nearly unmodified) in the program state. This input-to-state correspondence can be exploited to create a robust method to overcome common fuzzing roadblocks in a highly effective and efficient manner. Our prototype implementation, called REDQUEEN, is able to solve magic bytes and (nested) checksum tests automatically for a given binary executable. Additionally, we show that our techniques outperform various state-of-the-art tools on a wide variety of targets across different privilege levels (kernel-space and userland) with no platform-specific code. REDQUEEN is the first method to find more than 100% of the bugs planted in LAVA-M across all targets. Furthermore, we were able to discover 65 new bugs and obtained 16 CVEs in multiple programs and OS kernel drivers. Finally, our evaluation demonstrates that REDQUEEN is fast, widely applicable and outperforms concurrent approaches by up to three orders of magnitude.

T-Fuzz: fuzzing by program transformation (S&P 2018)

Abstract: Fuzzing is a simple yet effective approach to discover software bugs utilizing randomly generated inputs. However, it is limited by coverage and cannot find bugs hidden in deep execution paths of the program because the randomly generated inputs fail complex sanity checks, e.g., checks on magic values, checksums, or hashes. To improve coverage, existing approaches rely on imprecise heuristics or complex input mutation techniques (e.g., symbolic execution or taint analysis) to bypass sanity checks. Our novel method tackles coverage from a different angle: by removing sanity checks in the target program. T-Fuzz leverages a coverage-guided fuzzer to generate inputs. Whenever the fuzzer can no longer trigger new code paths, a light-weight, dynamic tracing based technique detects the input checks that the fuzzer-generated inputs fail. These checks are then removed from the target program. Fuzzing then continues on the transformed program, allowing the code protected by the removed checks to be triggered and potential bugs discovered. Fuzzing transformed programs to find bugs poses two challenges: (1) removal of checks leads to over-approximation and false positives, and (2) even for true bugs, the crashing input on the transformed program may not trigger the bug in the original program. As an auxiliary post-processing step, T-Fuzz leverages a symbolic execution-based approach to filter out false positives and reproduce true bugs in the original program. By transforming the program as well as mutating the input, T-Fuzz covers more code and finds more true bugs than any existing technique. We have evaluated T-Fuzz on the DARPA Cyber Grand Challenge dataset, LAVA-M dataset and 4 real-world programs (pngfix, tiffinfo, magick and pdftohtml). For the CGC dataset, T-Fuzz finds bugs in 166 binaries, Driller in 121, and AFL in 105. In addition, found 3 new bugs in previously-fuzzed programs and libraries.

FairFuzz: A Targeted Mutation Strategy for Increasing Greybox Fuzz Testing Coverage (ASE 2018)

Abstract: In recent years, fuzz testing has proven itself to be one of the most effective techniques for finding correctness bugs and security vulnerabilities in practice. One particular fuzz testing tool, American Fuzzy Lop (AFL), has become popular thanks to its ease-of-use and bug-finding power. However, AFL remains limited in the bugs it can find since it simply does not cover large regions of code. If it does not cover parts of the code, it will not find bugs there. We propose a two-pronged approach to increase the coverage achieved by AFL. First, the approach automatically identifies branches exercised by few AFL-produced inputs (rare branches), which often guard code that is empirically hard to cover by naively mutating inputs. The second part of the approach is a novel mutation mask creation algorithm, which allows mutations to be biased towards producing inputs hitting a given rare branch. This mask is dynamically computed during fuzz testing and can be adapted to other testing targets. We implement this approach on top of AFL in a tool named FairFuzz. We conduct evaluation on real-world programs against state-of-the-art versions of AFL. We find that on these programs FairFuzz achieves high branch coverage at a faster rate that state-of-the-art versions of AFL. In addition, on programs with nested conditional structure, it achieves sustained increases in branch coverage after 24 hours (average 10.6% increase). In qualitative analysis, we find that FairFuzz has an increased capacity to automatically discover keywords.

VUzzer: Application-aware Evolutionary Fuzzing (NDSS 2017)

Abstract: Fuzzing is an effective software testing technique to find bugs. Given the size and complexity of real-world applications, modern fuzzers tend to be either scalable, but not effective in exploring bugs that lie deeper in the execution, or capable of penetrating deeper in the application, but not scalable. In this paper, we present an application-aware evolutionary fuzzing strategy that does not require any prior knowledge of the application or input format. In order to maximize coverage and explore deeper paths, we leverage control- and data-flow features based on static and dynamic analysis to infer fundamental properties of the application. This enables much faster generation of interesting inputs compared to an application-agnostic approach. We implement our fuzzing strategy in VUzzer and evaluate it on three different datasets: DARPA Grand Challenge binaries (CGC), a set of real-world applications (binary input parsers), and the recently released LAVA dataset. On all of these datasets, VUzzer yields significantly better results than state-of-the-art fuzzers, by quickly finding several existing and new bugs.

Grammars \ Semantic \ Context-aware Fuzzing

FUZZILLI: Fuzzing for JavaScript JIT Compiler Vulnerabilities (NDSS 2023)

Paper

Abstract: JavaScript has become an essential part of the Internet infrastructure, and today's interactive web applications would be inconceivable without this programming language. On the downside, this interactivity implies that web applications rely on an ever-increasing amount of computationally intensive JavaScript code, which burdens the JavaScript engine responsible for efficiently executing the code. To meet these rising performance demands, modern JavaScript engines ship with sophisticated just-in-time (JIT) compilers. However, JIT compilers are a complex technology and, consequently, provide a broad attack surface for potential faults that might even be security-critical. Previous work on discovering software faults in JavaScript engines found many vulnerabilities, often using fuzz testing. Unfortunately, these fuzzing approaches are not designed to generate source code that actually triggers JIT semantics. Consequently, JIT vulnerabilities are unlikely to be discovered by existing methods. In this paper, we close this gap and present the first fuzzer that focuses on JIT vulnerabilities. More specifically, we present the design and implementation of an intermediate representation (IR) that focuses on discovering JIT compiler vulnerabilities. We implemented a complete prototype of the proposed approach and evaluated our fuzzer over a period of six months. In total, we discovered 17 confirmed security vulnerabilities. Our results show that targeted JIT fuzzing is possible and a dangerously neglected gap in fuzzing coverage for JavaScript engines.

FRAMESHIFTER: Manipulating HTTP/2 Frame Sequences with Fuzzing (Usenix Security2020)

Paper

Abstract: HTTP/2 adoption is rapidly climbing. However, in practice, Internet communications still rarely happen over end-to-end HTTP/2 channels. This is due to Content Delivery Networks and other reverse proxies, ubiquitous and necessary components of the Internet ecosystem, which only support HTTP/2 on the client's end, but not the forward connection to the origin server. Instead, proxy technologies predominantly rely on HTTP/2-to-HTTP/1 protocol conversion between the two legs of the connection.

We present the first systematic exploration of HTTP/2-to-HTTP/1 protocol conversion anomalies and their security implications. We develop a novel grammar-based fuzzer for HTTP/2, experiment with 12 popular reverse proxy technologies & CDNs through HTTP/2 frame sequence and content manipulation, and discover a plethora of novel web application attack vectors that lead to Request Blackholing, Denial-of-Service, Query-of-Death, and Request Smuggling attacks.

SGXFuzz: Efficiently Synthesizing Nested Structures for SGX Enclave Fuzzing (Usenix Security2022)

Paper

Abstract: Intel's Software Guard Extensions (SGX) provide a nonintrospectable trusted execution environment (TEE) to protect security-critical code from a potentially malicious OS. This protection can only be effective if the individual enclaves are secure, which is already challenging in regular software, and this becomes even more difficult for enclaves as the entire environment is potentially malicious. As such, many enclaves expose common vulnerabilities, e.g., memory corruption and SGXspecific vulnerabilities like null-pointer dereferences. While fuzzing is a popular technique to assess the security of software, dynamically analyzing enclaves is challenging as enclaves are meant to be non-introspectable. Further, they expect an allocated multi-pointer structure as input instead of a plain buffer.

In this paper, we present SGXFUZZ, a coverage-guided fuzzer that introduces a novel binary input structure synthesis method to expose enclave vulnerabilities even without source-code access. To obtain code coverage feedback from enclaves, we show how to extract enclave code from distribution formats. We also present an enclave runner that allows execution of the extracted enclave code as a user-space application at native speed, while emulating all relevant environment interactions of the enclave. We use this setup to fuzz enclaves using a state-of-the-art snapshot fuzzing engine that deploys our novel structure synthesis stage. This stage synthesizes multi-layer pointer structures and size fields incrementally on-the-fly based on fault signals. Furthermore, it matches the expected input format of the enclave without any prior knowledge. We evaluate our approach on 30 open- and closed-source enclaves and found a total of 79 new bugs and vulnerabilities.

Unicorn: Detect Runtime Error in Time-Series Databases With Hybrid Input Synthesis (ISSTA 2022)

Paper

Abstract: The ubiquitous use of time-series databases in the safety-critical Internet of Things domain demands strict security and correctness. One successful approach in database bug detection is fuzzing, where hundreds of bugs have been detected automatically in relational databases. However, it cannot be easily applied to time-series databases: the bulk of time-series logic is unreachable because of mismatched query specifications, and serious bugs are undetectable because of implicitly handled exceptions.

In this paper, we propose Unicorn to secure time-series databases with automated fuzzing. First, we design hybrid input synthesis to generate high-quality queries which not only cover time-series features but also ensure grammar correctness. Then, Unicorn uses proactive exception detection to discover minuscule-symptom bugs which hide behind implicit exception handling. With the specialized design oriented to time-series databases, Unicorn outperforms the state-of-the-art database fuzzers in terms of coverage and bugs. Specifically, Unicorn outperforms SQLsmith and SQLancer on widely used time-series databases IoTDB, KairosDB, TimescaleDB, TDEngine, QuestDB, and GridDB in the number of basic blocks by 21%-199% and 34%-693%, respectively. More importantly, Unicorn has discovered 42 previously unknown bugs.

Cooper: Testing the Binding Code of Scripting Languages with Cooperative Mutation (NDSS 2022)

Paper
Code

Abstract: Scripting languages like JavaScript are being integrated into commercial software to support easy file modification. For example, Adobe Acrobat accepts JavaScript to dynamically manipulate PDF files. To bridge the gap between the high-level scripts and the low-level languages (like C/C++) used to implement the software, a binding layer is necessary to transfer data and transform representations. However, due to the complexity of two sides, the binding code is prone to inconsistent semantics and security holes, which lead to severe vulnerabilities. Existing efforts for testing binding code merely focus on the script side, and thus miss bugs that require special program native inputs.

In this paper, we propose cooperative mutation, which modifies both the script code and the program native input to trigger bugs in binding code. Our insight is that many bugs are due to the interplay between the program initial state and the dynamic operations, which can only be triggered through two-dimensional mutations. We develop three novel techniques to enable practical cooperative mutation on popular scripting languages: we first cluster objects into semantics similar classes to reduce the mutation space of native inputs; then, we statistically infer the relationship between script code and object classes based on a large number of executions; at last, we use the inferred relationship to select proper objects and related script code for targeted mutation. We applied our tool, COOPER, on three popular systems that integrate scripting languages, including Adobe Acrobat, Foxit Reader and Microsoft Word. COOPER successfully found 134 previously unknown bugs. We have reported all of them to the developers. At the time of paper publishing, 59 bugs have been fixed and 33 of them are assigned CVE numbers. We are awarded totally 22K dollars bounty for 17 out of all reported bugs.

Fuzzing Class Specifications (ICSE 2022)

Paper

Abstract: Expressing class specifications via executable constraints is important for various software engineering tasks such as test generation, bug finding and automated debugging, but developers rarely write them. Techniques that infer specifications from code exist to fill this gap, but they are designed to support specific kinds of assertions and are difficult to adapt to support different assertion languages, e.g., to add support for quantification, or additional comparison operators, such as membership or containment.

To address the above issue, we present SpecFuzzer, a novel technique that combines grammar-based fuzzing, dynamic invariant detection, and mutation analysis, to automatically produce class specifications. SpecFuzzer uses: \emph{(i)} a fuzzer as a generator of candidate assertions derived from a grammar that is automatically obtained from the class definition; \emph{(ii)} a dynamic invariant detector - Daikon - to filter out assertions invalidated by a test suite; and \emph{(iii)} a mutation-based mechanism to cluster and rank assertions, so that similar constraints are grouped and then the stronger prioritized. Grammar-based fuzzing enables SpecFuzzer to be straightforwardly adapted to support different specification languages, by manipulating the fuzzing grammar, e.g., to include additional operators.

We evaluate our technique on a benchmark of 43 Java methods employed in the evaluation of the state-of-the-art techniques GAssert and EvoSpex. Our results show that SpecFuzzer can easily support a more expressive assertion language, over which is more effective than GAssert and EvoSpex in inferring specifications, according to standard performance metrics.

Efficient ECU Analysis Technology through Structure-aware CAN Fuzzing (Access 2022)

Paper

Abstract: Modern vehicles are equipped with a number of electronic control units (ECUs), which control vehicles efficiently by communicating with each other through the controller area network (CAN). However, the CAN is known to be vulnerable to cyber attacks because it does not have any security mechanisms. To find vulnerable CAN messages that can control safety-critical functions in ECUs, researchers have studied CAN fuzzing methods. In existing CAN fuzzing methods, fuzzing input values are generally generated at random without considering the structure of CAN messages, resulting in non-negligible CAN fuzzing time. In addition, existing fuzzing solutions have limited monitoring capabilities of the fuzzing results. In this paper, we propose a Structure-aware CAN Fuzzing protocol, in which the structure of CAN messages is considered and fuzzing input values are systematically generated to locate vulnerable functions in ECUs. Our proposed Structure-aware CAN Fuzzing system takes less time to run than existing solutions, meaning that problematic CAN messages that may have originated from SW implementation errors or CAN DBC (database CAN) design errors can be found quickly and, subsequently, appropriate action can be taken. Finally, we evaluated the performance of our Structure-aware CAN Fuzzing system on two real vehicles. We proved that our proposed method can find CAN messages that control safety-critical functions in ECUs faster than existing fuzzing solutions.

Semantic Image Fuzzing of AI Perception Systems

Paper

Abstract: Perception systems enable autonomous systems to interpret sensor readings obtained from the physical world. Testing of such systems aims to uncover misinterpretations that can severely impact an autonomous system's behavior. Current testing methods for perception systems, however, are inadequate: (1) when testing on real-world input data, the cost of human interpretation and annotation is very high, so test suites tend to be small; (2) the simulation-reality gap reduces the validity of test results based on simulated worlds; and (3) methods for synthesizing test cases with realistic inputs are limited in scope and lack semantic interpretations. To address these limitations, we developed a novel approach to fuzz testing perception systems based on semantic mutation of real-world sensor readings and their corresponding ground-truth interpretations. This enables mutations like adding a car driving on the street to an existing image while providing an oracle that accounts for that semantic change. We implemented our approach and evaluated its performance by generating 150,000 semantically mutated image inputs for five state-of-the-art perception systems. Our approach produced novel image inputs not found in the original test suite, and uncovered inputs that lead to significant issues in the analyzed systems at a very low cost.

SoFi: Reflection-Augmented Fuzzing for JavaScript Engines (CCS 2021)

Paper

Abstract: JavaScript engines have been shown prone to security vulnerabilities, which can lead to serious consequences due to their popularity. Fuzzing is an effective testing technique to discover vulnerabilities. The main challenge of fuzzing JavaScript engines is to generate syntactically and semantically valid inputs such that deep functionalities can be explored. However, due to the dynamic nature of JavaScript and the special features of different engines, it is quite challenging to generate semantically meaningful test inputs.

We observed that state-of-the-art semantic-aware JavaScript fuzzers usually require manually written rules to analyze the semantics for a JavaScript engine, which is labor-intensive, incomplete and engine-specific. Moreover, the error rate of generated test cases is still high. Another challenge is that existing fuzzers cannot generate new method calls that are not included in the initial seed corpus or pre-defined rules, which limits the bug-finding capability.

To this end, we propose a novel semantic-aware fuzzing technique named SoFi. To guarantee the validity of the generated test cases, SoFi adopts a fine-grained program analysis to identify available variables and infer types of these variables for the mutation. Moreover, an automatic repair strategy is proposed to repair syntax/semantic errors in invalid test cases. To improve the exploration capability of SoFi, we propose a reflection-based analysis to identify unseen attributes and methods of objects, which are further used in the mutation. With fine-grained analysis and reflection-based augmentation, SoFi can generate more valid and diverse test cases. Besides, SoFi is general in different JavaScript engines without any manual configuration (e.g., the grammar rules). The evaluation results have shown that SoFi outperforms state-of-the-art techniques in generating semantically valid inputs, improving code coverage and detecting more bugs. SoFi discovered 51 bugs in popular JavaScript engines, 28 of which have been confirmed or fixed by the developers and 10 CVE IDs have been assigned.

V-SHUTTLE: Scalable and Semantics-Aware Hypervisor Fuzzing (CCS 2021)

Paper

Abstract: With the wide application and deployment of cloud computing in enterprises, virtualization developers and security researchers are paying more attention to cloud computing security. The core component of cloud computing products is the hypervisor, which is also known as the virtual machine monitor (VMM) that can isolate multiple virtual machines in one host machine. However, compromising the hypervisor can lead to virtual machine escape and the elevation of privilege, allowing attackers to gain the permission of code execution in the host. Therefore, the security analysis and vulnerability detection of the hypervisor are critical for cloud computing enterprises. Importantly, virtual devices expose many interfaces to a guest user for communication, making virtual devices the most vulnerable part of a hypervisor. However, applying fuzzing to the virtual devices of a hypervisor is challenging because the data structures transferred by DMA are constructed in a nested form according to protocol specifications. Failure to understand the protocol of the virtual devices will make the fuzzing process stuck in the initial fuzzing stage, resulting in inefficient fuzzing.

In this paper, we propose a new framework called V-Shuttle to conduct hypervisor fuzzing, which performs scalable and semanticsaware hypervisor fuzzing. To address the above challenges, we first design a DMA redirection mechanism to significantly reduce the manual efforts to reconstruct virtual devices' protocol structures and make the fuzzing environment setup automated and scalable. Furthermore, we put forward a new fuzzing mutation scheduling mechanism called seedpool to make the virtual device fuzzing process semantics-aware and speed up the fuzzing process to achieve high coverage. Extensive evaluation on QEMU and VirtualBox, two of the most popular hypervisor platforms among the world, shows that V-Shuttle can efficiently reproduce existing vulnerabilities and find new vulnerabilities. We further carried out a long-term fuzzing campaign in QEMU/KVM and VirtualBox with V-Shuttle. In total, we discovered 35 new bugs with 17 CVEs assigne.

Token-Level Fuzzing (WiSec 2021)

Paper

Abstract: Fuzzing has become a commonly used approach to identifying bugs in complex, real-world programs. However, interpreters are notoriously difficult to fuzz effectively, as they expect highly structured inputs, which are rarely produced by most fuzzing mutations. For this class of programs, grammar-based fuzzing has been shown to be effective. Tools based on this approach can find bugs in the code that is executed after parsing the interpreter inputs, by following language-specific rules when generating and mutating test cases. Unfortunately, grammar-based fuzzing is often unable to discover subtle bugs associated with the parsing and handling of the language syntax. Additionally, if the grammar provided to the fuzzer is incomplete, or does not match the implementation completely, the fuzzer will fail to exercise important parts of the available functionality.

In this paper, we propose a new fuzzing technique, called Token-Level Fuzzing. Instead of applying mutations either at the byte level or at the grammar level, Token-Level Fuzzing applies mutations at the token level. Evolutionary fuzzers can leverage this technique to both generate inputs that are parsed successfully and generate inputs that do not conform strictly to the grammar. As a result, the proposed approach can find bugs that neither byte-level fuzzing nor grammar-based fuzzing can find. We evaluated Token-Level Fuzzing by modifying AFL and fuzzing four popular JavaScript engines, finding 29 previously unknown bugs, several of which could not be found with state-of-the-art byte-level and grammar-based fuzzers.

Extended grammar-based fuzzing algorithm for JavaScript Engines (2021)

Paper

Abstract: JavaScript engine security continues to be critical for user safety. Unfortunately, modern fuzzing algorithms cover only a small part of the entire engine. JavaScript engine requires highly structured input - JavaScript programs that are syntactically and semantically correct. The most of generated input struggle to pass syntax and semantic correctness checks. In this paper, we describe the extension of the grammar-based fuzzing algorithm. We propose a way of describing grammar for fuzzing using a set of JavaScript source codes. Grammars constructed with our method cover larger part of JavaScript language in comparison with grammars created by describing grammar rules. Another change of the basic algorithm is controlling the context in the mutation process. It allows filtering a lot of inputs that don't give new results. Our experiments show that the improved algorithm has increased speed of finding new paths in the target program.

Gramatron: Effective Grammar-Aware Fuzzing (ISSTA 2021)

Paper
Video

Abstract: Fuzzers aware of the input grammar can explore deeper program states using grammar-aware mutations. Existing grammar-aware fuzzers are ineffective at synthesizing complex bug triggers due to: (i) grammars introducing a sampling bias during input generation due to their structure, and (ii) the current mutation operators for parse trees performing localized small-scale changes. Gramatron uses grammar automatons in conjunction with aggressive mutation operators to synthesize complex bug triggers faster. We build grammar automatons to address the sampling bias. It restructures the grammar to allow for unbiased sampling from the input state space. We redesign grammar-aware mutation operators to be more aggressive, i.e., perform large-scale changes. Gramatron can consistently generate complex bug triggers in an efficient manner as compared to using conventional grammars with parse trees. Inputs generated from scratch by Gramatron have higher diversity as they achieve up to 24.2% more coverage relative to existing fuzzers. Gramatron makes input generation 98% faster and the input representations are 24% smaller. Our redesigned mutation operators are 6.4× more aggressive while still being 68% faster at performing these mutations. We evaluate Gramatron across three interpreters with 10 known bugs consisting of three complex bug triggers and seven simple bug triggers against two Nautilus variants. Gramatron finds all the complex bug triggers reliably and faster. For the simple bug triggers, Gramatron outperforms Nautilus four out of seven times. To demonstrate Gramatron's effectiveness in the wild, we deployed Gramatron on three popular interpreters for a 10-day fuzzing campaign where it discovered 10 new vulnerabilities.

One Engine to Fuzz 'em All: Generic Language Processor Testing with Semantic Validation (S&P 2021)

Abstract: Language processors, such as compilers and interpreters, are indispensable in building modern software. Errors in language processors can lead to severe consequences, like incorrect functionalities or even malicious attacks. However, it is not trivial to automatically test language processors to find bugs. Existing testing methods (or fuzzers) either fail to generate high-quality (i.e., semantically correct) test cases, or only support limited programming languages. In this paper, we propose POLYGLOT, a generic fuzzing framework that generates high-quality test cases for exploring processors of different programming languages. To achieve the generic applicability, POLYGLOT neutralizes the difference in syntax and semantics of programming languages with a uniform intermediate representation (IR). To improve the language validity, POLYGLOT performs constrained mutation and semantic validation to preserve syntactic correctness and fix semantic errors. We have applied POLYGLOT on 21 popular language processors of 9 programming languages, and identified 173 new bugs, 113 of which are fixed with 18 CVEs assigned. Our experiments show that POLYGLOT can support a wide range of programming languages, and outperforms existing fuzzers with up to 30x improvement in code coverage.

Growing A Test Corpus with Bonsai Fuzzing (ICSE 2021)

Paper

Abstract: This paper presents a coverage-guided grammar-based fuzzing technique for automatically generating a corpus of concise test inputs for programs such as compilers. We walk-through a case study of a compiler designed for education and the corresponding problem of generating meaningful test cases to provide to students. The prior state-of-the-art solution is a combination of fuzzing and test-case reduction techniques such as variants of delta-debugging. Our key insight is that instead of attempting to minimize convoluted fuzzer-generated test inputs, we can instead grow concise test inputs by construction using a form of iterative deepening. We call this approach Bonsai Fuzzing. Experimental results show that Bonsai Fuzzing can generate test corpora having inputs that are 16-45% smaller in size on average as compared to a fuzz-then-reduce approach, while achieving approximately the same code coverage and fault-detection capability.

Favocado: Fuzzing the Binding Code of JavaScript Engines Using Semantically Correct Test Cases (NDSS 2021)

Paper
Code

Abstract: JavaScript runtime systems include some specialized programming interfaces, called binding layers. Binding layers translate data representations between JavaScript and unsafe low-level languages, such as C and C++, by converting data between different types. Due to the wide adoption of JavaScript (and JavaScript engines) in the entire computing ecosystem, discovering bugs in JavaScript binding layers is critical. Nonetheless, existing JavaScript fuzzers cannot adequately fuzz binding layers due to two major challenges: Generating syntactically and semantically correct test cases and reducing the size of the input space for fuzzing.

In this paper, we propose Favocado, a novel fuzzing approach that focuses on fuzzing binding layers of JavaScript runtime systems. Favocado can generate syntactically and semantically correct JavaScript test cases through the use of extracted semantic information and careful maintaining of execution states. This way, test cases that Favocado generates do not raise unintended runtime exceptions, which substantially increases the chance of triggering binding code. Additionally, exploiting a unique feature (relative isolation) of binding layers, Favocado significantly reduces the size of the fuzzing input space by splitting DOM objects into equivalence classes and focusing fuzzing within each equivalence class. We demonstrate the effectiveness of Favocado in our experiments and show that Favocado outperforms a stateof-the-art DOM fuzzer. Finally, during the evaluation, we find 61 previously unknown bugs in four JavaScript runtime systems (Adobe Acrobat Reader, Foxit PDF Reader, Chromium, and WebKit). 33 of these bugs are security vulnerabilities.

CMFuzz: context-aware adaptive mutation for fuzzers (Empirical Software Engineering 2021)

Paper

Abstract: Mutation-based fuzzing is a simple yet effective technique to discover bugs and security vulnerabilities in software. Given a set of well-formed initial seeds, mutation-based fuzzers continually generate interesting seeds by applying specific mutation strategy in order to maximize code coverage or the number of unique bugs explored at any point-in-time. However, existing fuzzers remain limited in the paths it could cover since it simply follows a uniform distribution to choose mutation operators. In this paper, we proposed a novel context-aware adaptive mutation scheme, namely CMFuzz, which utilizes a contextual bandit algorithm LinUCB to effectively choose optimal mutation operators for various seed files. To this end, CMFuzz dynamically extracts and encodes file characteristics, which allows mutation-based fuzzers to perform context-aware mutation. We apply this scheme on top of several state-of-the-art fuzzers, i.e., PTfuzz, AFL, and AFLFast, and implement CMFuzz-PT, CMFuzz-AFL, and CMFuzz-AFLFast, respectively. We conduct evaluation on 12 real-world open source applications and LAVA-M dataset against their counterparts. Extensive evaluations demonstrate that CMFuzz-based fuzzers achieve higher code coverage and find more crashes at a faster rate than their counterparts on most cases. Furthermore, we also utilize other mainstream bandit algorithms, e.g., Thompson Sample and epsilon-greedy, and implement Thompson-PT and Greedy-PT based on PTfuzz to examine the performance of proposed model. CMFuzz-PT significantly outperforms Thompson-PT especially in terms of unique crashes and paths, i.e., found 1.79× unique crashes and 1.29× unique paths on average. Compared to Greedy-PT, our approach still increases the amount of unique crashes and paths by 1.11× and 1.05×, respectively.

Generating Highly-structured Input Data by Combining Search-based Testing and Grammar-based Fuzzing (ASE 2020)

Paper

Abstract: Software testing is an important and time-consuming task that is often done manually. In the last decades, researchers have come up with techniques to generate input data (e.g., fuzzing) and automate the process of generating test cases (e.g., search-based testing). However, these techniques are known to have their own limitations: search-based testing does not generate highly-structured data; grammar-based fuzzing does not generate test case structures. To address these limitations, we combine these two techniques. By applying grammar-based mutations to the input data gathered by the search-based testing algorithm, it allows us to co-evolve both aspects of test case generation. We evaluate our approach, called G-EvoSuite, by performing an empirical study on 20 Java classes from the three most popular JSON parsers across multiple search budgets. Our results show that the proposed approach on average improves branch coverage for JSON related classes by 15% (with a maximum increase of 50%) without negatively impacting other classes.

Montage: A Neural Network Language Model-Guided JavaScript Engine Fuzzer (Usenix Security2020)

Paper

Abstract: JavaScript (JS) engine vulnerabilities pose significant security threats affecting billions of web browsers. While fuzzing is a prevalent technique for finding such vulnerabilities, there have been few studies that leverage the recent advances in neural network language models (NNLMs). In this paper, we present Montage, the first NNLM-guided fuzzer for finding JS engine vulnerabilities. The key aspect of our technique is to transform a JS abstract syntax tree (AST) into a sequence of AST subtrees that can directly train prevailing NNLMs. We demonstrate that Montage is capable of generating valid JS tests, and show that it outperforms previous studies in terms of finding vulnerabilities. Montage found 37 real-world bugs, including three CVEs, in the latest JS engines, demonstrating its efficacy in finding JS engine bugs.

Fuzzing JavaScript Engines with Aspect-preserving Mutation (S&P 2020)

Abstract: Fuzzing is a practical, widely-deployed technique to find bugs in complex, real-world programs like JavaScript engines. We observed, however, that existing fuzzing approaches, either generative or mutational, fall short in fully harvesting high-quality input corpora such as known proof of concept (PoC) exploits or unit tests. Existing fuzzers tend to destruct subtle semantics or conditions encoded in the input corpus in order to generate new test cases because this approach helps in discovering new code paths of the program. Nevertheless, for JavaScript-like complex programs, such a conventional design leads to test cases that tackle only shallow parts of the complex codebase and fails to reach deep bugs effectively due to the huge input space.

In this paper, we advocate a new technique, called an aspect preserving mutation, that stochastically preserves the desirable properties, called aspects, that we prefer to be maintained across mutation. We demonstrate the aspect preservation with two mutation strategies, namely, structure and type preservation, in our fully-fledged JavaScript fuzzer, called DIE. Our evaluation shows that DIE's aspect-preserving mutation is more effective in discovering new bugs (5.7× more unique crashes) and producing valid test cases (2.4× fewer runtime errors) than the state-ofthe-art JavaScript fuzzers. DIE newly discovered 48 high-impact bugs in ChakraCore, JavaScriptCore, and V8 (38 fixed with 12 CVEs assigned as of today). The source code of DIE is publicly available as an open-source project.

Language-Agnostic Generation of Compilable Test Programs (ICST 2020)

Paper
Code

Abstract: Testing is an integral part of the development of compilers and other language processors. To automatically create large sets of test programs, random program generators, or fuzzers, have emerged. Unfortunately, existing approaches are either language-specific (and thus require a rewrite for each language) or may generate programs that violate rules of the respective programming language (which limits their usefulness). This work introduces *Smith, a language-agnostic framework for the generation of valid, compilable test programs. It takes as input an abstract attribute grammar that specifies the syntactic and semantic rules of a programming language. It then creates test programs that satisfy all these rules. By aggressively pruning the search space and keeping the construction as local as possible, *Smith can generate huge, complex test programs in short time. We present four case studies covering four real-world programming languages (C, Lua, SQL, and SMT-LIB 2) to show that *Smith is both efficient and effective, while being flexible enough to support programming languages that differ considerably. We found bugs in all four case studies. For example, *Smith detected 165 different crashes in older versions of GCC and LLVM. *Smith and the language grammars are available online.

Smart Greybox Fuzzing (TSE 2019)

Paper

Abstract: Coverage-based greybox fuzzing (CGF) is one of the most successful approaches for automated vulnerability detection. Given a seed file (as a sequence of bits), a CGF randomly flips, deletes or copies some bits to generate new files. CGF iteratively constructs (and fuzzes) a seed corpus by retaining those generated files which enhance coverage. However, random bitflips are unlikely to produce valid files (or valid chunks in files), for applications processing complex file formats. In this work, we introduce smart greybox fuzzing (SGF) which leverages a high-level structural representation of the seed file to generate new files. We define innovative mutation operators that work on the virtual file structure rather than on the bit level which allows SGF to explore completely new input domains while maintaining file validity. We introduce a novel validity-based power schedule that enables SGF to spend more time generating files that are more likely to pass the parsing stage of the program, which can expose vulnerabilities much deeper in the processing logic. Our evaluation demonstrates the effectiveness of SGF. On several libraries that parse complex chunk-based files, our tool AFLSMART achieves substantially more branch coverage (up to 87% improvement), and exposes more vulnerabilities than baseline AFL. Our tool AFLSMART has discovered 42 zero-day vulnerabilities in widely-used, well-tested tools and libraries; so far 17 CVEs were assigned.

Semantic Fuzzing with Zest (ISSTA 2019)

Abstract: Programs expecting structured inputs often consist of both a syntactic analysis stage, which parses raw input, and a semantic analysis stage, which conducts checks on the parsed input and executes the core logic of the program. Generator-based testing tools in the lineage of QuickCheck are a promising way to generate random syntactically valid test inputs for these programs. We present Zest, a technique which automatically guides QuickCheck-like randominput generators to better explore the semantic analysis stage of test programs. Zest converts random-input generators into deterministic parametric generators. We present the key insight that mutations in the untyped parameter domain map to structural mutations in the input domain. Zest leverages program feedback in the form of code coverage and input validity to perform feedback-directed parameter search. We evaluate Zest against AFL and QuickCheck on five Java programs: Maven, Ant, BCEL, Closure, and Rhino. Zest covers 1.03x-2.81x as many branches within the benchmarks semantic analysis stages as baseline techniques. Further, we find 10 new bugs in the semantic analysis stages of these benchmarks. Zest is the most effective technique in finding these bugs reliably and quickly, requiring at most 10 minutes on average to find each bug.

Field-aware Evolutionary Fuzzing Based on Input Specifications and Vulnerability Metrics (2019)

Paper

Abstract: Evolutionary fuzzing technology based on genetic algorithm has become one of the most effective vulnerability discovery techniques due to its fast and scalable advantages. How to effectively mutate the seed input plays a crucial role in improving the efficiency of the fuzzing. A good mutation strategy can increase code coverage and vulnerability triggering probability. Existing fuzzing tools generally focus on how to mutate smartly to improve code coverage to find more vulnerabilities (such as passing the branch with magic bytes), but they still face two challenges which substantially reduces the efficiency of vulnerability discovery. First, the input space is huge and current fuzzers are not aware of the input format, resulting in many mutated inputs are invalid. Second, they believe all bytes are equal and mutate them sequentially, wasting lots of time testing some uninteresting bytes. To this end, this paper proposes a field-aware mutation strategy that can find more vulnerabilities by generating fewer but more effective inputs. Specifically, we extract the field and type information of the seed input through the existing input specifications to ensure that the mutation is performed in field level instead of byte level and the optimal mutation strategy is selected. At the same time, the input fields are scored by code assessment based on vulnerability metrics, thus the more important fields (i.e., fields that are more likely to trigger the vulnerability) are prioritized to be mutated. We implemented a prototype tool, FaFuzzer, and evaluated it on two different datasets consisting of a variety of real-world applications. Experiments show that our field-aware strategy can find more vulnerabilities with fewer inputs than existing tools, while maintaining high code coverage. We found many unknown bugs in five widely used real-world applications and reported them to the relevant vendors.

Parser-Directed Fuzzing (PLDI 2019)

Abstract: To be effective, software test generation needs to well cover the space of possible inputs. Traditional fuzzing generates large numbers of random inputs, which however are unlikely to contain keywords and other specific inputs of non-trivial input languages. Constraint-based test generation solves conditions of paths leading to uncovered code, but fails on programs with complex input conditions because of path explosion. In this paper, we present a test generation technique specifically directed at input parsers. We systematically produce inputs for the parser and track comparisons made; after every rejection, we satisfy the comparisons leading to rejection. This approach effectively covers the input space: Evaluated on five subjects, from CSV files to JavaScript, our pFuzzer prototype covers more tokens than both random-based and constraint-based approaches, while requiring no symbolic analysis and far fewer tests than random fuzzers.

GRIMOIRE: Synthesizing Structure while Fuzzing (USENIX Security2019)

Paper

Abstract: In the past few years, fuzzing has received significant attention from the research community. However, most of this attention was directed towards programs without a dedicated parsing stage. In such cases, fuzzers which leverage the input structure of a program can achieve a significantly higher code coverage compared to traditional fuzzing approaches. This advancement in coverage is achieved by applying large-scale mutations in the application's input space. However, this improvement comes at the cost of requiring expert domain knowledge, as these fuzzers depend on structure input specifications (e. g., grammars). Grammar inference, a technique which can automatically generate such grammars for a given program, can be used to address this shortcoming. Such techniques usually infer a program's grammar in a pre-processing step and can miss important structures that are uncovered only later during normal fuzzing.

In this paper, we present the design and implementation of GRIMOIRE, a fully automated coverage-guided fuzzer which works without any form of human interaction or preconfiguration; yet, it is still able to efficiently test programs that expect highly structured inputs. We achieve this by performing large-scale mutations in the program input space using grammar-like combinations to synthesize new highly structured inputs without any pre-processing step. Our evaluation shows that GRIMOIRE outperforms other coverageguided fuzzers when fuzzing programs with highly structured inputs. Furthermore, it improves upon existing grammarbased coverage-guided fuzzers. Using GRIMOIRE, we identified 19 distinct memory corruption bugs in real-world programs and obtained 11 new CVEs.

Life after Speech Recognition: Fuzzing Semantic Misinterpretation for Voice Assistant Applications (NDSS 2019)

Paper

Abstract: Popular Voice Assistant (VA) services such as Amazon Alexa and Google Assistant are now rapidly appifying their platforms to allow more flexible and diverse voice-controlled service experience. However, the ubiquitous deployment of VA devices and the increasing number of third-party applications have raised security and privacy concerns. While previous works such as hidden voice attacks mostly examine the problems of VA services' default Automatic Speech Recognition (ASR) component, our work analyzes and evaluates the security of the succeeding component after ASR, i.e., Natural Language Understanding (NLU), which performs semantic interpretation (i.e., text-to-intent) after ASR's acoustic-to-text processing. In particular, we focus on NLU's Intent Classifier which is used in customizing machine understanding for third-party VA Applications (or vApps). We find that the semantic inconsistency caused by the improper semantic interpretation of an Intent Classifier can create the opportunity of breaching the integrity of vApp processing when attackers delicately leverage some common spoken errors. In this paper, we design the first linguistic-model-guided fuzzing tool, named LipFuzzer, to assess the security of Intent Classifier and systematically discover potential misinterpretation-prone spoken errors based on vApps' voice command templates. To guide the fuzzing, we construct adversarial linguistic models with the help of Statistical Relational Learning (SRL) and emerging Natural Language Processing (NLP) techniques. In evaluation, we have successfully verified the effectiveness and accuracy of LipFuzzer. We also use LipFuzzer to evaluate both Amazon Alexa and Google Assistant vApp platforms. We have identified that a large portion of real-world vApps are vulnerable based on our fuzzing result.

SLF: Fuzzing without Valid Seed Inputs (ICSE 2019)

Abstract: Fuzzing is an important technique to detect software bugs and vulnerabilities. It works by mutating a small set of seed inputs to generate a large number of new inputs. Fuzzers' performance often substantially degrades when valid seed inputs are not available. Although existing techniques such as symbolic execution can generate seed inputs from scratch, they have various limitations hindering their applications in real-world complex software without source code. In this paper, we propose a novel fuzzing technique that features the capability of generating valid seed inputs. It piggy-backs on AFL to identify input validity checks and the input fields that have impact on such checks. It further classifies these checks according to their relations to the input. Such classes include arithmetic relation, object offset, data structure length and so on. A multi-goal search algorithm is developed to apply class specific mutations in order to satisfy inter-dependent checks all together. We evaluate our technique on 20 popular benchmark programs collected from other fuzzing projects and the Google fuzzer test suite, and compare it with existing fuzzers AFL and AFLFast, symbolic execution engines KLEE and S2E, and a hybrid tool Driller that combines fuzzing with symbolic execution. The results show that our technique is highly effective and efficient, out-performing the other tools.

Superion: Grammar-Aware Greybox Fuzzing (ICSE 2019)

Abstract: In recent years, coverage-based greybox fuzzing has proven itself to be one of the most effective techniques for finding security bugs in practice. Particularly, American Fuzzy Lop (AFL for short) is deemed to be a great success in fuzzing relatively simple test inputs. Unfortunately, when it meets structured test inputs such as XML and JavaScript, those grammar-blind trimming and mutation strategies in AFL hinder the effectiveness and efficiency. To this end, we propose a grammar-aware coverage-based grey-box fuzzing approach to fuzz programs that process structured inputs. Given the grammar (which is often publicly available) of test inputs, we introduce a grammar-aware trimming strategy to trim test inputs at the tree level using the abstract syntax trees (ASTs) of parsed test inputs. Further, we introduce two grammar-aware mutation strategies (i.e., enhanced dictionary-based mutation and tree-based mutation). Specifically, tree-based mutation works via replacing subtrees using the ASTs of parsed test inputs. Equipped with grammar-awareness, our approach can carry the fuzzing exploration into width and depth. We implemented our approach as an extension to AFL, named Superion; and evaluated the effectiveness of Superion on real-life large-scale programs (a XML engine libplist and three JavaScript engines WebKit, Jerryscript and ChakraCore). Our results have demonstrated that Superion can improve the code coverage (i.e., 16.7% and 8.8% in line and function coverage) and bug-finding capability (i.e., 30 new bugs, among which we discovered 21 new vulnerabilities with 16 CVEs assigned and 3.2K USD bug bounty rewards received) over AFL and jsfunfuzz.

ProFuzzer: On-the-fly Input Type Probing for Better Zero-day Vulnerability Discovery (S&P 2019)

Abstract: Existing mutation based fuzzers tend to randomly mutate the input of a program without understanding its underlying syntax and semantics. In this paper, we propose a novel on-the-fly probing technique (called ProFuzzer) that automatically recovers and understands input fields of critical importance to vulnerability discovery during a fuzzing process and intelligently adapts the mutation strategy to enhance the chance of hitting zero-day targets. Since such probing is transparently piggybacked to the regular fuzzing, no prior knowledge of the input specification is needed. During fuzzing, individual bytes are first mutated and their fuzzing results are automatically analyzed to link those related together and identify the type for the field connecting them; these bytes are further mutated together following type-specific strategies, which substantially prunes the search space. We define the probe types generally across all applications, thereby making our technique application agnostic. Our experiments on standard benchmarks and real-world applications show that ProFuzzer substantially outperforms AFL and its optimized version AFLFast, as well as other state-of-art fuzzers including VUzzer, Driller and QSYM. Within two months, it exposed 42 zero-days in 10 intensively tested programs, generating 30 CVEs.

CodeAlchemist: Semantics-Aware Code Generation to Find Vulnerabilities in JavaScript Engines (NDSS 2019)

Abstract: JavaScript engines are an attractive target for attackers due to their popularity and flexibility in building exploits. Current state-of-the-art fuzzers for finding JavaScript engine vulnerabilities focus mainly on generating syntactically correct test cases based on either a predefined context-free grammar or a trained probabilistic language model. Unfortunately, syntactically correct JavaScript sentences are often semantically invalid at runtime. Furthermore, statically analyzing the semantics of JavaScript code is challenging due to its dynamic nature: JavaScript code is generated at runtime, and JavaScript expressions are dynamically-typed. To address this challenge, we propose a novel test case generation algorithm that we call semantics-aware assembly, and implement it in a fuzz testing tool termed CodeAlchemist. Our tool can generate arbitrary JavaScript code snippets that are both semantically and syntactically correct, and it effectively yields test cases that can crash JavaScript engines. We found numerous vulnerabilities of the latest JavaScript engines with CodeAlchemist and reported them to the vendors.

NAUTILUS: Fishing for Deep Bugs with Grammars (NDSS 2019)

Abstract: Fuzzing is a well-known method for efficiently identifying bugs in programs.Unfortunately, when fuzzing targets that require highly-structured inputs such as interpreters, many fuzzing methods struggle to pass the syntax checks. More specifically, interpreters often process inputs in multiple stages: first syntactic, then semantic correctness is checked. Only if these checks are passed, the interpreted code gets executed. This prevents fuzzers from executing ``deeper'' --- and hence potentially more interesting --- code. Typically two valid inputs that lead to the execution of different features in the target application require too many mutations for simple mutation-based fuzzers to discover: making small changes like bit flips usually only leads to the execution of error paths in the parsing engine. So-called grammar fuzzers are able to pass the syntax checks by using Context-Free Grammars. Using feedback can significantly increase the efficiency of fuzzing engines. Hence, it is commonly used in state-of-the-art mutational fuzzers that do not use grammars. Yet, grammar fuzzers do not make use of code coverage, i.e., they do not know whether any input triggers new functionality or not.

In this paper, we propose NAUTILUS, a method to efficiently fuzz programs that require highly-structured inputs by combining the use of grammars with the use of code coverage feedback. This allows us to recombine aspects of interesting inputs that were learned individually, and to dramatically increase the probability that any generated input will be accepted by the parser. We implemented a proof-of-concept fuzzer that we tested on multiple targets, including ChakraCore (the JavaScript engine of Microsoft Edge), PHP, mruby, and Lua. NAUTILUS identified multiple bugs in all of the targets: Seven in mruby, three in PHP, two in ChakraCore, and one in Lua. Reporting these bugs was awarded with a sum of 2600 USD and 6 CVEs were assigned. Our experiments show that combining context-free grammars and feedback-driven fuzzing significantly outperforms state-of-the-art approaches like American Fuzzy Lop (AFL) by an order of magnitude and grammar fuzzers by more than a factor of two when measuring code coverage.

TIFF: Using Input Type Inference To Improve Fuzzing (ACSAC 2018)

Abstract: Developers commonly use fuzzing techniques to hunt down all manner of memory corruption vulnerabilities during the testing phase. Irrespective of the fuzzer, input mutation plays a central role in providing adequate code coverage, as well as in triggering bugs. However, each class of memory corruption bugs requires a different trigger condition. While the goal of a fuzzer is to find bugs, most existing fuzzers merely approximate this goal by targeting their mutation strategies toward maximizing code coverage.

In this work, we present a new mutation strategy that maximizes the likelihood of triggering memory-corruption bugs by generating fewer, but better inputs. In particular, our strategy achieves bug- directed mutation by inferring the type of the input bytes. To do so, it tags each offset of the input with a basic type (e.g., 32-bit integer, string, array etc.), while deriving mutation rules for specific classes of bugs, We infer types by means of in-memory data-structure identification and dynamic taint analysis, and implement our novel mutation strategy in a fully functional fuzzer which we call TIFF (Type Inference-based Fuzzing Framework). Our evaluation on real-world applications shows that type-based fuzzing triggers bugs much earlier than existing solutions, while maintaining high code coverage. For example, on several real-world applications and libraries (e.g., poppler, mpg123 etc.), we find real bugs (with known CVEs) in almost half of the time and upto an order of magnitude fewer inputs than state-of-the-art fuzzers.

Skyfire: Data-Driven Seed Generation for Fuzzing (S&P 2017)

Paper

Abstract: Programs that take highly-structured files as inputs normally process inputs in stages: syntax parsing, semantic checking, and application execution. Deep bugs are often hidden in the application execution stage, and it is non-trivial to automatically generate test inputs to trigger them. Mutation-based fuzzing generates test inputs by modifying well-formed seed inputs randomly or heuristically. Most inputs are rejected at the early syntax parsing stage. Differently, generation-based fuzzing generates inputs from a specification (e.g., grammar). They can quickly carry the fuzzing beyond the syntax parsing stage. However, most inputs fail to pass the semantic checking (e.g., violating semantic rules), which restricts their capability of discovering deep bugs. In this paper, we propose a novel data-driven seed generation approach, named Skyfire, which leverages the knowledge in the vast amount of existing samples to generate well-distributed seed inputs for fuzzing programs that process highly-structured inputs. Skyfire takes as inputs a corpus and a grammar, and consists of two steps. The first step of Skyfire learns a probabilistic context-sensitive grammar (PCSG) to specify both syntax features and semantic rules, and then the second step leverages the learned PCSG to generate seed inputs. We fed the collected samples and the inputs generated by Skyfire as seeds of AFL to fuzz several open-source XSLT and XML engines (i.e., Sablotron, libxslt, and libxml2). The results have demonstrated that Skyfire can generate well-distributed inputs and thus significantly improve the code coverage (i.e., 20% for line coverage and 15% for function coverage on average) and the bug-finding capability of fuzzers. We also used the inputs generated by Skyfire to fuzz the closed-source JavaScript and rendering engine of Internet Explorer 11. Altogether, we discovered 19 new memory corruption bugs (among which there are 16 new vulnerabilities and received 33.5k USD bug bounty rewards) and 32 denial-of-service bugs.

Exploit Generation

ETHPLOIT: From Fuzzing to Efficient Exploit Generation against Smart Contracts (SANER2020)

Paper
Slides

Abstract: Smart contracts, programs running on blockchain systems, leverage diverse decentralized applications (DApps). Unfortunately, well-known smart contract platforms, Ethereum for example, face serious security problems. Exploits to contracts may cause enormous financial losses, which emphasize the importance of smart contract testing. However, current exploit generation tools have difficulty to solve hard constraints in execution paths and cannot simulate the blockchain behaviors very well. These problems cause a loss of coverage and accuracy of exploit generation.

To overcome the problems, we design and implement ETHPLOIT, a smart contract exploit generator based on fuzzing. ETHPLOIT adopts static taint analysis to generate exploit targeted transaction sequences, a dynamic seed strategy to pass hard constraints and an instrumented Ethereum Virtual Machine to simulate blockchain behaviors. We evaluate ETHPLOIT on 45,308 smart contracts and discovered 554 exploitable contracts. ETHPLOIT automatically generated 644 exploits without any false positive and 306 of them cannot be generated by previous exploit generation tools.

Gollum: Modular and Greybox Exploit Generation for Heap Overflows in Interpreters (CCS 2019)

Paper

Abstract: We present the first approach to automatic exploit generation for heap overflows in interpreters. It is also the first approach to exploit generation in any class of program that integrates a solution for automatic heap layout manipulation. At the core of the approach is a novel method for discovering exploit primitives---inputs to the target program that result in a sensitive operation, such as a function call or a memory write, utilizing attacker-injected data. To produce an exploit primitive from a heap overflow vulnerability, one has to discover a target data structure to corrupt, ensure an instance of that data structure is adjacent to the source of the overflow on the heap, and ensure that the post-overflow corrupted data is used in a manner desired by the attacker. Our system addresses all three tasks in an automatic, greybox, and modular manner. Our implementation is called GOLLUM, and we demonstrate its capabilities by producing exploits from 10 unique vulnerabilities in the PHP and Python interpreters, 5 of which do not have existing public exploits.

From proof-of-concept to exploitable (Cybersecurity 2019)

Abstract: Exploitability assessment of vulnerabilities is important for both defenders and attackers. The ultimate way to assess the exploitability is crafting a working exploit. However, it usually takes tremendous hours and significant manual efforts. To address this issue, automated techniques can be adopted. Existing solutions usually explore in depth the crashing paths, i.e., paths taken by proof-of-concept (PoC) inputs triggering vulnerabilities, and assess exploitability by finding exploitable states along the paths. However, exploitable states do not always exist in crashing paths. Moreover, existing solutions heavily rely on symbolic execution and are not scalable in path exploration and exploit generation.

In this paper, we propose a novel solution to generate exploit for userspace programs or facilitate the process of crafting a kernel UAF exploit. Technically, we utilize oriented fuzzing to explore diverging paths from vulnerability point. For userspace programs, we adopt a control-flow stitching solution to stitch crashing paths and diverging paths together to generate exploit. For kernel UAF, we leverage a lightweight symbolic execution to identify, analyze and evaluate the system calls valuable and useful for exploiting vulnerabilities.

We have developed a prototype system and evaluated it on a set of 19 CTF (capture the flag) programs and 15 realworld Linux kernel UAF vulnerabilities. Experiment results showed it could generate exploit for most of the userspace test set, and it could also facilitate security mitigation bypassing and exploitability evaluation for kernel test set.

Revery: From Proof-of-Concept to Exploitable (CCS 2018)

Abstract: Automatic exploit generation is an open challenge. Existing solutions usually explore in depth the crashing paths, i.e., paths taken by proof-of-concept (POC) inputs triggering vulnerabilities, and generate exploits when exploitable states are found along the paths. However, exploitable states do not always exist in crashing paths. Moreover, existing solutions heavily rely on symbolic execution and are not scalable in path exploration and exploit generation. In addition, few solutions could exploit heap-based vulnerabilities. In this paper, we propose a new solution revery to search for exploitable states in paths diverging from crashing paths, and generate control-flow hijacking exploits for heap-based vulnerabilities. It adopts three novel techniques:(1) a digraph to characterize a vulnerability's memory layout and its contributor instructions;(2) a fuzz solution to explore diverging paths, which have similar memory layouts as the crashing paths, in order to search more exploitable states and generate corresponding diverging inputs;(3) a stitch solution to stitch crashing paths and diverging paths together, and synthesize EXP inputs able to trigger both vulnerabilities and exploitable states. We have developed a prototype of revery based on the binary analysis engine angr, and evaluated it on a set of 19 real world CTF (capture the flag) challenges. Experiment results showed that it could generate exploits for 9 (47%) of them, and generate EXP inputs able to trigger exploitable states for another 5 (26%) of them.

SemFuzz: Semantics-based Automatic Generation of Proof-of-Concept Exploits (CCS 2017)

Paper

Abstract: Patches and related information about software vulnerabilities are often made available to the public, aiming to facilitate timely fixes. Unfortunately, the slow paces of system updates (30 days on average) often present to the attackers enough time to recover hidden bugs for attacking the unpatched systems. Making things worse is the potential to automatically generate exploits on input-validation flaws through reverse-engineering patches, even though such vulnerabilities are relatively rare (e.g., 5% among all Linux kernel vulnerabilities in last few years). Less understood, however, are the implications of other bug-related information (e.g., bug descriptions in CVE), particularly whether utilization of such information can facilitate exploit generation, even on other vulnerability types that have never been automatically attacked. In this paper, we seek to use such information to generate proof-of-concept (PoC) exploits for the vulnerability types never automatically attacked. Unlike an input validation flaw that is often patched by adding missing sanitization checks, fixing other vulnerability types is more complicated, usually involving replacement of the whole chunk of code. Without understanding of the code changed, automatic exploit becomes less likely. To address this challenge, we present SemFuzz, a novel technique leveraging vulnerability-related text (e.g., CVE reports and Linux git logs) to guide automatic generation of PoC exploits. Such an end-to-end approach is made possible by natural-language processing (NLP) based information extraction and a semantics-based fuzzing process guided by such information. Running over 112 Linux kernel flaws reported in the past five years, SemFuzz successfully triggered 18 of them, and further discovered one zero-day and one undisclosed vulnerabilities. These flaws include use-after-free, memory corruption, information leak, etc., indicating that more complicated flaws can also be automatically attacked. This finding calls into question the way vulnerability-related information is shared today.

ExploitMeter: Combining Fuzzing with Machine Learning for Automated Evaluation of Software Exploitability (PAC 2017)

Paper

Abstract: Exploitable software vulnerabilities pose severe threats to its information security and privacy. Although a great amount of efforts have been dedicated to improving software security, research on quantifying software exploitability is still in its infancy. In this work, we propose ExploitMeter, a fuzzing-based framework of quantifying software exploitability that facilitates decision-making for software assurance and cyber insurance. Designed to be dynamic, efficient and rigorous, ExploitMeter integrates machine learning-based prediction and dynamic fuzzing tests in a Bayesian manner. Using 100 Linux applications, we conduct extensive experiments to evaluate the performance of ExploitMeter in a dynamic environment.

Parallel / Ensemble Fuzzing

UltraFuzz: Towards Resource-saving in Distributed Fuzzing (TSE 2022)

Paper
Code

Abstract: Recent research has sought to improve fuzzing performance via parallel computing. However, researchers focus on improving efficiency while ignoring the increasing cost of testing resources. Parallel fuzzing in the distributed environment amplifies the resource-wasting problem caused by the random nature of fuzzing. In the parallel mode, owing to the lack of an appropriate task dispatching scheme and timely fuzzing status synchronization among different fuzzing instances, task conflicts and workload imbalance occur, making the resource-wasting problem severe. In this paper, we design UltraFuzz, a fuzzer for resource-saving in distributed fuzzing. Based on centralized dynamic scheduling, UltraFuzz can dispatch tasks and schedule power globally and reasonably to avoid resource-wasting. Besides, UltraFuzz can elastically allocate computing power for fuzzing and seed evaluation, thereby avoiding the potential bottleneck of seed evaluation that blocks the fuzzing process. UltraFuzz was evaluated using real-world programs, and the results show that with the same testing resource, UltraFuzz outperforms state-of-the-art tools, such as AFL, AFL-P, PAFL, and EnFuzz. Most importantly, the experiment reveals certain results that seem counter-intuitive, namely that parallel fuzzing can achieve "super-linear acceleration" when compared with single-core fuzzing. We conduct additional experiments to reveal the deep reasons behind this phenomenon and dig deep into the inherent advantages of parallel fuzzing over serial fuzzing, including the global optimization of seed energy scheduling and the escape of local optimal seed. Additionally, 24 real-world vulnerabilities were discovered using UltraFuzz.

Towards Systematic and Dynamic Task Allocation for Collaborative Parallel Fuzzing (ASE 2021 NIER)

Paper
Code

Abstract: Parallel coverage-guided greybox fuzzing is the most common setup for vulnerability discovery at scale. However, so far it has received little attention from the research community compared to single-mode fuzzing, leaving open several problems particularly in its task allocation strategies. Current approaches focus on managing micro tasks, at the seed input level, and their task division algorithms are either ad-hoc or static. In this paper, we leverage research on graph partitioning and search algorithms to propose a systematic and dynamic task allocation solution that works at the macro-task level. First, we design an attributed graph to capture both the program structures (e.g., program call graph) and fuzzing information (e.g., branch hit counts, bug discovery probability). Second, our graph partitioning algorithm divides the global program search space into sub-search-spaces. Finally our search algorithm prioritizes these sub-search-spaces (i.e., tasks) and explores them to maximize code coverage and number of bugs found. The results are collected to update the graph and guide further iterations of partitioning and exploration. We implemented a prototype tool called AFLTeam. In our preliminary experiments on well-tested benchmarks, AFLTeam achieved higher code coverage (up to 16.4% branch coverage improvement) compared to the default parallel mode of AFL and discovered 2 zero-day bugs in FFmpeg and JasPer toolkits.

CollabFuzz: A Framework for Collaborative Fuzzing (EuroSec 2021)

Paper

Abstract: In the recent past, there has been lots of work on improving fuzz testing. In prior work, EnFuzz showed that by sharing progress among different fuzzers, they can perform better than the sum of their parts. In this paper, we continue this line of work and present CollabFuzz, a collaborative fuzzing framework allowing multiple different fuzzers to collaborate under an informed scheduling policy based on a number of central analyses. More specifically, CollabFuzz is a generic framework that allows a user to express different test case scheduling policies, such as the collaborative approach presented by EnFuzz. CollabFuzz can control which tests cases are handed out to what fuzzer and allows the orchestration of different fuzzers across the network. Furthermore, it allows the centralized analysis of the test cases generated by the various fuzzers under its control, allowing to implement scheduling policies based on the results of arbitrary program (e.g., data-flow) analysis.

Improving Web Application Vulnerability Detection Leveraging Ensemble Fuzzing (ENASE 2021)

Paper

Abstract: The vast majority of online services we use nowadays provide their web application to the users. The correctness of the source code of these applications is crucial to prevent attackers from exploiting its vulnerabilities, leading to severe consequences like the disclosure of sensitive information or the degradation of the availability of the application. Currently, multiple existent solutions analyse and detect vulnerabilities in the source code. Attackers, however, do not usually have access to the source code and must work with the information that is made public. Their goals are clear - exploit vulnerabilities without accessing the code -, and they resort of black-box fuzzing tools to achieve such. In this paper, we propose an ensemble fuzzing approach to check the correctness of the web applications from the point of view of an attacker and, in a posterior phase, analyse the source code to correlate with the collected information. The approach focuses first on the quality of fuzzers' crawlers and afterwards on fuzzers capabilities of exploiting the results of all crawlers between them, in order to provide better coverage and precision in the detection of web vulnerabilities. Our preliminary results show that the ensemble performs better than fuzzers individually.

Cupid: Automatic Fuzzer Selection for Collaborative Fuzzing (ACSAC 2020)

Paper

Abstract: Combining the strengths of individual fuzzing methods is an appealing idea to find software faults more efficiently, especially when the computing budget is limited. In prior work, EnFuzz introduced the idea of ensemble fuzzing and devised three heuristics to classify properties of fuzzers in terms of diversity. Based on these heuristics, the authors manually picked a combination of different fuzzers that collaborate.

In this paper, we generalize this idea by collecting and applying empirical data from single, isolated fuzzer runs to automatically identify a set of fuzzers that complement each other when executed collaboratively. To this end, we present Cupid, a collaborative fuzzing framework allowing automated, data-driven selection of multiple complementary fuzzers for parallelized and distributed fuzzing. We evaluate the automatically selected target-independent combination of fuzzers by Cupid on Google's fuzzer-test-suite, a collection of real-world binaries, as well as on the synthetic Lava-M dataset. We find that Cupid outperforms two expert-guided, targetspecific and hand-picked combinations on Google's fuzzer-test-suite in terms of branch coverage, and improves bug finding on Lava-M by 10%. Most importantly, we improve the latency for obtaining 95% and 99% of the coverage by 90% and 64%, respectively. Furthermore, Cupid reduces the amount of CPU hours needed to find a high-performing combination of fuzzers by multiple orders of magnitude compared to an exhaustive evaluation.

EnFuzz: Ensemble Fuzzing with Seed Synchronization among Diverse Fuzzers (USENIX Security2019)

Abstract: Fuzzing is widely used for software vulnerability detection. There are various kinds of fuzzers with different fuzzing strategies, and most of them perform well on their targets. However, in industry practice and empirical study, the performance and generalization ability of those well-designed fuzzing strategies are challenged by the complexity and diversity of real-world applications. In this paper, inspired by the idea of ensemble learning, we first propose an ensemble fuzzing approach EnFuzz, that integrates multiple fuzzing strategies to obtain better performance and generalization ability than that of any constituent fuzzer alone. First, we define the diversity of the base fuzzers and choose those most recent and well-designed fuzzers as base fuzzers. Then, EnFuzz ensembles those base fuzzers with seed synchronization and result integration mechanisms. For evaluation, we implement EnFuzz , a prototype basing on four strong open-source fuzzers (AFL, AFLFast, AFLGo, FairFuzz), and test them on Google's fuzzing test suite, which consists of widely used real-world applications. The 24-hour experiment indicates that, with the same resources usage, these four base fuzzers perform variously on different applications, while EnFuzz shows better generalization ability and always outperforms others in terms of path coverage, branch coverage and crash discovery. Even compared with the best cases of AFL, AFLFast, AFLGo and FairFuzz, EnFuzz discovers 26.8%, 117%, 38.8% and 39.5% more unique crashes, executes 9.16%, 39.2%, 19.9% and 20.0% more paths and covers 5.96%, 12.0%, 21.4% and 11.1% more branches respectively.

PAFL: Extend FuzzingOptimizations of Single Mode to Industrial Parallel Mode (ESEC/FSE 2018)

Abstract: Researchers have proposed many optimizations to improve the efficiency of fuzzing, and most optimized strategies work very well on their targets when running in single mode with instantiating one fuzzer instance. However, in real industrial practice, most fuzzers run in parallel mode with instantiating multiple fuzzer instances, and those optimizations, unfortunately, fail to maintain the efficiency improvements.

In this paper, we present PAFL, a framework that utilizes efficient guiding information synchronization and task division to extend those existing fuzzing optimizations of single-mode to industrial parallel mode. With an additional data structure to store the guiding information, the synchronization ensures the information is shared and updated among different fuzzer instances timely. Then, the task division promotes the diversity of fuzzer instances by splitting the fuzzing task into several sub-tasks based on branch bitmap. We first evaluate PAFL using 12 different real-world programs from Google fuzzer-test-suite. Results show that in parallel mode, two AFL improvers - AFLFast and FairFuzz do not outperform AFL, which is different from the case in a single mode. However, when augmented with PAFL, the performance of AFLFast and FairFuzz in parallel mode improves. They cover 8% and 17% more branches, trigger 79% and 52% more unique crashes. For further evaluation of more widely-used software systems from GitHub, optimized fuzzers augmented with PAFL find more real bugs, and 25 of which are security-critical vulnerabilities registered as CVEs in the US National Vulnerability Database.

Sanitizer-guided Fuzzing

ParmeSan: Sanitizer-guided Greybox Fuzzing (USENIX Security2020)

Abstract: One of the key questions when fuzzing is where to look for vulnerabilities. Coverage-guided fuzzers indiscriminately optimize for covering as much code as possible given that bug coverage often correlates with code coverage. Since code coverage overapproximates bug coverage, this approach is less than ideal and may lead to non-trivial time-to-exposure (TTE) of bugs. Directed fuzzers try to address this problem by directing the fuzzer to a basic block with a potential vulnerability. This approach can greatly reduce the TTE for a specific bug, but such special-purpose fuzzers can then greatly underapproximate overall bug coverage.

In this paper, we present sanitizer-guided fuzzing, a new design point in this space that specifically optimizes for bug coverage. For this purpose, we make the key observation that while the instrumentation performed by existing software sanitizers are regularly used for detecting fuzzer-induced error conditions, they can further serve as a generic and effective mechanism to identify interesting basic blocks for guiding fuzzers. We present the design and implementation of ParmeSan, a new sanitizer-guided fuzzer that builds on this observation. We show that ParmeSan greatly reduces the TTE of real-world bugs, and finds bugs 37% faster than existing state-of-the-art coverage-based fuzzers (Angora) and 288% faster than directed fuzzers (AFLGo), while still covering the same set of bugs.

State / Sequence Guided Fuzzing

SWaTEval: An Evaluation Framework for Stateful Web Application Testing (ICISSP 2023)

Paper
Code

Abstract: Web applications are an easily accessible and valuable target for attackers. Therefore, web applications need to be examined for vulnerabilities. Modern web applications usually behave in a stateful manner and hence have an underlying state machine that determines their behavior based on the current state. To thoroughly test a web application, it is necessary to consider all aspects of a web application, including its internal states. In a blackbox setting, which we presuppose for this work, however, the internal state machine must be inferred before it can be used for testing. For state machine inference it is necessary to choose a similarity measure for web pages. Some approaches for automated blackbox stateful testing for web applications have already been proposed. It is, however, unclear how these approaches perform in comparison. We therefore present our evaluation framework for stateful web application testing, SWaTEval. In our evaluation, we show that SWaTEval is able to reproduce evaluation results from literature, demonstrating that SWaTEval is suitable for conducting meaningful evaluations. Further, we use SWaTEval to evaluate various approaches to similarity measures for web pages, including a new method based on the euclidean distance that we propose in this paper. These similarity measures are an important part of the automated state machine inference necessary for stateful blackbox testing. We show that the choice of similarity measure has an impact on the performance of the state machine inference regarding the number of correctly identified states, and that our newly proposed similarity measure leads to the highest number of correctly identified states.

Stateful Greybox Fuzzing (USENIX Security 2022)

Paper

Abstract: Many protocol implementations are reactive systems, where the protocol process is in continuous interaction with other processes and the environment. If a bug can be exposed only in a certain state, a fuzzer needs to provide a specific sequence of events as inputs that would take protocol into this state before the bug is manifested. We call these bugs as "stateful" bugs. Usually, when we are testing a protocol implementation, we do not have a detailed formal specification of the protocol to rely upon. Without knowledge of the protocol, it is inherently difficult for a fuzzer to discover such stateful bugs. A key challenge then is to cover the state space without an explicit specification of the protocol. In this work, we posit that manual annotations for state identification can be avoided for stateful protocol fuzzing. Specifically, we rely on a programmatic intuition that the state variables used in protocol implementations often appear in enum type variables whose values (the state names) come from named constants. In our analysis of the Top-50 most widely used open-source protocol implementations, we found that every implementation uses state variables that are assigned named constants (with easy to comprehend names such as INIT, READY) to represent the current state. In this work, we propose to automatically identify such state variables and track the sequence of values assigned to them during fuzzing to produce a "map" of the explored state space. Our experiments confirm that our stateful fuzzer discovers stateful bugs twice as fast as the baseline greybox fuzzer that we extended. Starting from the initial state, our fuzzer exercises one order of magnitude more state/transition sequences and covers code two times faster than the baseline fuzzer. Several zero-day bugs in prominent protocol implementations were found by our fuzzer, and 8 CVEs have been assigned.

Linear-time Temporal Logic guided Greybox Fuzzing (ICSE 2022)

Paper

Abstract: Software model checking is a verification technique which is widely used for checking temporal properties of software systems. Even though it is a property verification technique, its common usage in practice is in 'bug finding', that is, finding violations of temporal properties. Motivated by this observation and leveraging the recent progresses in fuzzing, we build a greybox fuzzing framework to find violations of Linear-time Temporal Logic (LTL) properties.

Our framework takes as input a sequential program written in C, and an LTL property. It finds violations, or counter-example traces, of the LTL property in stateful software systems; however, it does not achieve verification. Our work substantially extends directed greybox fuzzing to witness arbitrarily complex event orderings. We note that existing directed greybox fuzzing approaches are limited to witnessing reaching a location or witnessing simple event orderings like use-after-free. At the same time, compared to model checkers, our approach finds the counter-examples faster, thereby finding more counterexamples within a given time budget.

Our LTL-fuzzer tool, built on top of the AFL fuzzer, is shown to be effective in detecting bugs in well-known protocol implementations, such as OpenSSL and Telnet. We use LTL-fuzzer to reproduce known vulnerabilities (CVEs), to find 15 zero-day bugs by checking properties extracted from RFCs (for which 9 CVEs have been assigned), and to find violations of both safety as well as liveness properties in real-world protocol implementations. Our work represents a practical advance over software model checkers - while simultaneously also representing a conceptual advance over existing greybox fuzzers. Our work provides a starting point for understanding the unexplored synergies between software model checking and greybox fuzzing.

Rtkaller: State-aware Task Generation for RTOS Fuzzing (EMSOFT 2021)

Paper

Abstract: A real-time operating system (RTOS) is an operating system designed to meet certain real-time requirements. It is widely used in embedded applications, and its correctness is safety-critical. However, the validation of RTOS is challenging due to its complex real-time features and large code base.

In this paper, we propose Rtkaller, a state-aware kernel fuzzer for the vulnerability detection in RTOS. First, Rtkaller implements an automatic task initialization to transform the syscall sequences into initial tasks with more real-time information. Then, a coverage-guided task mutation is designed to generate those tasks that explore more in-depth real-time related code for parallel execution. Moreover, Rtkaller realizes a task modification to correct those tasks that may hang during fuzzing. We evaluated it on recent versions of rt-Linux, which is one of the most widely used RTOS. Compared to the state-of-the-art kernel fuzzers Syzkaller and Moonshine, Rtkaller achieves the same code coverage at the speed of 1.7X and 1.6X , gains an increase of 26.1% and 22.0% branch coverage within 24 hours respectively. More importantly, Rtkaller has confirmed 28 previously unknown vulnerabilities that are missed by other fuzzers.

Typestate-Guided Fuzzer for Discovering Use-after-Free Vulnerabilities (ICSE 2020)

Abstract: Existing coverage-based fuzzers usually use the individual control flow graph (CFG) edge coverage to guide the fuzzing process, which has shown great potential in finding vulnerabilities. However, CFG edge coverage is not effective in discovering vulnerabilities such as use-after-free (UaF). This is because, to trigger UaF vulnerabilities, one needs not only to cover individual edges, but also to traverse some long sequence of edges in a particular order, which is challenging for existing fuzzers. To this end, we first propose to model UaF vulnerabilities as typestate properties, then develop a typestate-guided fuzzer, named UAFL, for discovering vulnerabilities violating typestate properties. %Our approach works in two phases. Given a typestate property, we first perform a static typestate analysis to find operation sequences potentially violating the property. Then, the fuzzing process is guided by the operation sequences in order to progressively generate test cases triggering property violations. In addition, we also adopt the information flow analysis to improve the efficiency of the fuzzing process. We performed a thorough evaluation of UAFL on 14 widely-used real-world programs. The experiment results show that UAFL substantially outperforms the state-of-the-art fuzzers, including AFL, AFLFast, FairFuzz, MOpt, Angora and QSYM, in terms of the time taken to discover vulnerabilities. We discovered 10 previously unknown vulnerabilities, and received 5 new CVEs.

IJON: Exploring Deep State Spaces via Fuzzing (S&P 2020)

Abstract: Although current fuzz testing (fuzzing) methods are highly effective, there are still many situations such as complex state machines where fully automated approaches fail. State-of-the-art fuzzing methods offer very limited ability for a human to interact and aid the fuzzer in such cases. More specifically, most current approaches are limited to adding a dictionary or new seed inputs to guide the fuzzer. When dealing with complex programs, these mechanisms are unable to uncover new parts of the codebase.

In this paper, we propose IJON, an annotation mechanism that a human analyst can use to guide the fuzzer. In contrast to the two aforementioned techniques, this approach allows a more systematic exploration of the program's behavior based on the data representing the internal state of the program. As a consequence, using only a small (usually one line) annotation, a user can help the fuzzer to solve previously unsolvable challenges. We extended various AFL-based fuzzers with the ability to annotate the source code of the target application with guidance hints. Our evaluation demonstrates that such simple annotations are able to solve problems that - to the best of our knowledge - no other current fuzzer or symbolic execution based tool can overcome. For example, with our extension, a fuzzer is able to play and solve games such as Super Mario Bros. or resolve more complex patterns such as hash map lookups. To further demonstrate the capabilities of our annotations, we use AFL combined with IJON to uncover both novel security issues and issues that previously required a custom and comprehensive grammar to be uncovered. Lastly, we show that using IJON and AFL, one can solve many challenges from the CGC data set that resisted all fully automated and human-guided attempts so far.

MemFuzz: Using Memory Accesses to Guide Fuzzing (ICST 2019)

Abstract: Fuzzing is a form of random testing that is widely used for finding bugs and vulnerabilities. State of the art approaches commonly leverage information about the control flow of prior executions of the program under test to decide which inputs to mutate further. By relying solely on control flow information to characterize executions, such approaches may miss relevant differences. We propose augmenting evolutionary fuzzing by additionally leveraging information about memory accesses performed by the target program. The resulting approach can leverage more sophisticated information about the execution of the target program, enhancing the effectiveness of the evolutionary fuzzing. We implement our approach as a modification of the widely used AFL fuzzer and evaluate our implementation on three widely used target applications. We find distinct crashes from those detected by AFL for all three targets in our evaluation.

Rust Fuzzing

Rust Library Fuzzing via API Dependency Graph Traversal (ASE 2021)

Paper
Code

Abstract: Robustness is a key concern for Rust library development because Rust promises no risks of undefined behaviors if developers use safe APIs only. Fuzzing is a practical approach for examining the robustness of programs. However, existing fuzzing tools are not directly applicable to library APIs due to the absence of fuzz targets. It mainly relies on human efforts to design fuzz targets case by case which is labor-intensive. To address this problem, this paper proposes a novel automated fuzz target generation approach for fuzzing Rust libraries via API dependency graph traversal. We identify several essential requirements for library fuzzing, including validity and effectiveness of fuzz targets, high API coverage, and efficiency. To meet these requirements, we first employ breadth-first search with pruning to find API sequences under a length threshold, then we backward search longer sequences for uncovered APIs, and finally we optimize the sequence set as a set covering problem. We implement our fuzz target generator and conduct fuzzing experiments with AFL++ on several real-world popular Rust projects. Our tool finally generates 7 to 118 fuzz targets for each library with API coverage up to 0.92. We exercise each target with a threshold of 24 hours and find 30 previously-unknown bugs from seven libraries.

RUSTY: A Fuzzing Tool for Rust (ACSAC 2020)

Paper

Abstract: Rust is known as one of the most popular programming languages on the Stack Overflow website in 2020, indicating that many programmers have had the opportunity to use Rust in different projects. There are many reasons for this success, mostly due to its performance and safety. Rust is a friendly compiler with useful error messages, which provides excellent documentation with an integrated package manager. Furthermore, it is memory-efficient and fast without a garbage collector. Hence, the Rust compiler can power performance-critical services run on embedded devices. Likewise, due to Rusts rich type system and ownership model, it guarantees memory-safety and thread-safety enabling the end-users to reduce various bug types at compile-time.

However, even though all the advantages come with Rust, we found there are still various security issues in this ecosystem that can compromise the safety of Rust programs. To perform our security analysis, we created RUSTY that perform bug fuzzing based on the combination of concolic testing and property-based testing. is also the first kind of its own for the Rust ecosystem. Our preliminary evaluation collected multiple trendy Rust projects on GitHub, and RUSTY could successfully identify various memory security issues in this benchmark suite.

Regression Fuzzing

R2Z2: Detecting Rendering Regressions in Web Browsers through Differential Fuzz Testing (ICSE 2022)

Paper

Abstract: A rendering regression is a bug introduced by a web browser where a web page no longer functions as users expect. Such rendering bugs critically harm the usability of web browsers as well as web applications. The unique aspect of rendering bugs is that they affect the presented visual appearance of web pages, but those web pages have no pre-defined correct appearance. Therefore, it is challenging to automatically detect errors in their appearance. In practice, web browser vendors rely on non-trivial and time-prohibitive manual analysis to detect and handle rendering regressions.

This paper proposes R2Z2, an automated tool to find rendering regressions. R2Z2 uses the differential fuzz testing approach, which repeatedly compares the rendering results of two different versions of a browser while providing the same HTML as input. If the rendering results are different, R2Z2 further performs cross browser compatibility testing to check if the rendering difference is indeed a rendering regression. After identifying a rendering regression, R2Z2 will perform an in-depth analysis to aid in fixing the regression. Specifically, R2Z2 performs a delta-debugging-like analysis to pinpoint the exact browser source code commit causing the regression, as well as inspecting the rendering pipeline stages to pinpoint which pipeline stage is responsible. We implemented a prototype of R2Z2 particularly targeting the Chrome browser. So far, R2Z2 found 11 previously undiscovered rendering regressions in Chrome, all of which were confirmed by the Chrome developers. Importantly, in each case, R2Z2 correctly reported the culprit commit. Moreover, R2Z2 correctly pin-pointed the culprit rendering pipeline stage in all but one case.

FADATest: Fast and Adaptive Performance Regression Testing of Dynamic Binary Translation Systems (ICSE 2022)

Paper

Abstract: Dynamic binary translation (DBT) is the cornerstone of many important applications. In practice, however, it is quite difficult to maintain the performance efficiency of a DBT system due to its inherent complexity. Although performance regression testing isan effective approach to detect potential performance regression issues, it is not easy to apply performance regression testing to DBT systems, because of the natural differences between DBT systems and common software systems and the limited availability of effective test programs. In this paper, we present FADATest, which devises several novel techniques to address these challenges. Specifically,FADATest automatically generates adaptable test programs from existing real benchmark programs of DBT systems according to the runtime characteristics of the benchmarks. The test programs can then be used to achieve highly efficient and adaptive performance regression testing of DBT systems. We have implemented a prototype of FADATest. Experimental results on SPEC CPU 2017 and PARSEC show that FADATest can successfully uncover the same performance regression issues across the evaluated versions of two popular DBT systems, QEMU and Valgrind, as the original benchmark programs. Moreover, the testing efficiency is improved significantly on two different hardware platforms powered by x86-64 and AArch64, respectively.

Regression Greybox Fuzzing (CCS 2021)

Paper

Abstract: What you change is what you fuzz! In an empirical study of all fuzzer-generated bug reports in OSSFuzz, we found that four in every five bugs have been introduced by recent code changes. That is, 77% of 23k bugs are regressions. For a newly added project, there is usually an initial burst of new reports at 2-3 bugs per day. However, after that initial burst, and after weeding out most of the existing bugs, we still get a constant rate of 3-4 bug reports per week. The constant rate can only be explained by an increasing regression rate. Indeed, the probability that a reported bug is a regression (i.e., we could identify the bug-introducing commit) increases from 20% for the first bug to 92% after a few hundred bug reports.

In this paper, we introduce regression greybox fuzzing (RGF) a fuzzing approach that focuses on code that has changed more recently or more often. However, for any active software project, it is impractical to fuzz sufficiently each code commit individually. Instead, we propose to fuzz all commits simultaneously, but code present in more (recent) commits with higher priority. We observe that most code is never changed and relatively old. So, we identify means to strengthen the signal from executed code-of-interest. We also extend the concept of power schedules to the bytes of a seed and introduce Ant Colony Optimization to assign more energy to those bytes which promise to generate more interesting inputs.

Our large-scale fuzzing experiment demonstrates the validity of our main hypothesis and the efficiency of regression greybox fuzzing. We conducted our experiments in a reproducible manner within Fuzzbench, an extensible fuzzer evaluation platform. Our experiments involved 3+ CPU-years worth of fuzzing campaigns and 20 bugs in 15 open-source C programs available on OSSFuzz.

DeltaFuzz: Historical Version Information Guided Fuzz Testing (Journal of Computer Science and Technology 2021)

Video
Paper

Abstract: With the widespread use of agile software development methods, such as agile and scrum, software is iteratively updated more frequently. To ensure the quality of the software, regression testing is conducted before new versions are released. Moreover, to improve the efficiency of regression testing, testing efforts should be concentrated on the modified and impacted parts of a program. However, the costs of manually constructing new test cases for the modified and impacted parts are relatively expensive. Fuzz testing is an effective method for generating test data automatically, but it is usually devoted to achieving higher code coverage, which makes fuzz testing unsuitable for direct regression testing scenarios. For this reason, we propose a fuzz testing method based on the guidance of historical version information. First, the differences between the program being tested and the historical version are analyzed, and the results of the analysis are used to locate change points. Then, change impact analysis is performed to find the corresponding impacted basic blocks. Finally, the fitness values of test cases are calculated according to the execution traces, and new test cases are generated iteratively by the genetic algorithm. Based on the proposed method, we implement a prototype tool DeltaFuzz and conduct experiments on six open-source projects. Compared with the fuzzing tool AFLGo, AFLFast and AFL, DeltaFuzz could cover the target faster, and the time taken by DeltaFuzz was reduced by 20.59%, 30.05% and 32.61%, respectively.

HyDiff: Hybrid Differential Software Analysis (ICSE 2020)

Abstract: Detecting regression bugs in software evolution, analyzing side-channels in programs and evaluating robustness in deep neural networks (DNNs) can all be seen as instances of differential software analysis, where the goal is to generate diverging executions of program paths. Two executions are said to be diverging if the observable program behavior differs, e.g., in terms of program output, execution time, or (DNN) classification. The key challenge of differential software analysis is to simultaneously reason about multiple program paths, often across program variants.

This paper presents HyDiff, the first hybrid approach for differential software analysis. HyDiff integrates and extends two very successful testing techniques: Feedback-directed greybox fuzzing for efficient program testing and shadow symbolic execution for systematic program exploration. HyDiff extends greybox fuzzing with divergence-driven feedback based on novel cost metrics that take into account the control flow graph of the program. Furthermore HyDiff extends shadow symbolic execution by applying four-way forking in a systematic exploration and still having the ability to incorporate concrete inputs in the analysis. HyDiff applies divergence revealing heuristics based on resource consumption and control-flow information to efficiently guide the symbolic exploration, which allows its efficient usage beyond regression testing applications. We introduce differential metrics such as output, decision and cost difference, as well as patch distance, to assist the fuzzing and symbolic execution components in maximizing the execution divergence.

We implemented our approach on top of the fuzzer AFL and the symbolic execution framework Symbolic PathFinder. We illustrate HyDiff on regression and side-channel analysis for Java bytecode programs, and further show how to use HyDiff for robustness analysis of neural networks.

Directed Fuzzing

MC^2: Rigorous and Efficient Directed Greybox Fuzzing (CCS 2022)

Paper

Abstract: Directed greybox fuzzing is a popular technique for targeted software testing that seeks to find inputs that reach a set of target sites in a program. Most existing directed greybox fuzzers do not provide any theoretical analysis of their performance or optimality.

In this paper, we introduce a complexity-theoretic framework to pose directed greybox fuzzing as an oracle-guided search problem where some feedback about the input space (e.g., how close an input is to the target sites) is received by querying an oracle. Our framework assumes that each oracle query can return arbitrary content with a large but constant amount of information. Therefore, we use the number of oracle queries required by a fuzzing algorithm to find a target-reaching input as the performance metric. Using our framework, we design a randomized directed greybox fuzzing algorithm that makes a logarithmic (wrt. the number of all possible inputs) number of queries in expectation to find a target-reaching input. We further prove that the number of oracle queries required by our algorithm is optimal, i.e., no fuzzing algorithm can improve (i.e., minimize) the query count by more than a constant factor.

We implement our approach in MC^2 and outperform state-of-the-art directed greybox fuzzers on challenging benchmarks (Magma and Fuzzer Test Suite) by up to two orders of magnitude (i.e., 134\times) on average. MC^2 also found 15 previously undiscovered bugs that other state-of-the-art directed greybox fuzzers failed to find.

WindRanger: A Directed Greybox Fuzzer driven by Deviation Basic Block (ICSE 2022)

Paper

Abstract: Directed grey-box fuzzing (DGF) is a security testing technique that aims to steer the fuzzer towards predefined target sites in the program. To gain directness, DGF prioritizes the seeds whose execution traces are closer to the target sites. Therefore, evaluating the distance between the execution trace of a seed and the target sites (aka, the seed distance) is important for DGF. The first directed grey-box fuzzer, AFLGo, uses an approach of calculating the basic block level distances during static analysis and accumulating the distances of the executed basic blocks to compute the seed distance. Following AFLGo, most of the existing state-of-the-art DGF techniques use all the basic blocks on the execution trace and only the control flow information for seed distance calculation. However, not every basic block is equally important and there are certain basic blocks where the execution trace starts to deviate from the target sites (aka, deviation basic blocks).

In this paper, we propose a technique called WindRanger which leverages deviation basic blocks to facilitate DGF. To identify the deviation basic blocks, WindRanger applies both static reachability analysis and dynamic filtering. To conduct directed fuzzing, WindRanger uses the deviation basic blocks and their related data flow information for seed distance calculation, mutation, seed prioritization as well as explore-exploit scheduling. We evaluated WindRanger on 3 datasets consisting of 29 programs. The experiment results show that WindRanger outperforms AFLGo, AFL, and Fairfuzz by reaching the target sites 21%, 34%, and 37% faster and detecting the target crashes 44%, 66%, and 77% faster respectively. Moreover, we found a 0-day vulnerability with a CVE ID assigned in ffmpeg (a popular multimedia library extensively fuzzed by OSS-fuzz) with WindRanger by supplying manually identified suspect locations as the target sites.

BEACON: Directed Grey-Box Fuzzing with Provable Path Pruning (S&P 2022)

Paper
Code

Abstract: Unlike coverage-based fuzzing that gives equal attention to every part of a code, directed fuzzing aims to direct a fuzzer to a specific target in the code, e.g., the code with potential vulnerabilities. Despite much progress, we observe that existing directed fuzzers are still not efficient as they often symbolically or concretely execute a lot of program paths that cannot reach the target code. They thus waste a lot of computational resources. This paper presents BEACON, which can effectively direct a greybox fuzzer in the sea of paths in a provable manner. That is, assisted by a lightweight static analysis that computes abstracted preconditions for reaching the target, we can prune 82.94% of the executing paths at runtime with negligible analysis overhead (ă5h) but with the guarantee that the pruned paths must be spurious with respect to the target. We have implemented our approach, BEACON, and compared it to five state-of-the-art (directed) fuzzers in the application scenario of vulnerability reproduction. The evaluation results demonstrate that BEACON is 11.50x faster on average than existing directed grey-box fuzzers and it can also improve the speed of the conventional coverage-guided fuzzers, AFL, AFL++, and Mopt, to reproduce specific bugs with 6.31x ,11.86x, and 10.92x speedup, respectively. More interestingly, when used to test the vulnerability patches, BEACON found 14 incomplete fixes of existing CVE-identified vulnerabilities and 8 new bugs while 10 of them are exploitable with new CVE ids assigned

Improving Configurability of Unit-level Continuous Fuzzing: An Industrial Case Study with SAP HANA (ASE 2021 Industry)

Paper

Abstract: This paper presents industrial experiences on enhancing the configurability of a fuzzing framework for effective continuous fuzzing of the SAP HANA components. We propose five new mutation scheduling strategies for effective uses of grammar-aware mutators in the unit-level fuzzing framework, and three new seed corpus selection strategies to configure a fuzzing campaign to check on changed code in priority. The empirical results show that the proposed extension gives users chances to improve fuzzing effectiveness and efficiency by configuring the framework specifically for each target component

KCFuzz: Directed Fuzzing Based on Keypoint Coverage (ICAIS 2021)

Paper

Abstract: Directed fuzzing, as an efficient method to focus on a specific set of targets in the program, often works better than random fuzzing when combined with a researcher's empirical judgment. However, the current directed fuzzing work is not efficient enough. In previous studies, some have generated closer seed inputs by guiding the execution path through the distance from the target region, but the distance guided algorithm is less robust. Some studies used selective symbolic execution for directed testing to alleviate the path explosion problem, but it brings a higher false-positive rate. In this paper, we propose a keypoint coverage-based fuzzing (KCFuzz) method, which extracts the keypoint list using a control flow graph, obtains the keypoint list coverage information through runtime instrumentation, calculates the test priority of the seeds based on the overall coverage and keypoint coverage using an energy scheduling algorithm, and continuously generates test inputs closer to the target according to the specified mutation strategy. On this basis, a hybrid testing framework is implemented, using keypoint coverage directed fuzzing to generate a seed queue covering keypoints, using offspring generation strategies and hybrid execution technology, and further exploring the new state of the program according to changes in overall and keypoint coverage. The experimental results show that the KCFuzz method can efficiently induce the generation of seed queues to reach the target region, and at the same time, the depth and validity of the exploration paths are higher than those of the most advanced directed fuzzing methods such as AFLGo.

Constraint-guided Directed Greybox Fuzzing (USENIX Security2021)

Paper

Abstract: Directed greybox fuzzing is an augmented fuzzing technique intended for the targeted usages such as crash reproduction and proof-of-concept generation, which gives directedness to fuzzing by driving the seeds toward the designated program locations called target sites. However, we find that directed greybox fuzzing can still suffer from the long fuzzing time before exposing the targeted crash, because it does not consider the ordered target sites and the data conditions. This paper presents constraint-guided directed greybox fuzzing that aims to satisfy a sequence of constraints rather than merely reaching a set of target sites. Constraint-guided greybox fuzzing defines a constraint as the combination of a target site and the data conditions, and drives the seeds to satisfy the constraints in the specified order. We automatically generate the constraints with seven types of crash dumps and four types of patch changelogs, and evaluate the prototype system CAFL against the representative directed greybox fuzzing system AFLGo with 47 real-world crashes and 12 patch changelogs. The evaluation shows CAFL outperforms AFLGo by 2.88x for crash reproduction, and better performs in PoC generation as the constraints get explicit.

Constructing More Complete Control Flow Graphs Utilizing Directed Gray-Box Fuzzing (MDPI 2021)

Paper

Abstract: Control Flow Graphs (CFGs) provide fundamental data for many program analyses, such as malware analysis, vulnerability detection, code similarity analysis, etc. Existing techniques for constructing control flow graphs include static, dynamic, and hybrid analysis, which each having their own advantages and disadvantages. However, due to the difficulty of resolving indirect jump relations, the existing techniques are limited in completeness. In this paper, we propose a practical technique that applies static analysis and dynamic analysis to construct more complete control flow graphs. The main innovation of our approach is to adopt directed gray-box fuzzing (DGF) instead of coverage-based gray-box fuzzing (CGF) used in the existing approach to generate test cases that can exercise indirect jumps. We first employ a static analysis to construct the static CFGs without indirect jump relations. Then, we utilize directed gray-box fuzzing to generate test cases and resolve indirect jump relations by monitoring the execution traces of these test cases. Finally, we combine the static CFGs with indirect jump relations to construct more complete CFGs. In addition, we also propose an iterative feedback mechanism to further improve the completeness of CFGs. We have implemented our technique in a prototype and evaluated it through comparing with the existing approaches on eight benchmarks. The results show that our prototype can resolve more indirect jump relations and construct more complete CFGs than existing approaches.

Binary-level Directed Fuzzing for Use-After-Free Vulnerabilities (RAID 2020)

Abstract: Directed fuzzing focuses on automatically testing specific parts of the code by taking advantage of additional information such as (partial) bug stack trace, patches or risky operations. Key applications include bug reproduction, patch testing and static analysis report verification. Although directed fuzzing has received a lot of attention recently, hard-to-detect vulnerabilities such as Use-Afer-Free (UAF) are still not well addressed, more especially at the binary level. We propose UAFuzz, the first (binary-level) directed greybox fuzzer dedicated to UAF bugs. The technique features a fuzzing engine tailored to UAF specifics, a lightweight code instrumentation and an efficient bug triage step. Experimental evaluation for bug reproduction on real cases demonstrates that UAFuzz significantly outperforms state-of-the-art directed fuzzers in terms of fault detection rate, time to exposure and bug triaging. UAFuzz has also been proven effective in patch testing, leading to the discovery of 20 new bugs in Perl, GPAC and GNU Patch (including a buggy patch) - all of them have been acknowledged and 14 have been fixed. Last but not least, we provide to the community the first fuzzing benchmark dedicated to UAF, built on both real codes and real bugs.

Ankou: Guiding Grey-box Fuzzing towards Combinatorial Difference (ICSE 2020)

Paper
Code

Abstract: Grey-box fuzzing is an evolutionary process, which maintains and evolves a population of test cases with the help of a fitness function. Fitness functions used by current grey-box fuzzers are not informative in that they cannot distinguish different program executions as long as those executions achieve the same coverage. The problem is that the current fitness functions only consider a union of data, but not the combination of them. As such, fuzzers often get stuck in a local optimum during their search. In this paper, we introduce Ankou, the first grey-box fuzzer that recognizes different \emph{combinations} of execution information, and present several scalability challenges encountered while designing and implementing Ankou. Our experimental results show that Ankou is $1.94\times$ and $8.0\times$ more effective in finding bugs than AFL and Angora, respectively.

RDFuzz: Accelerating Directed Fuzzing with Intertwined Schedule and Optimized Mutation (2020)

Paper

Abstract: Directed fuzzing is a practical technique, which concentrates its testing energy on the process toward the target code areas, while costing little on other unconcerned components. It is a promising way to make better use of available resources, especially in testing large-scale programs. However, by observing the state-of-the-art-directed fuzzing engine (AFLGo), we argue that there are two universal limitations, the balance problem between the exploration and the exploitation and the blindness in mutation toward the target code areas. In this paper, we present a new prototype RDFuzz to address these two limitations. In RDFuzz, we first introduce the frequency-guided strategy in the exploration and improve its accuracy by adopting the branch-level instead of the path-level frequency. Then, we introduce the input-distance-based evaluation strategy in the exploitation stage and present an optimized mutation to distinguish and protect the distance sensitive input content. Moreover, an intertwined testing schedule is leveraged to perform the exploration and exploitation in turn. We test RDFuzz on 7 benchmarks, and the experimental results demonstrate that RDFuzz is skilled at driving the program toward the target code areas, and it is not easily stuck by the balance problem of the exploration and the exploitation.

TOFU: Target-Oriented FUzzer (Arxiv 2020)

Abstract: Program fuzzing - providing randomly constructed inputs to a computer program - has proved to be a powerful way to uncover bugs, find security vulnerabilities, and generate test inputs that increase code coverage. In many applications, however, one is interested in a target-oriented approach - one wants to find an input that causes the program to reach a specific target point in the program. We have created TOFU (for Target-Oriented FUzzer) to address the directed fuzzing problem. TOFU's search is biased according to a distance metric that scores each input according to how close the input's execution trace gets to the target locations. TOFU is also input-structure aware (i.e., the search makes use of a specification of a superset of the program's allowed inputs). Our experiments on xmllint show that TOFU is 28% faster than AFLGo, while reaching 45% more targets. Moreover, both distance-guided search and exploitation of knowledge of the input structure contribute significantly to TOFU's performance.

Sequence coverage directed greybox fuzzing (ICPC 2019)

Paper

Abstract: Existing directed fuzzers are not efficient enough. Directed symbolic-execution-based whitebox fuzzers, e.g. BugRedux, spend lots of time on heavyweight program analysis and constraints solving at runtime. Directed greybox fuzzers, such as AFLGo, perform well at runtime, but considerable calculation during instrumentation phase hinders the overall performance.

In this paper, we propose Sequence-coverage Directed Fuzzing (SCDF), a lightweight directed fuzzing technique which explores towards the user-specified program statements efficiently. Given a set of target statement sequences of a program, SCDF aims to generate inputs that can reach the statements in each sequence in order and trigger bugs in the program. Moreover, we present a novel energy schedule algorithm, which adjusts on demand a seed's energy according to its ability of covering the given statement sequences calculated on demand. We implement the technique in a tool LOLLY in order to achieve efficiency both at instrumentation time and at runtime. Experiments on several real-world software projects demonstrate that LOLLY outperforms two well-established tools on efficiency and effectiveness, i.e., AFLGo-a directed greybox fuzzer and BugRedux-a directed symbolic-execution-based whitebox fuzzer.

Hawkeye: Towards a Desired Directed Grey-box Fuzzer (CCS 2018)

Abstract: Grey-box fuzzing is a practically effective approach to test real-world programs. However, most existing grey-box fuzzers lack directedness, i.e. the capability of executing towards user-specified target sites in the program. To emphasize existing challenges in directed fuzzing, we propose Hawkeye to feature four desired properties of directed grey-box fuzzers. Owing to a novel static analysis on the program under test and the target sites, Hawkeye precisely collects the information such as the call graph, function and basic block level distances to the targets. During fuzzing, Hawkeye evaluates exercised seeds based on both static information and the execution traces to generate the dynamic metrics, which are then used for seed prioritization, power scheduling and adaptive mutating. These strategies help Hawkeye to achieve better directedness and gravitate towards the target sites. We implemented Hawkeye as a fuzzing framework and evaluated it on various real-world programs under different scenarios. The experimental results showed that Hawkeye can reach the target sites and reproduce the crashes much faster than state-of-the-art grey-box fuzzers such as AFL and AFLGo. Specially, Hawkeye can reduce the time to exposure for certain vulnerabilities from about 3.5 hours to 0.5 hour. By now, Hawkeye has detected more than 41 previously unknown crashes in projects such as Oniguruma, MJS with the target sites provided by vulnerability prediction tools; all these crashes are confirmed and 15 of them have been assigned CVE IDs.

RFUZZ: Coverage-Directed Fuzz Testing of RTL on FPGAs (ICCAD 2018)

Paper

Abstract: Dynamic verification is widely used to increase confidence in the correctness of RTL circuits during the pre-silicon design phase. Despite numerous attempts over the last decades to automate the stimuli generation based on coverage feedback, Coverage Directed Test Generation (CDG) has not found the widespread adoption that one would expect. Based on new ideas from the software testing community around coverage-guided mutational fuzz testing, we propose a new approach to the CDG problem which requires minimal setup and takes advantage of FPGA-accelerated simulation for rapid testing. We provide test input and coverage definitions that allow fuzz testing to be applied to RTL circuit verification. In addition we propose and implement a series of transformation passes that make it feasible to reset arbitrary RTL designs quickly, a requirement for deterministic test execution. Alongside this paper we provide rfuzz, a fully featured implementation of our testing methodology which we make available as open-source software to the research community. An empirical evaluation of rfuzz shows promising results on archiving coverage for a wide range of different RTL designs ranging from communication IPs to an industry scale 64-bit CPU.

Directed Greybox Fuzzing (CCS 2017)

Abstract: Existing Greybox Fuzzers (GF) cannot be effectively directed, for instance, towards problematic changes or patches, towards critical system calls or dangerous locations, or towards functions in the stack-trace of a reported vulnerability that we wish to reproduce. In this paper, we introduce Directed Greybox Fuzzing (DGF) which generates inputs with the objective of reaching a given set of target program locations efficiently. We develop and evaluate a simulated annealing-based power schedule that gradually assigns more energy to seeds that are closer to the target locations while reducing energy for seeds that are further away. Experiments with our implementation AFLGo demonstrate that DGF outperforms both directed symbolic-execution-based whitebox fuzzing and undirected greybox fuzzing. We show applications of DGF to patch testing and crash reproduction, and discuss the integration of AFLGo into Google's continuous fuzzing platform OSS-Fuzz. Due to its directedness, AFLGo could find 39 bugs in several well-fuzzed, security-critical projects like LibXML2. 17 CVEs were assigned.

Addressing Collision:

CollAFL: Path Sensitive Fuzzing (S&P 2018)

Abstract: Coverage-guided fuzzing is a widely used and ef- fective solution to find software vulnerabilities. Tracking code coverage and utilizing it to guide fuzzing are crucial to coverage- guided fuzzers. However, tracking full and accurate path coverage is infeasible in practice due to the high instrumentation overhead. Popular fuzzers (e.g., AFL) often use coarse coverage information, e.g., edge hit counts stored in a compact bitmap, to achieve highly efficient greybox testing. Such inaccuracy and incompleteness in coverage introduce serious limitations to fuzzers. First, it causes path collisions, which prevent fuzzers from discovering potential paths that lead to new crashes. More importantly, it prevents fuzzers from making wise decisions on fuzzing strategies. In this paper, we propose a coverage sensitive fuzzing solution CollAFL. It mitigates path collisions by providing more accurate coverage information, while still preserving low instrumentation overhead. It also utilizes the coverage information to apply three new fuzzing strategies, promoting the speed of discovering new paths and vulnerabilities. We implemented a prototype of CollAFL based on the popular fuzzer AFL and evaluated it on 24 popular applications. The results showed that path collisions are common, i.e., up to 75% of edges could collide with others in some applications, and CollAFL could reduce the edge collision ratio to nearly zero. Moreover, armed with the three fuzzing strategies, CollAFL outperforms AFL in terms of both code coverage and vulnerability discovery. On average, CollAFL covered 20% more program paths, found 320% more unique crashes and 260% more bugs than AFL in 200 hours. In total, CollAFL found 157 new security bugs with 95 new CVEs assigned.

Performance Fuzzing

Understanding and Detecting Performance Bugs in Markdown Compilers (ASE 2021)

Paper

Abstract: Markdown compilers are widely used for translating plain Markdown text into formatted text, yet they suffer from performance bugs that cause performance degradation and resource exhaustion. Currently, there is little knowledge and understanding about these performance bugs in the wild. In this work, we first conduct a comprehensive study of known performance bugs in Markdown compilers. We identify that the ways Markdown compilers handle the language's context-sensitive features are the dominant root cause of performance bugs. To detect unknown performance bugs, we develop MDPERFFUZZ, a fuzzing framework with a syntax-tree based mutation strategy to efficiently generate test cases to manifest such bugs. It equips an execution trace similarity algorithm to de-duplicate the bug reports. With MDPERFFUZZ, we successfully identified 216 new performance bugs in real-world Markdown compilers and applications. Our work demonstrates that the performance bugs are a common, severe, yet previously overlooked security problem.

HotFuzz: Discovering Algorithmic Denial-of-Service Vulnerabilities Through Guided Micro-Fuzzing (NDSS 2020)

Paper

Abstract: Contemporary fuzz testing techniques focus on identifying memory corruption vulnerabilities that allow adversaries to achieve either remote code execution or information disclosure. Meanwhile, Algorithmic Complexity (AC)vulnerabilities, which are a common attack vector for denial-of-service attacks, remain an understudied threat. In this paper, we present HotFuzz, a framework for automatically discovering AC vulnerabilities in Java libraries. HotFuzz uses micro-fuzzing, a genetic algorithm that evolves arbitrary Java objects in order to trigger the worst-case performance for a method under test. We define Small Recursive Instantiation (SRI) as a technique to derive seed inputs represented as Java objects to micro-fuzzing. After micro-fuzzing, HotFuzz synthesizes test cases that triggered AC vulnerabilities into Java programs and monitors their execution in order to reproduce vulnerabilities outside the fuzzing framework. HotFuzz outputs those programs that exhibit high CPU utilization as witnesses for AC vulnerabilities in a Java library. We evaluate HotFuzz over the Java Runtime Environment (JRE), the 100 most popular Java libraries on Maven, and challenges contained in the DARPA Space and Time Analysis for Cybersecurity (STAC) program. We evaluate SRI's effectiveness by comparing the performance of micro-fuzzing with SRI, measured by the number of AC vulnerabilities detected, to simply using empty values as seed inputs. In this evaluation, we verified known AC vulnerabilities, discovered previously unknown AC vulnerabilities that we responsibly reported to vendors, and received confirmation from both IBM and Oracle. Our results demonstrate that micro-fuzzing finds AC vulnerabilities in real-world software, and that micro-fuzzing with SRI-derived seed inputs outperforms using empty values.

MemLock: Memory Usage Guided Fuzzing (ICSE2020)

Abstract: Uncontrolled memory consumption is a kind of critical software security weaknesses. It can also become a security-critical vulnerability when attackers can control the input to consume a large amount of memory to launch a Denial-of-Service attack. However, detecting such vulnerability is challenging due to the fact that it requires long executions with well-crafted inputs to trigger excessive memory consumption, while the state-of-the-art testing techniques have mostly focused on code coverage. To tackle this challenge, we propose a feedback-directed fuzzing technique, named MemLock, to automatically generate those memory-consuming inputs to trigger memory consumption bugs. The fuzzing process is guided with memory consumption information so that the approach is general and does not require any domain knowledge. We perform a thorough evaluation for MemLock on 14 widely-used real-world programs. Our experiment results show that MemLock substantially outperforms the state-of-the-art fuzzing techniques, including AFL, AFLfast, PerfFuzz, FairFuzz and QSYM, in discovering memory consumption bugs. During the experiments, we discovered several previously unknown excessive memory consumption vulnerabilities and received 15 new CVEs.

Singularity: Pattern Fuzzing for Worst Case Complexity (FSE 2018)

Paper

Abstract: We describe a new blackbox complexity testing technique for determining the worst-case asymptotic complexity of a given application. The key idea is to look for an input pattern - rather than a concrete input - that maximizes the asymptotic resource usage of the program. Because input patterns can be described concisely as programs in a restricted language, our method transforms the complexity testing problem to optimal program synthesis. In particular, we express these input patterns using a new model of computation called Recurrent Computation Graph (RCG) and solve the optimal synthesis problem by developing a genetic programming algorithm that operates on RCGs. We have implemented the proposed ideas in a tool called Singularity and evaluate it on a diverse set of benchmarks. Our evaluation shows that Singularity can effectively discover the worst-case complexity of various algorithms and that it is more scalable compared to existing state-of-the-art techniques. Furthermore, our experiments also corroborate that Singularity can discover previously unknown performance bugs and availability vulnerabilities in real-world applications such as Google Guava and JGraphT.

PerfFuzz: Automatically Generating Pathological Inputs (ISSTA 2018)

Paper
Code

Abstract: Performance problems in software can arise unexpectedly when programs are provided with inputs that exhibit worst-case behavior. A large body of work has focused on diagnosing such problems via statistical profiling techniques. But how does one find these inputs in the first place? We present PerfFuzz, a method to automatically generate inputs that exercise pathological behavior across program locations, without any domain knowledge. PerfFuzz generates inputs via feedback-directed mutational fuzzing. Unlike previous approaches that attempt to maximize only a scalar characteristic such as the total execution path length, PerfFuzz uses multi-dimensional feedback and independently maximizes execution counts for all program locations. This enables PerfFuzz to (1) find a variety of inputs that exercise distinct hot spots in a program and (2) generate inputs with higher total execution path length than previous approaches by escaping local maxima. PerfFuzz is also effective at generating inputs that demonstrate algorithmic complexity vulnerabilities. We implement PerfFuzz on top of AFL, a popular coverage-guided fuzzing tool, and evaluate PerfFuzz on four real-world C programs typically used in the fuzzing literature. We find that PerfFuzz outperforms prior work by generating inputs that exercise the most-hit program branch 5x to 69x times more, and result in 1.9x to 24.7x longer total execution paths.

Badger: Complexity Analysis with Fuzzing and Symbolic Execution (ISSTA 2018)

Paper
Slides

Abstract: Hybrid testing approaches that involve fuzz testing and symbolic execution have shown promising results in achieving high code coverage, uncovering subtle errors and vulnerabilities in a variety of software applications. In this paper we describe Badger - a new hybrid approach for complexity analysis, with the goal of discovering vulnerabilities which occur when the worst-case time or space complexity of an application is significantly higher than the average case. Badger uses fuzz testing to generate a diverse set of inputs that aim to increase not only coverage but also a resource-related cost associated with each path. Since fuzzing may fail to execute deep program paths due to its limited knowledge about the conditions that influence these paths, we complement the analysis with a symbolic execution, which is also customized to search for paths that increase the resource-related cost. Symbolic execution is particularly good at generating inputs that satisfy various program conditions but by itself suffers from path explosion. Therefore, Badger uses fuzzing and symbolic execution in tandem, to leverage their benefits and overcome their weaknesses. We implemented our approach for the analysis of Java programs, based on Kelinci and Symbolic PathFinder. We evaluated Badger on Java applications, showing that our approach is significantly faster in generating worst-case executions compared to fuzzing or symbolic execution on their own.

SlowFuzz: Automated Domain-Independent Detection of Algorithmic Complexity Vulnerabilities (CCS 2017)

Abstract: Algorithmic complexity vulnerabilities occur when the worst-case time/space complexity of an application is significantly higher than the respective average case for particular user-controlled inputs. When such conditions are met, an attacker can launch Denial-of-Service attacks against a vulnerable application by providing inputs that trigger the worst-case behavior. Such attacks have been known to have serious effects on production systems, take down entire websites, or lead to bypasses of Web Application Firewalls. Unfortunately, existing detection mechanisms for algorithmic complexity vulnerabilities are domain-specific and often require significant manual effort. In this paper, we design, implement, and evaluate SlowFuzz, a domain-independent framework for automatically finding algorithmic complexity vulnerabilities. SlowFuzz automatically finds inputs that trigger worst-case algorithmic behavior in the tested binary. SlowFuzz uses resource-usage-guided evolutionary search techniques to automatically find inputs that maximize computational resource utilization for a given application.

Enhancing Memory Error:

OBSan: An Out-Of-Bound Sanitizer to Harden DNN Executables (NDSS 2023)

Paper

Abstract: The rapid adoption of deep neural network (DNN) models on a variety of hardware platforms has boosted the development of deep learning (DL) compilers. DL compilers take as input the high-level DNN model specifications and generate optimized DNN executables for diverse hardware architectures like CPUs and GPUs. Despite the emerging adoption of DL compilers in real-world scenarios, no solutions exist to protect DNN executables. To fill this critical gap, this paper introduces OBSAN, a fast sanitizer designed to check out-of-bound (OOB) behavior of DNN executables. From a holistic view, DNN incorporates bidirectional computation: forward propagation that predicts an output based on an input, and backward propagation that characterizes how the forward prediction is made. Both neuron activations in forward propagation and the gradients in backward propagation should fall within valid ranges, and deviations from the valid ranges would be considered as OOB.

OOB is primarily related to unsafe behavior of DNNs, which root from anomalous inputs and may cause mispredictions or even exploitation via adversarial examples (AEs). We thus design OBSAN, which includes two variants, FOBSAN and BOBSAN, that can detect OOB in the forward and backward propagations, respectively. Each OBSAN is designed as extra passes of DL compilers to integrate with large-scale DNN models, and we design various optimization schemes to reduce the overhead of OBSAN. Evaluations over various anomalous inputs show that OBSAN manifests promising OOB detectability with low overhead. We further present two downstream applications to show how OBSAN prevents online AE generation and facilitates feedback-driven fuzz testing toward DNN executables.

SANRAZOR: Reducing Redundant Sanitizer Checks in C/C++ Programs (OSDI 2021)

Abstract: Sanitizers detect unsafe actions such as invalid memory accesses by inserting checks that are validated during a program's execution. Despite their extensive use for debugging and vulnerability discovery, sanitizer checks often induce a high runtime cost. One important reason for the high cost is, as we observe in this paper, that many sanitizer checks are redundant - the same safety property is repeatedly checked - leading to unnecessarily wasted computing resources. To help more profitably utilize sanitizers, we introduce SanRazor, a practical tool aiming to effectively detect and remove redundant sanitizer checks. SanRazor adopts a novel hybrid approach - it captures both dynamic code coverage and static data dependencies of checks, and uses the extracted information to perform a redundant check analysis. Our evaluation on the SPEC benchmarks shows that SanRazor can reduce the overhead of sanitizers significantly, from 73.8% to 28.0-62.0% for AddressSanitizer, and from 160.1% to 36.6-124.4% for UndefinedBehaviorSanitizer (depending on the applied reduction scheme). Our further evaluation on 38 CVEs from 10 commonly-used programs shows that SanRazor reduced checks suffice to detect at least 33 out of the 38 CVEs. Furthermore, by combining SanRazor with an existing sanitizer reduction tool ASAP, we show synergistic effect by reducing the runtime cost to only 7.0% with a reasonable tradeoff of security.

Unleashing Fuzzing Through Comprehensive, Efficient, and Faithful Exploitable-Bug Exposing

Paper

Abstract: Fuzzing has become an essential means of finding software bugs. Bug finding through fuzzing requires two partsexploring code paths to reach bugs and exposing bugs when they are reached. Existing fuzzing research has primarily focused on improving code coverage but not on exposing bugs. Sanitizers such as AddressSanitizer (ASAN) and MemorySanitizer (MSAN) have been the dominating tools for exposing bugs. However, sanitizer-based bug exposing has the following limitations. (1) sanitizers are not compatible with each other. (2) sanitizers incur significant runtime overhead. (3) sanitizers may generate false positives, and (4) exposed bugs may not be exploitable. To address these limitations, we propose EXPOZZER, a fuzzing system that can expose bugs comprehensively, efficiently, and faithfully. The intuition of EXPOZZER is to detect bugs through divergences in a properly diversified dual-execution environment, which does not require maintaining or checking execution metadata. We design a practical and deterministic dual-execution engine, a co-design for dual-execution and fuzzers, bug-sensitive diversification, comprehensive and efficient divergence detection to ensure the effectiveness of EXPOZZER. The results of evaluations show that EXPOZZER can detect not only CVE-assigned vulnerabilities reliably, but also new vulnerabilities in well-tested real-world programs.EXPOZZER is 10 times faster than MemorySanitizer and is similar to AddressSanitizer.

HDR-Fuzz: Detecting Buffer Overruns using AddressSanitizer Instrumentation and Fuzzing (2021)

Paper

Abstract: Buffer-overruns are a prevalent vulnerability in software libraries and applications. Fuzz testing is one of the effective techniques to detect vulnerabilities in general. Greybox fuzzers such as AFL automatically generate a sequence of test inputs for a given program using a fitness-guided search process. A recently proposed approach in the literature introduced a buffer-overrun specific fitness metric called "headroom", which tracks how close each generated test input comes to exposing the vulnerabilities. That approach showed good initial promise, but is somewhat imprecise and expensive due to its reliance on conservative points-to analysis. Inspired by the approach above, in this paper we propose a new ground-up approach for detecting buffer-overrun vulnerabilities. This approach uses an extended version of ASAN (Address Sanitizer) that runs in parallel with the fuzzer, and reports back to the fuzzer test inputs that happen to come closer to exposing buffer-overrun vulnerabilities. The ASAN-style instrumentation is precise as it has no dependence on points-to analysis. We describe in this paper our approach, as well as an implementation and evaluation of the approach.

Enhancing Memory Error Detection for Large-Scale Applications and Fuzz Testing (NDSS 2018)

Abstract: Memory errors are one of the most common vulnerabilities for the popularity of memory unsafe languages including C and C++. Once exploited, it can easily lead to system crash (i.e., denial-of-service attacks) or allow adversaries to fully compromise the victim system. This paper proposes MEDS, a practical memory error detector. MEDS significantly enhances its detection capability by approximating two ideal properties, called an infinite gap and an infinite heap. The approximated infinite gap of MEDS setups large inaccessible memory region between objects (i.e., 4 MB), and the approximated infinite heap allows MEDS to fully utilize virtual address space (i.e., 45-bits memory space). The key idea of MEDS in achieving these properties is a novel user-space memory allocation mechanism, MEDSALLOC. MEDSALLOC leverages a page aliasing mechanism, which allows MEDS to maximize the virtual memory space utilization but minimize the physical memory uses. To highlight the detection capability and practical impacts of MEDS, we evaluated and then compared to Google's state-of-the-art detection tool, AddressSanitizer. MEDS showed three times better detection rates on four real-world vulnerabilities in Chrome and Firefox. More importantly, when used for a fuzz testing, MEDS was able to identify 68.3% more memory errors than AddressSanitizer for the same amount of a testing time, highlighting its practical aspects in the software testing area. In terms of performance overhead, MEDS slowed down 108% and 86% compared to native execution and AddressSanitizer, respectively, on real-world applications including Chrome, Firefox, Apache, Nginx, and OpenSSL.

AddressSanitizer: A Fast Address Sanity Checker (USENIX Security2012)

Memory access bugs, including buffer overflows and uses of freed heap memory, remain a serious problem for programming languages like C and C++. Many memory error detectors exist, but most of them are either slow or detect a limited set of bugs, or both. This paper presents AddressSanitizer, a new memory error detector. Our tool finds out-of-bounds accesses to heap, stack, and global objects, as well as use-after-free bugs. It employs a specialized memory allocator and code instrumentation that is simple enough to be implemented in any compiler, binary translation system, or even in hardware. AddressSanitizer achieves efficiency without sacrificing comprehensiveness. Its average slowdown is just 73% yet it accurately detects bugs at the point of occurrence. It has found over 300 previously unknown bugs in the Chromium browser and many bugs in other software.

Schedule (Power & Mutation)

SLIME: Program-sensitive Energy Allocation for Fuzzing (ISSTA 2022)

Abstract: The energy allocation strategy is one of the most popular techniques in fuzzing to improve code coverage and vulnerability discovery. The core intuition is that fuzzers should allocate more computational energy to the seed files that have high efficiency to trigger unique paths and crashes after mutation. Existing solutions usually define several properties, e.g., the execution speed, the file size, and the number of the triggered edges in the control flow graph, to serve as the key measurements in their allocation logics to estimate the potential of a seed. The efficiency of a property is usually assumed to be the same across different programs. However, we find that this assumption is not always valid. As a result, the state-of-the-art energy allocation solutions with static energy allocation logics are hard to achieve desirable performance on different programs.

To address the above problem, we propose a novel program-sensitive solution, named SLIME, to enable adaptive energy allocation on the seed files with various properties for each program. Specifically, SLIME first designs multiple property-aware queues, with each queue containing the seed files with a specific property. Second, to improve the return of investment, SLIME leverages a customized Upper Confidence Bound Variance-aware (UCB-V) algorithm to statistically select a property queue with the most estimated reward, i.e., finding the most new unique execution paths and crashes. Finally, SLIME mutates the seed files in the selected property queue to perform property-adaptive fuzzing on a program. We evaluate SLIME against state-of-the-art open source fuzzers AFL, MOPT, AFL++, AFL++HIER, EcoFuzz, and TortoiseFuzz on 9 real-world programs. The results demonstrate that SLIME discovers up to 3.53X more unique vulnerabilities than the baselines, and achieves better coverage performance. We will open source a prototype of SLIME to facilitate future fuzzing research.

MobFuzz: Adaptive Multi-objective Optimization in Gray-box Fuzzing (NDSS 2022)

Paper

Abstract: Coverage-guided gray-box fuzzing (CGF) is an efficient software testing technique. There are usually multiple objectives to optimize in CGF. However, existing CGF methods cannot successfully find the optimal values for multiple objectives simultaneously. In this paper, we propose a gray-box fuzzer for multi-objective optimization (MOO) called MobFuzz. We model the multi-objective optimization process as a multi-player multi-armed bandit (MPMAB). First, it adaptively selects the objective combination that contains the most appropriate objectives for the current situation. Second, our model deals with the power schedule, which adaptively allocates energy to the seeds under the chosen objective combination. In MobFuzz, we propose an evolutionary algorithm called NIC to optimize our chosen objectives simultaneously without incurring additional performance overhead. To prove the effectiveness of MobFuzz, we conduct experiments on 12 real-world programs and the MAGMA data set. Experiment results show that multi-objective optimization in MobFuzz outperforms single-objective fuzzing in the baseline fuzzers. In contrast to them, MobFuzz can select the optimal objective combination and increase the values of multiple objectives up to 107%, with at most a 55% reduction in the energy consumption. Moreover, MobFuzz has up to 6% more program coverage and finds 3x more unique bugs than the baseline fuzzers. The NIC algorithm has at least a 2x improvement with a performance overhead of approximately 3%.

Effective Seed Scheduling for Fuzzing with Graph Centrality Analysis (S&P 2022)

Paper
Code

Abstract: Seed scheduling, the order in which seeds are selected, can greatly affect the performance of a fuzzer. Existing approaches schedule seeds based on their historical mutation data, but ignore the structure of the underlying Control Flow Graph (CFG). Examining the CFG can help seed scheduling by revealing the potential edge coverage gain from mutating a seed.

An ideal strategy will schedule seeds based on a count of all reachable and feasible edges from a seed through mutations, but computing feasibility along all edges is prohibitively expensive. Therefore, a seed scheduling strategy must approximate this count. We observe that an approximate count should have 3 properties - (i) it should increase if there are more edges reachable from a seed; (ii) it should decrease if mutation history information suggests an edge is hard to reach or is located far away from currently visited edges; and (iii) it should be efficient to compute over large CFGs.

We observe that centrality measures from graph analysis naturally provide these three properties and therefore can efficiently approximate the likelihood of reaching unvisited edges by mutating a seed. We therefore build a graph called the edge horizon graph that connects seeds to their closest unvisited nodes and compute the seed node's centrality to measure the potential edge coverage gain from mutating a seed.

We implement our approach in K-scheduler and compare with many popular seed scheduling strategies. We find that K-scheduler increases feature coverage by 25.89% compared to Entropic and edge coverage by 4.21% compared to the next-best AFL-based seed scheduler, in arithmetic mean on 12 Google FuzzBench programs. It also finds 3 more previously-unknown bugs than the next-best AFL-based seed scheduler.

Path Transitions Tell More: Optimizing Fuzzing Schedules via Runtime Program States (ICSE 2022)

Paper

Abstract: Coverage-guided Greybox Fuzzing (CGF) is one of the most successful and widely-used techniques for bug hunting. Two major approaches are adopted to optimize CFG: (i) to reduce search space of inputs by inferring relationships between input bytes and path constraints; (ii) to formulate fuzzing processes (e.g., path transitions) and build up probability distributions to optimize power schedules, i.e., the number of inputs generated per seed. However, the former is subjective to the inference results which may include extra bytes for a path constraint, thereby limiting the efficiency of path constraints resolution, code coverage discovery, and bugs exposure; the latter formalization, concentrating on power schedules for seeds alone, is inattentive to the schedule for bytes in a seed.

In this paper, we propose a lightweight fuzzing framework, Truzz, to optimize existing CGFs. To address two aforementioned challenges, Truzz identifies the bytes related to the validation checks (i.e., the checks guarding error-handling code), and protects those bytes from being frequently mutated, making most generated inputs examine the functionalities of programs, in lieu of being rejected by validation checks. The byte-wise relationship determination mitigates the problem of loading extra bytes when fuzzers infer the byte-constraint relation. Furthermore, the proposed path transition within Truzz can efficiently prioritize the seed as the new path, harvesting many new edges, likely belongs to a code region with many undiscovered code lines. To evaluate our approach, we implemented 4 state-of-the-art fuzzers, AFL, AFLFast, NEUZZ, and Greyone, in Truzz. The experimental results show that Truzz can generate up to 33% more inputs flowing into functional code, in addition to roughly 83% more edge coverage identified than the vanilla fuzzers. Finally, our approach exposes 9 bugs in 8 target programs, and three of the bugs have not been identified by the vanilla fuzzers.

FEFuzzer: Hybrid Files Fuzzing Tool (2022)

Paper

Abstract: Fuzzing is one of the most effective techniques for finding bugs and mining potential vulnerabilities in applications. In the fuzzing test, how to select test cases directly affects the result of the testing. In this paper, we propose a new method that can compute the effectiveness of test cases based on the genetic algorithm. A fuzzer can filter out more effective test cases by using the new method and recycles them, which can improve the efficiency of the fuzzing test. Through several experiments, we summarize four parameters (CPU occupancy rate, memory occupancy rate, a crash of the test target and execution time of processing a test case) which can be used in our new method. Meanwhile, we develop a fuzzer named FEFuzzer and implement the new method in FEFuzzer which has many new features than other fuzzing tools. FEFuzzer can not only generate more than 130 different standard file samples but also can automatically select effective cases and recycle them in fuzzing tests. Finally, we do some fuzzing experiments by using FEFuzzer and zzuf. By analyzing the result of experiments, we find that FEFuzzer is more efficient and FEFuzzer can find out more bugs than zzuf.

Seed Selection for Successful Fuzzing (ISSTA 2021)

Paper

Abstract: Mutation-based greybox fuzzing - unquestionably the most widely-used fuzzing technique - relies on a set of non-crashing seed inputs (a corpus) to bootstrap the bug-finding process. When evaluating a fuzzer, common approaches for constructing this corpus include: (i) using an empty file; (ii) using a single seed representative of the target's input format; or (iii) collecting a large number of seeds (e.g., by crawling the Internet). Little thought is given to how this seed choice affects the fuzzing process, and there is no consensus on which approach is best (or even if a best approach exists).

To address this gap in knowledge, we systematically investigate and evaluate how seed selection affects a fuzzer's ability to find bugs in real-world software. This includes a systematic review of seed selection practices used in both evaluation and deployment contexts, and a large-scale empirical evaluation (over 33 CPU-years) of six seed selection approaches. These six seed selection approaches include three corpus minimization techniques (which select the smallest subset of seeds that trigger the same range of instrumentation data points as a full corpus).

Our results demonstrate that fuzzing outcomes vary significantly depending on the initial seeds used to bootstrap the fuzzer, with minimized corpora outperforming singleton, empty, and large (in the order of thousands of files) seed sets. Consequently, we encourage seed selection to be foremost in mind when evaluating/deploying fuzzers, and recommend that (a) seed choice be carefully considered and explicitly documented, and (b) never to evaluate fuzzers with only a single seed.

MooFuzz: Many-Objective Optimization Seed Schedule for Fuzzer (Mathematics 2021)

Paper

Abstract: Coverage-based Greybox Fuzzing (CGF) is a practical and effective solution for finding bugs and vulnerabilities in software. A key challenge of CGF is how to select conducive seeds and allocate accurate energy. To address this problem, we propose a novel many-objective optimization solution, MooFuzz, which can identify different states of the seed pool and continuously gather different information about seeds to guide seed schedule and energy allocation. First, MooFuzz conducts risk marking in dangerous positions of the source code. Second, it can automatically update the collected information, including the path risk, the path frequency, and the mutation information. Next, MooFuzz classifies seed pool into three states and adopts different objectives to select seeds. Finally, we design an energy recovery mechanism to monitor energy usage in the fuzzing process and reduce energy consumption. We implement our fuzzing framework and evaluate it on seven real-world programs. The experimental results show that MooFuzz outperforms other state-of-the-art fuzzers, including AFL, AFLFast, FairFuzz, and PerfFuzz, in terms of path discovery and bug detection.

EcoFuzz: Adaptive Energy-Saving Greybox Fuzzing as a Variant of the Adversarial Multi-Armed Bandit (USENIX Security2020)

Paper

Abstract: Fuzzing is one of the most effective approaches for identifying security vulnerabilities. As a state-of-the-art coverage-based greybox fuzzer, AFL is a highly effective and widely used technique. However, AFL allocates excessive energy (i.e.,the number of test cases generated by the seed) to seeds that exercise the high-frequency paths and can not adaptively adjust the energy allocation, thus wasting a significant amount of energy. Moreover, the current Markov model for modeling coverage-based greybox fuzzing is not profound enough. This paper presents a variant of the Adversarial Multi-Armed Bandit model for modeling AFL's power schedule process. We first explain the challenges in AFL's scheduling algorithm by using the reward probability that generates a test case for discovering a new path. Moreover, we illustrated the three states of the seeds set and developed a unique adaptive scheduling algorithm as well as a probability-based search strategy. These approaches are implemented on top of AFL in an adaptive energy-saving greybox fuzzer called EcoFuzz. EcoFuzz is examined against other six AFL-type tools on 14 real-world subjects over 490 CPU days. According to the results, EcoFuzz could attain 214% of the path coverage of AFL with reducing 32% test cases generation of that of AFL. Besides, EcoFuzz identified 12 vulnerabilities in GNU Binutils and other software. We also extended EcoFuzz to test some IoT devices and found a new vulnerability in the SNMP component.

MEUZZ: Smart Seed Scheduling for Hybrid Fuzzing (2020)

Paper

Abstract: Seed scheduling is a prominent factor in determining the yields of hybrid fuzzing. Existing hybrid fuzzers schedule seeds based on fixed heuristics that aim to predict input utilities. However, such heuristics are not generalizable as there exists no one-size-fits-all rule applicable to different programs. They may work well on the programs from which they were derived, but not others. To overcome this problem, we design a Machine learning-Enhanced hybrid fUZZing system (MEUZZ), which employs supervised machine learning for adaptive and generalizable seed scheduling. MEUZZ determines which new seeds are expected to produce better fuzzing yields based on the knowledge learned from past seed scheduling decisions made on the same or similar programs. MEUZZ's learning is based on a series of features extracted via code reachability and dynamic analysis, which incurs negligible runtime overhead (in microseconds). Moreover, MEUZZ automatically infers the data labels by evaluating the fuzzing performance of each selected seed. As a result, MEUZZ is generally applicable to, and performs well on, various kinds of programs. Our evaluation shows MEUZZ significantly outperforms the state-of-the-art grey-box and hybrid fuzzers, achieving 27.1% more code coverage than QSYM. The learned models are reusable and transferable, which boosts fuzzing performance by 7.1% on average and improves 68% of the 56 cross-program fuzzing campaigns. MEUZZ discovered 47 deeply hidden and previously unknown bugs--with 21 confirmed and fixed by the developers--when fuzzing 8 well-tested programs with the same configurations as used in previous work.

Greybox Fuzzing Based on Ant Colony Algorithm (AINA 2020)

Paper

Abstract: Greybox fuzzing technology is a kind of fuzzing technology that is commonly used now and effective. This fuzzing technology can guide the direction of fuzzing by acquiring the execution information of some paths in the program. However, the gray box fuzzy testing technology commonly used in the market today evaluates the seed of a sample by its path depth, execution time, and whether there is a new path to judge the quality of a sample, which is often not comprehensive. This article will propose a sample seed screening technology that uses ant colony algorithm to control gray box fuzzy test. By estimating the transition probability between the basic block and the basic block, we can determine what kind of seed sample is more likely to mutate into a new sample file. Based on this, the order and degree of fuzzing of the samples are determined, so as to improve the efficiency of fuzzing.

Suzzer: A Vulnerability-Guided Fuzzer Based on Deep Learning (Inscrypt 2019)

Paper

Abstract: Fuzzing is a simple and effective way to find software bugs. Most state-of-the-art fuzzers focus on improving code coverage to enhance the possibility of causing crashes. However, a software program oftentimes has only a fairly small portion that contains vulnerabilities, leading coverage-based fuzzers to work poorly most of the time. To address this challenge, we propose Suzzer, a vulnerability-guided fuzzer, to concentrate on testing code blocks that are more likely to contain bugs. Suzzer has a light-weight static analyzer to extract ACFG vector from target programs. In order to determine which code blocks are more vulnerable, Suzzer is equipped with prediction models which get the prior probability of each ACFG vector. The prediction models will guide Suzzer to generate test inputs with higher vulnerability scores, thus improving the efficiency of finding bugs. We evaluate Suzzer using two different datasets: artificial LAVA-M dataset and a set of real-world programs. The results demonstrate that in the best case of short-term fuzzing, Suzzer saved 64.5% of the time consumed to discover vulnerabilities compared to VUzzer.

MOPT: Optimize Mutation Scheduling for Fuzzers (USENIX Security2019)

Abstract: MOpt is a novel mutation scheduling scheme, which enables mutation-based fuzzers to discover vulnerabilities more efficiently. MOPT utilizes a customized Particle Swarm Optimization (PSO) algorithm to find the optimal selection probability distribution of operators with respect to fuzzing effectiveness, and provides a pacemaker fuzzing mode to accelerate the convergence speed of PSO. MOPT provides a good rationality, compatibility and steadiness, while introducing negligible costs. We make MOpt-AFL (one of the applications of MOpt-based fuzzers), seed sets used in the evaluation, results and the technical report with more details publicly available to facilitate the research in this area, which is available at: https://github.com/puppet-meteor/MOpt-AFL .

Cerebro: Context-aware Adaptive Fuzzing for Effective Vulnerability Detection (FSE 2019)

Paper

Abstract: Existing greybox fuzzers mainly utilize program coverage as the goal to guide the fuzzing process. To maximize their outputs, coverage-based greybox fuzzers need to evaluate the quality of seeds properly, which involves making two decisions: 1) which is the most promising seed to fuzz next (seed prioritization), and 2) how many efforts should be made to the current seed (power scheduling). In this paper, we present our fuzzer, Cerebro, to address the above challenges. For the seed prioritization problem, we propose an online multi-objective based algorithm to balance various metrics such as code complexity, coverage, execution time, etc. To address the power scheduling problem, we introduce the concept of input potential to measure the complexity of uncovered code and propose a cost-effective algorithm to update it dynamically. Unlike previous approaches where the fuzzer evaluates an input solely based on the execution traces that it has covered, Cerebro is able to foresee the benefits of fuzzing the input by adaptively evaluating its input potential. We perform a thorough evaluation for Cerebro on 6 different real-world programs. The experiments show that Cerebro can find more vulnerabilities and achieve better coverage than state-of-the-art fuzzers such as AFL and AFLFast. Additionally, we found 15 previously unknown bugs in mjs (a light-weight Javascript engine for embedded systems), Intel XED (Intel X86 Encoder Decoder) during the experiments and 1 new CVE in Radare2 (a popular reverse engineering framework). Furthermore, all the new bugs are confirmed by the developers and fixed.

Coverage-based Greybox Fuzzing as Markov Chain (CCS 2016)

Abstract: Coverage-based Greybox Fuzzing (CGF) is a random testing approach that requires no program analysis. A new test is generated by slightly mutating a seed input. If the test exercises a new and interesting path, it is added to the set of seeds; otherwise, it is discarded. We observe that most tests exercise the same few "high-frequency" paths and develop strategies to explore significantly more paths with the same number of tests by gravitating towards low-frequency paths. We explain the challenges and opportunities of CGF using a Markov chain model which specifies the probability that fuzzing the seed that exercises path i generates an input that exercises path j. Each state (i.e., seed) has an energy that specifies the number of inputs to be generated from that seed. We show that CGF is considerably more efficient if energy is inversely proportional to the density of the stationary distribution and increases monotonically every time that seed is chosen. Energy is controlled with a power schedule. We implemented the exponential schedule by extending AFL. In 24 hours, AFLFAST exposes 3 previously unreported CVEs that are not exposed by AFL and exposes 6 previously unreported CVEs 7x faster than AFL. AFLFAST produces at least an order of magnitude more unique crashes than AFL.

Program-Adaptive Mutational Fuzzing (S&P 2015)

[Paper](./Paper/SP15_Adaptive .pdf)

Abstract: We present the design of an algorithm to maximize the number of bugs found for black-box mutational fuzzing given a program and a seed input. The major intuition is to leverage white-box symbolic analysis on an execution trace for a given program-seed pair to detect dependencies among the bit positions of an input, and then use this dependency relation to compute a probabilistically optimal mutation ratio for this program-seed pair. Our result is promising: we found an average of 38.6% more bugs than three previous fuzzers over 8 applications using the same amount of fuzzing time.

Learning-based Fuzzing

Evaluating and Improving Neural Program-Smoothing-based Fuzzing (ICSE 2022)

Paper

Abstract: Fuzzing nowadays has been commonly modeled as an optimization problem, e.g., maximizing code coverage under a given time budget via typical search-based solutions such as evolutionary algorithms. However, such solutions are widely argued to cause inefficient computing resource usage, i.e., inefficient mutations. To address such issue, two neural program-smoothing-based fuzzers, Neuzz and MTFuzz, have been recently proposed by approximating program branching behaviors via a neural network model which inputs byte sequences of a seed and outputs vectors representing program branching behaviors. Moreover, assuming that mutating the bytes with larger gradients can better explore branching behaviors, they develop strategies to mutate such bytes for generating new seeds as test cases. Although they have been shown to be effective in their original papers, they were only evaluated upon a limited dataset. In addition, it is still unclear how their key technical components and whether other factors can impact their performance. To further investigate neural program-smoothing-based fuzzing, we first construct a large-scale benchmark with a total of 28 influential open-source projects. Then, we extensively evaluate Neuzz and MTFuzz on such benchmark where the results suggest that they can incur quite inconsistent edge coverage performance. Moreover, neither neural network models or mutation strategies can be consistently effective, and the power of their gradient guidance mechanisms have been compromised so far. Inspired by such findings, we propose an improved technique, namely RESuzz, upon neural program-smoothing-based fuzzers by enhancing their adopted gradient guidance mechanisms along with appending the AFL havoc mechanism. Our evaluation results indicate that it can significantly increase the edge coverage performance of Neuzz and MTFuzz. Furthermore, we also reveal multiple practical guidelines to advance future research.

Better Pay Attention Whilst Fuzzing (2022)

Paper

Abstract: Fuzzing is one of the prevailing methods for vulnerability detection. However, even state-of-the-art fuzzing methods become ineffective after some period of time, i.e., the coverage hardly improves as existing methods are ineffective to focus the attention of fuzzing on covering the hard-to-trigger program paths. In other words, they cannot generate inputs that can break the bottleneck due to the fundamental difficulty in capturing the complex relations between the test inputs and program coverage. In particular, existing fuzzers suffer from the following main limitations: 1) lacking an overall analysis of the program to identify the most "rewarding" seeds, and 2) lacking an effective mutation strategy which could continuously select and mutates the more relevant "bytes" of the seeds. In this work, we propose an approach called ATTuzz to address these two issues systematically. First, we propose a lightweight dynamic analysis technique which estimates the "reward" of covering each basic block and selects the most rewarding seeds accordingly. Second, we mutate the selected seeds according to a neural network model which predicts whether a certain "rewarding" block will be covered given certain mutation on certain bytes of a seed. The model is a deep learning model equipped with attention mechanism which is learned and updated periodically whilst fuzzing. Our evaluation shows that ATTuzz significantly outperforms 5 state-of-the-art grey-box fuzzers on 13 popular real-world programs at achieving higher edge coverage and finding new bugs. In particular, ATTuzz achieved 2X edge coverage and 4X bugs detected than AFL over 24-hour runs. Moreover, ATTuzz persistently improves the edge coverage in the long run, i.e., achieving 50% more coverage than AFL in 5 days.

Automated Conformance Testing for JavaScript Engines via Deep Compiler Fuzzing (PLDI 2021)

Paper

Abstract: JavaScript (JS) is a popular, platform-independent programming language. To ensure the interoperability of JS programs across different platforms, the implementation of a JS engine should conform to the ECMAScript standard. However, doing so is challenging as there are many subtle definitions of API behaviors, and the definitions keep evolving. We present COMFORT, a new compiler fuzzing framework for detecting JS engine bugs and behaviors that deviate from the ECMAScript standard. COMFORT leverages the recent advance in deep learning-based language models to automatically generate JS test code. As a departure from prior fuzzers, COMFORT utilizes the well-structured ECMAScript specifications to automatically generate test data along with the test programs to expose bugs that could be overlooked by the developers or manually written test cases. COMFORT then applies differential testing methodologies on the generated test cases to expose standard conformance bugs. We apply COMFORT to ten mainstream JS engines. In 200 hours of automated concurrent testing runs, we discover bugs in all tested JS engines. We had identified 158 unique JS engine bugs, of which 129 have been verified, and 115 have already been fixed by the developers. Furthermore, 21 of the Comfort-generated test cases have been added to Test262, the official ECMAScript conformance test suite.

Reinforcement Learning-based Hierarchical Seed Scheduling for Greybox Fuzzing (NDSS 2021)

Video
Paper

Abstract: Coverage metrics play an essential role in greybox fuzzing. Recent work has shown that fine-grained coverage metrics could allow a fuzzer to detect bugs that cannot be covered by traditional edge coverage. However, fine-grained coverage metrics will also select more seeds, which cannot be efficiently scheduled by existing algorithms. This work addresses this problem by introducing a new concept of multi-level coverage metric and the corresponding reinforcement-learning-based hierarchical scheduler. Evaluation of our prototype on DARPA CGC showed that our approach outperforms AFL and AFLFast significantly: it can detect 20% more bugs, achieve higher coverage on 83 out of 180 challenges, and achieve the same coverage on 60 challenges. More importantly, it can detect the same number of bugs and achieve the same coverage faster. On FuzzBench, our approach achieves higher coverage than AFL++ (Qemu) on 10 out of 20 projects.

Neural Network Guided Evolutionary Fuzzing for Finding Traffic Violations of Autonomous Vehicles (arXiv 2021)

Paper

Abstract: Self-driving cars and trucks, autonomous vehicles (AVs), should not be accepted by regulatory bodies and the public until they have much higher confidence in their safety and reliability -- which can most practically and convincingly be achieved by testing. But existing testing methods are inadequate for checking the end-to-end behaviors of AV controllers against complex, real-world corner cases involving interactions with multiple independent agents such as pedestrians and human-driven vehicles. While test-driving AVs on streets and highways fails to capture many rare events, existing simulation-based testing methods mainly focus on simple scenarios and do not scale well for complex driving situations that require sophisticated awareness of the surroundings. To address these limitations, we propose a new fuzz testing technique, called AutoFuzz, which can leverage widely-used AV simulators' API grammars. to generate semantically and temporally valid complex driving scenarios (sequences of scenes). AutoFuzz is guided by a constrained Neural Network (NN) evolutionary search over the API grammar to generate scenarios seeking to find unique traffic violations. Evaluation of our prototype on one state-of-the-art learning-based controller and two rule-based controllers shows that AutoFuzz efficiently finds hundreds of realistic traffic violations resembling real-world crashes. Further, fine-tuning the learning-based controller with the traffic violations found by AutoFuzz successfully reduced the traffic violations found in the new version of the AV controller software.

OmniFuzz: A Flexible Framework for Expediting Bug Finding by Leveraging Past (Mis-)Behavior to Discover New Bugs (ACSAC 2020)

Paper

Abstract: Among various fuzzing approaches, coverage-guided grey-box fuzzing is perhaps the most prominent, due to its ease of use and effectiveness. Using this approach, the selection of inputs focuses on maximizing program coverage, e.g., in terms of the different branches that have been traversed. In this work, we begin with the observation that selecting any input that explores a new path, and giving equal weight to all paths, can lead to severe inefficiencies. For instance, although seemingly "new" crashes involving previously unexplored paths may be discovered, these often have the same root cause and actually correspond to the same bug.

To address these inefficiencies, we introduce a framework that incorporates a tighter feedback loop to guide the fuzzing process in exploring truly diverse code paths. Our framework employs (i) a vulnerability-aware selection of coverage metrics for enhancing the effectiveness of code exploration, (ii) crash deduplication information for early feedback, and (iii) a configurable input culling strategy that interleaves multiple strategies to achieve comprehensiveness. A novel aspect of our work is the use of hardware performance counters to derive coverage metrics. We present an approach for assessing and selecting the hardware events that can be used as a meaningful coverage metric for a target program. The results of our empirical evaluation using real-world programs demonstrate the effectiveness of our approach: in some cases, we explore fewer than 50% of the paths compared to a base fuzzer (AFL, MOpt, and Fairfuzz), yet on average, we improve new bug discovery by 31%, and find the same bugs (as the base) 3.3 times faster. Moreover, although we specifically chose applications that have been subject to recent fuzzing campaigns, we still discovered 9 new vulnerabilities.

Learning Input Tokens for Effective Fuzzing (ISSTA 2020)

Paper

Abstract: Modern fuzzing tools like AFL operate at a lexical level: They explore the input space of tested programs one byte after another. For inputs with complex syntactical properties, this is very inefficient, as keywords and other tokens have to be composed one character at a time. Fuzzers thus allow to specify dictionaries listing possible tokens the input can be composed from; such dictionaries speed up fuzzers dramatically. Also, fuzzers make use of dynamic tainting to track input tokens and infer values that are expected in the input validation phase. Unfortunately, such tokens are usually implicitly converted to program specific values which causes a loss of the taints attached to the input data in the lexical phase.

In this paper we present a technique to extend dynamic tainting to not only track explicit data flows but also taint implicitly converted data without suffering from taint explosion. This extension makes it possible to augment existing techniques and automatically infer a set of tokens and seed inputs for the input language of a program given nothing but the source code. Specifically targeting the lexical analysis of an input processor, our lFuzzer test generator systematically explores branches of the lexical analysis, producing a set of tokens that fully cover all decisions seen. The resulting set of tokens can be directly used as a dictionary for fuzzing. Along with the token extraction seed inputs are generated which give further fuzzing processes a head start. In our experiments, the lFuzzer-AFL combination achieves up to 17% more coverage on complex input formats like JSON, LISP, tinyC, and JavaScript compared to AFL.

MTFuzz: Fuzzing with a Multi-task Neural Network (FSE 2020)

Paper

Abstract: Fuzzing is a widely used technique for detecting software bugs and vulnerabilities. Most popular fuzzers generate new inputs using an evolutionary search to maximize code coverage. Essentially, these fuzzers start with a set of seed inputs, mutate them to generate new inputs, and identify the promising inputs using an evolutionary fitness function for further mutation. Despite their success, evolutionary fuzzers tend to get stuck in long sequences of unproductive mutations. In recent years, machine learning (ML) based mutation strategies have reported promising results. However, the existing ML-based fuzzers are limited by the lack of quality and diversity of the training data. As the input space of the target programs is high dimensional and sparse, it is prohibitively expensive to collect many diverse samples demonstrating successful and unsuccessful mutations to train the model. In this paper, we address these issues by using a Multi-Task Neural Network that can learn a compact embedding of the input space based on diverse training samples for multiple related tasks (i.e., predicting different types of coverage). The compact embedding can be used to guide the mutation process effectively by focusing most of the mutations on the parts of the embedding where the gradient is high. Our results show that MTFuzz uncovers 11 previously unseen bugs and achieves an average of 2x more edge coverage compared with 5 state-of-the-art fuzzer on 10 real-world programs.

FuzzGuard: Filtering out Unreachable Inputs in Directed Grey-box Fuzzing through Deep Learning (USENIX Security2020)

Paper

Abstract: Recently, directed grey-box fuzzing (DGF) becomes much more popular in the field of software testing. Different from coverage-based fuzzing whose goal is to increase code coverage for triggering more bugs, DGF is designed to test only the potential buggy code whose location is known (e.g., testing the high-risk code such as string operations). For the efficiency, all the inputs generated by an ideal DGF should reach the buggy code, hoping to trigger the bug. Any unreachable input will waste the time spent on execution. Unfortunately, in a real situation, large numbers of the generated inputs miss the target, greatly impacting the efficiency of testing, especially when the buggy code is hard to reach.

In this paper, we propose a deep-learning-based approach to predict the reachability of inputs and filter out those unreachable ones, which works together with DGF fuzzers instead of replacing them. In the process of combining deep learning with fuzzing, we design a suite of new techniques (e.g., step-forwarding approach, representative data selection) to solve the problems of unbalanced labeled data and low efficiency. We implemented a prototype called FuzzGuard and evaluated it using 45 real vulnerabilities. The results show that FuzzGuard boosts the fuzzing efficiency of the state-of-the-art DGF (e.g., AFLGo) up to 17 times. We also found 19 undisclosed bugs, and 4 zero-day bugs (we have got CVE numbers). Finally, we design an approach to understand the extracted features of the deep neural network model, and find them correlate with the constraints in target programs, which indeed impacts the execution on the code level.

LearnAFL: Greybox Fuzzing With Knowledge Enhancement (Access 2019)

Paper

Abstract: Mutation-based greybox fuzzing is a highly effective and widely used technique to find bugs in software. Provided initial seeds, fuzzers continuously generate test cases to test the software by mutating a seed input. However, the majority of them are 'invalid' because the mutation may destroy the format of the seeds. In this paper, we present a knowledge-learn evolutionary fuzzer based on AFL, which is called LearnAFL. LearnAFL does not require any prior knowledge of the application or input format. Based on our format generation theory, LearnAFL can learn partial format knowledge of some paths by analyzing the test cases that exercise the paths. Then LearnAFL uses these format information to mutate the seeds, which is efficient to explore deeper paths and reduce the test cases exercising high-frequency paths than AFL. We compared LearnAFL with AFL and some other state-of-the-art fuzzers on ten real-world programs. The result showed that LearnAFL could reach branch coverage 120% and 110% of that of AFL and FairFuzz, respectively. LearnAFL also found 8 unknown vulnerabilities in GNU Binutils, Libpng and Gif2png, all of which have been reported to the vendors. Besides, we compared the format information learned from the initial seed of an ELF file with a format standard of ELF files. The result showed that LearnAFL learns about 64% part of the file format without any prior knowledge.

NeuFuzz: Efficient Fuzzing With Deep Neural Network (Access 2019)

Paper

Abstract: Coverage-guided gray box fuzzing is one of the most popular and effective techniques for discovering vulnerabilities due to its nature of high speed and scalability. However, the existing techniques generally focus on code coverage but not on vulnerable code. These techniques aim to cover as many paths as possible rather than to explore paths that are more likely to be vulnerable. When selecting the seeds to test, the existing fuzzers usually treat all seed inputs equally, ignoring the fact that paths exercised by different seed inputs are not equally vulnerable. This results in wasting time testing uninteresting paths rather than vulnerable paths, thus reducing the efficiency of vulnerability detection. In this paper, we present a solution, NeuFuzz, using the deep neural network to guide intelligent seed selection during gray box fuzzing to alleviate the aforementioned limitation. In particular, the deep neural network is used to learn the hidden vulnerability pattern from a large number of vulnerable and clean program paths to train a prediction model to classify whether paths are vulnerable. The fuzzer then prioritizes seed inputs that are capable of covering the likely to be vulnerable paths and assigns more mutation energy (i.e., the number of inputs to be generated) to these seeds. We implemented a prototype of NeuFuzz based on an existing fuzzer PTfuzz and evaluated it on two different test suites: LAVA-M and nine real-world applications. The experimental results showed that NeuFuzz can find more vulnerabilities than the existing fuzzers in less time. We have found 28 new security bugs in these applications, 21 of which have been assigned as CVE IDs.

Learning-Guided Network Fuzzing for Testing Cyber-Physical System Defences (ASE 2019)

Paper

Abstract: The threat of attack faced by cyber-physical systems (CPSs), especially when they play a critical role in automating public infrastructure, has motivated research into a wide variety of attack defence mechanisms. Assessing their effectiveness is challenging, however, as realistic sets of attacks to test them against are not always available. In this paper, we propose smart fuzzing, an automated, machine learning guided technique for systematically finding 'test suites' of CPS network attacks, without requiring any expertise in the system's control programs or physical processes. Our approach uses predictive machine learning models and metaheuristic search algorithms to guide the fuzzing of actuators so as to drive the CPS into different unsafe physical states. We demonstrate the efficacy of smart fuzzing by implementing it for two real-world CPS testbeds - a water purification plant and a water distribution system - finding attacks that drive them into 27 different unsafe states involving water flow, pressure, and tank levels, including six that were not covered by an established attack benchmark. Finally, we use our approach to test the effectiveness of an invariant-based defence system for the water treatment plant, finding two attacks that were not detected by its physical invariant checks, highlighting a potential weakness that could be exploited in certain conditions.

NEUZZ: Efficient Fuzzing with Neural Program Smoothing (S&P 2019)

Paper

Abstract: Fuzzing has become the de facto standard technique for finding software vulnerabilities. However, even state-of-the-art fuzzers are not very efficient at finding hard-to-trigger software bugs. Most popular fuzzers use evolutionary guidance to generate inputs that can trigger different bugs. Such evolutionary algorithms, while fast and simple to implement, often get stuck in fruitless sequences of random mutations. Gradient-guided optimization presents a promising alternative to evolutionary guidance. Gradient-guided techniques have been shown to significantly outperform evolutionary algorithms at solving high-dimensional structured optimization problems in domains like machine learning by efficiently utilizing gradients or higher-order derivatives of the underlying function. However, gradient-guided approaches are not directly applicable to fuzzing as real-world program behaviors contain many discontinuities, plateaus, and ridges where the gradient-based methods often get stuck. We observe that this problem can be addressed by creating a smooth surrogate function approximating the target program's discrete branching behavior. In this paper, we propose a novel program smoothing technique using surrogate neural network models that can incrementally learn smooth approximations of a complex, real-world program's branching behaviors. We further demonstrate that such neural network models can be used together with gradient-guided input generation schemes to significantly increase the efficiency of the fuzzing process. Our extensive evaluations demonstrate that NEUZZ significantly outperforms 10 state-of-the-art graybox fuzzers on 10 popular real-world programs both at finding new bugs and achieving higher edge coverage. NEUZZ found 31 previously unknown bugs (including two CVEs) that other fuzzers failed to find in 10 real-world programs and achieved 3X more edge coverage than all of the tested graybox fuzzers over 24 hour runs. Furthermore, NEUZZ also outperformed existing fuzzers on both LAVA-M and DARPA CGC bug datasets.

V-Fuzz: Vulnerability-Oriented Evolutionary Fuzzing (Arxiv 2019)

Paper

Abstract: Fuzzing is a technique of finding bugs by executing a software recurrently with a large number of abnormal inputs. Most of the existing fuzzers consider all parts of a software equally, and pay too much attention on how to improve the code coverage. It is inefficient as the vulnerable code only takes a tiny fraction of the entire code. In this paper, we design and implement avulnerability-oriented evolutionary fuzzing prototype named V-Fuzz, which aims to find bugs efficiently and quickly in a limited time. V-Fuzz consists of two main components: a neural network-based vulnerability prediction model and a vulnerability-oriented evolutionary fuzzer. Given a binary program to V-Fuzz, the vulnerability prediction model will give a prior estimation on which parts ofthe software are more likely to be vulnerable. Then, the fuzzer leverages an evolutionary algorithm to generate inputs which tend to arrive at the vulnerable locations, guided by the vulnerability prediction result. Experimental results demonstrate that V-Fuzz can find bugs more efficiently than state-of-the-art fuzzers. Moreover, V-Fuzz has discovered 10 CVEs, and 3 of them are newly discovered. We reported the new CVEs, and they have been confirmed and fixed.

Compiler Fuzzing through Deep Learning (ISSTA 2018)

Paper
Code

Abstract: Random program generation - fuzzing - is an effective technique for discovering bugs in compilers but successful fuzzers require extensive development effort for every language supported by the compiler, and often leave parts of the language space untested. We introduce DeepSmith, a novel machine learning approach to accelerating compiler validation through the inference of generative models for compiler inputs. Our approach infers a learned model of the structure of real world code based on a large corpus of open source code. Then, it uses the model to automatically generate tens of thousands of realistic programs. Finally, we apply established differential testing methodologies on them to expose bugs in compilers. We apply our approach to the OpenCL programming language, automatically exposing bugs with little effort on our side. In 1,000 hours of automated testing of commercial and open source compilers, we discover bugs in all of them, submitting 67 bug reports. Our test cases are on average two orders of magnitude smaller than the state-of-the-art, require 3.03× less time to generate and evaluate, and expose bugs which the state-of-the-art cannot. Our random program generator, comprising only 500 lines of code, took 12 hours to train for OpenCL versus the state-of-the-art taking 9 man months to port from a generator for C and 50,000 lines of code. With 18 lines of code we extended our program generator to a second language, uncovering crashes in Solidity compilers in 12 hours of automated testing.

Deep Reinforcement Fuzzing (SPW 2018)

Paper

Abstract: Fuzzing is the process of finding security vulnerabilities in input-processing code by repeatedly testing the code with modified inputs. In this paper, we formalize fuzzing as a reinforcement learning problem using the concept of Markov decision processes. This in turn allows us to apply state-of-the-art deep Q -learning algorithms that optimize rewards, which we define from runtime properties of the program under test. By observing the rewards caused by mutating with a specific set of actions performed on an initial program input, the fuzzing agent learns a policy that can next generate new higher-reward inputs. We have implemented this new approach, and preliminary empirical evidence shows that reinforcement fuzzing can outperform baseline random fuzzing.

FuzzerGym: A Competitive Framework for Fuzzing and Learning (arxiv 2018)

Paper

Abstract: Fuzzing is a commonly used technique designed to test software by automatically crafting program inputs. Currently, the most successful fuzzing algorithms emphasize simple, low-overhead strategies with the ability to efficiently monitor program state during execution. Through compile-time instrumentation, these approaches have access to numerous aspects of program state including coverage, data flow, and heterogeneous fault detection and classification. However, existing approaches utilize blind random mutation strategies when generating test inputs. We present a different approach that uses this state information to optimize mutation operators using reinforcement learning (RL). By integrating OpenAI Gym with libFuzzer we are able to simultaneously leverage advancements in reinforcement learning as well as fuzzing to achieve deeper coverage across several varied benchmarks. Our technique connects the rich, efficient program monitors provided by LLVM Santizers with a deep neural net to learn mutation selection strategies directly from the input data. The cross-language, asynchronous architecture we developed enables us to apply any OpenAI Gym compatible deep reinforcement learning algorithm to any fuzzing problem with minimal slowdown.

Learn&Fuzz: Machine Learning for Input Fuzzing (ASE 2017)

Paper

Abstract: Fuzzing consists of repeatedly testing an application with modified, or fuzzed, inputs with the goal of finding security vulnerabilities in input-parsing code. In this paper, we show how to automate the generation of an input grammar suitable for input fuzzing using sample inputs and neural-network-based statistical machine-learning techniques. We present a detailed case study with a complex input format, namely PDF, and a large complex security-critical parser for this format, namely, the PDF parser embedded in Microsoft's new Edge browser. We discuss and measure the tension between conflicting learning and fuzzing goals: learning wants to capture the structure of well-formed inputs, while fuzzing wants to break that structure in order to cover unexpected code paths and find bugs. We also present a new algorithm for this learn&fuzz challenge which uses a learnt input probability distribution to intelligently guide where to fuzz inputs.

Fuzzing Deep Learning Compilers and Inference Engines

Graph-based Fuzz Testing for Deep Learning Inference Engines (ICSE 2021)

Paper

Abstract: With the wide use of Deep Learning (DL) systems, academy and industry begin to pay attention to their quality. Testing is one of the major methods of quality assurance. However, existing testing techniques focus on the quality of DL models but lacks attention to the core underlying inference engines (i.e., frameworks and libraries). Inspired by the success stories of fuzz testing, we design a graph-based fuzz testing method to improve the quality of DL inference engines. This method is naturally followed by the graph structure of DL models. An operator-level coverage based on graph theory is introduced and six different mutations are implemented to generate diversified DL models by exploring combinations of model structures, parameters, and data. The Monte Carlo Tree Search (MCTS) is used to drive DL model generation without a training process. The experimental results show that the MCTS outperforms the random method in boosting operator-level coverage and detecting exceptions. Our method has discovered more than 40 different exceptions in three types of undesired behaviors: model conversion failure, inference failure, output comparison failure. The mutation strategies are useful to generate new valid test inputs, by up to an 8.2% more operator-level coverage on average and 8.6 more exceptions captured.

Coverage-Guided Tensor Compiler Fuzzing with Joint IR-Pass Mutation (OOPSLA 2022)

Abstract: In the past decade, Deep Learning (DL) systems have been widely deployed in various application domains to facilitate our daily life, e.g., natural language processing, healthcare, activity recognition, and autonomous driving. Meanwhile, it is extremely challenging to ensure the correctness of DL systems (e.g., due to their intrinsic nondeterminism), and bugs in DL systems can cause serious consequences and may even threaten human lives. In the literature, researchers have explored various techniques to test, analyze, and verify DL models, since their quality directly affects the corresponding system behaviors. Recently, researchers have also proposed novel techniques for testing the underlying operator-level DL libraries (such as TensorFlow and PyTorch), which provide general binary implementations for each high-level DL operator and are the foundation for running DL models on different hardware platforms. However, there is still limited work targeting the reliability of the emerging tensor compilers (also known as DL compilers), which aim to automatically compile high-level tensor computation graphs directly into high-performance binaries for better efficiency, portability, and scalability than traditional operator-level libraries. Therefore, in this paper, we target the important problem of tensor compiler testing, and have proposed Tzer, a practical fuzzing technique for the widely used TVM tensor compiler. Tzer focuses on mutating the low-level Intermediate Representation (IR) for TVM due to the limited mutation space for the high-level IR. More specifically, Tzer leverages both general-purpose and tensor-compiler-specific mutators guided by coverage feedback for diverse and evolutionary IR mutation; furthermore, since tensor compilers provide various passes (i.e., transformations) for IR optimization, Tzer also performs pass mutation in tandem with IR mutation for more effective fuzzing. Our experimental results show that Tzer substantially outperforms existing fuzzing techniques on tensor compiler testing, with 75% higher coverage and 50% more valuable tests than the 2nd-best technique. Also, different components of Tzer have been validated via ablation study. To date, Tzer has detected 49 previously unknown bugs for TVM, with 37 bugs confirmed and 25 bugs fixed (PR merged).

NNSmith: Generating Diverse and Valid Test Cases for Deep Learning Compilers (ASPLOS 2023)

Abstract: Deep-learning (DL) compilers such as TVM and TensorRT are increasingly being used to optimize deep neural network (DNN) models to meet performance, resource utilization and other requirements. Bugs in these compilers can result in models whose semantics differ from the original ones, producing incorrect results that corrupt the correctness of downstream applications. However, finding bugs in these compilers is challenging due to their complexity. In this work, we propose a new fuzz testing approach for finding bugs in deep-learning compilers. Our core approach consists of (i) generating diverse yet valid DNN test models that can exercise a large part of the compiler's transformation logic using light-weight operator specifications; (ii) performing gradient-based search to find model inputs that avoid any floating-point exceptional values during model execution, reducing the chance of missed bugs or false alarms; and (iii) using differential testing to identify bugs. We implemented this approach in NNSmith which has found 72 new bugs for TVM, TensorRT, ONNXRuntime, and PyTorch to date. Of these 58 have been confirmed and 51 have been fixed by their respective project maintainers.

Fuzzing Deep Learning Compilers with HirGen (ISSTA 2023)

Abstract: Deep Learning (DL) compilers are widely adopted to optimize advanced DL models for efficient deployment on diverse hardware. Their quality has profound effect on the quality of compiled DL models. A recent bug study shows that the optimization of high-level intermediate representation (IR) is the most error-prone compilation stage. Bugs in this stage are accountable for 44.92% of the whole collected ones. However, existing testing techniques do not consider high-level optimization related features (e.g. high-level IR), and are therefore weak in exposing bugs at this stage. To bridge this gap, we propose HirGen, an automated testing technique that aims to effectively expose coding mistakes in the optimization of high-level IR. The design of HirGen includes 1) three coverage criteria to generate diverse and valid computational graphs; 2) full use of high-level IRs language features to generate diverse IRs; 3) three test oracles inspired from both differential testing and metamorphic testing. HirGen has successfully detected 21 bugs that occur at TVM, with 17 bugs confirmed and 12 fixed. Further, we construct four baselines using the state-of-the-art DL compiler fuzzers that can cover the high-level optimization stage. Our experiment results show that HirGen can detect 10 crashes and inconsistencies that cannot be detected by the baselines in 48 hours. We further validate the usefulness of our proposed coverage criteria and test oracles in evaluation.

Fuzzing Deep Learning Libraries

DocTer: Documentation-Guided Fuzzing for Testing Deep Learning API Functions (ISSTA 2022)

Paper

Abstract: Input constraints are useful for many software development tasks. For example, input constraints of a function enable the generation of valid inputs, i.e., inputs that follow these constraints, to test the function deeper. API functions of deep learning (DL) libraries have DL-specific input constraints, which are described informally in the free-form API documentation. Existing constraint-extraction techniques are ineffective for extracting DL-specific input constraints.

To fill this gap, we design and implement a new technique, DocTer, to analyze API documentation to extract DL-specific input constraints for DL API functions. DocTer features a novel algorithm that automatically constructs rules to extract API parameter constraints from syntactic patterns in the form of dependency parse trees of API descriptions. These rules are then applied to a large volume of API documents in popular DL libraries to extract their input parameter constraints. To demonstrate the effectiveness of the extracted constraints, DocTer uses the constraints to enable the automatic generation of valid and invalid inputs to test DL API functions.

Our evaluation on three popular DL libraries (TensorFlow, PyTorch, and MXNet) shows that DocTer's precision in extracting input constraints is 85.4%. DocTer detects 94 bugs from 174 API functions, including one previously unknown security vulnerability that is now documented in the CVE database, while a baseline technique without input constraints detects only 59 bugs. Most (63) of the 94 bugs are previously unknown, 54 of which have been fixed or confirmed by developers after we report them. In addition, DocTer detects 43 inconsistencies in documents, 39 of which are fixed or confirmed.

Fuzzing Deep-Learning Libraries via Automated Relational API Inference (ESEC/FSE 2022)

Paper

Abstract: Deep Learning (DL) has gained wide attention in recent years. Meanwhile, bugs in DL systems can lead to serious consequences, and may even threaten human lives. As a result, a growing body of research has been dedicated to DL model testing. However, there is still limited work on testing DL libraries, e.g., PyTorch and TensorFlow, which serve as the foundations for building, training, and running DL models. Prior work on fuzzing DL libraries can only generate tests for APIs which have been invoked by documentation examples, developer tests, or DL models, leaving a large number of APIs untested. In this paper, we propose RelFuzz, which leverages API relations to further advance DL library testing. Our basic hypothesis is that for a DL library under test, there may exist a number of APIs sharing similar input parameters and outputs; in this way, we can easily 'borrow' test inputs from invoked APIs to test other relational APIs. Furthermore, we formalize the notion of value equivalence and status equivalence for relational APIs to serve as the oracle for effective bug finding. We have implemented RelFuzz as a fully automated fuzzing technique, which 1) automatically infers potential API relations based on API syntactic/semantic information, 2) synthesizes concrete test programs for invoking relational APIs, 3) validates the inferred relational APIs via representative test inputs, and finally 4) performs fuzzing on the verified relational APIs to find potential inconsistencies. Our evaluation on two of the most popular DL libraries, PyTorch and TensorFlow, demonstrates that RelFuzz can cover 157% more APIs than state-of-the-art FreeFuzz. To date, RelFuzz has detected 162 bugs in total, with 106 already confirmed by the developers as previously unknown bugs. Surprisingly, RelFuzz was able to detect 13.5% of the high-priority bugs for the entire PyTorch issue-tracking system since our first PyTorch bug report. Also, besides the 162 code bugs, we were also able to detect 14 documentation bugs (all confirmed).

Free Lunch for Testing: Fuzzing Deep-Learning Libraries from Open Source (ICSE 2022)

Paper
Code

Abstract: Deep learning (DL) systems can make our life much easier, and thus is gaining more and more attention from both academia and industry. Meanwhile, bugs in DL systems can be disastrous, and can even threaten human lives in safety-critical applications. To date, a huge body of research efforts have been dedicated to testing DL models. However, interestingly, there is still limited work for testing the underlying DL libraries, which are the foundation for building, optimizing, and running the DL models. One potential reason is that test generation for the underlying DL libraries can be rather challenging since their public APIs are mainly exposed in Python, making it even hard to automatically determine the API input parameter types due to dynamic typing. In this paper, we propose FreeFuzz, the first approach to fuzzing DL libraries via mining from open source. More specifically, FreeFuzz obtains code/models from three different sources: 1) code snippets from the library documentation, 2) library developer tests, and 3) DL models in the wild. Then, FreeFuzz automatically runs all the collected code/models with instrumentation to collect the dynamic information for each covered API, including the types and values of each parameter during invocation, and shapes of input/output tensors. Lastly, FreeFuzz will leverage the traced dynamic information to perform fuzz testing for each covered API. The extensive study of FreeFuzz on PyTorch and TensorFlow, two of the most popular DL libraries, shows that FreeFuzz is able to automatically trace valid dynamic information for fuzzing 1158 popular APIs, 9X more than state-of-the-art LEMON with 3.5X lower overhead than LEMON. Furthermore, FreeFuzz is able to detect 35 bugs for PyTorch and TensorFlow (with 31 confirmed by developers).

Muffin: Testing Deep Learning Libraries via Neural Architecture Fuzzing (ICSE 2022)

Paper

Abstract: Deep learning (DL) techniques are shown to be effective in many challenging tasks, and are hence widely-adopted in practice. However, previous work has shown that DL libraries, the basis of building and executing DL models, contain bugs and can cause severe consequences. Unfortunately, existing approaches still cannot comprehensively excise DL libraries. They utilize existing trained models and only detect bugs in model inference phase. In this work we propose Muffin to address these issues. To this end, Muffin applies a specifically-designed model fuzzing approach, which allows it to generate tremendous diverse DL models to explore the target library, instead of relying only on existing trained models. Muffin makes differential testing feasible in the model training phase by tailoring a set of metrics to measure the inconsistency between different DL libraries. In this way, Muffin can best excise the library code to detect more bugs. To evaluate the effectiveness of Muffin, we conduct experiments on three widely-used DL libraries. The results demonstrate that Muffin can detect 39 new bugs in the latest release versions of these libraries.

Fuzzing Machine Learning Model

MDPFuzz: Testing Models Solving Markov Decision Processes (ISSTA 2022)

Paper

Abstract: The Markov decision process (MDP) provides a mathematical framework for modeling sequential decision-making problems, many of which are crucial to security and safety, such as autonomous driving and robot control. The rapid development of artificial intelligence research has created efficient methods for solving MDPs, such as deep neural networks (DNNs), reinforcement learning (RL), and imitation learning (IL). However, these popular models solving MDPs are neither thoroughly tested nor rigorously reliable.

We present MDPFuzz, the first blackbox fuzz testing framework for models solving MDPs. MDPFuzz forms testing oracles by checking whether the target model enters abnormal and dangerous states. During fuzzing, MDPFuzz decides which mutated state to retain by measuring if it can reduce cumulative rewards or form a new state sequence. We design efficient techniques to quantify the 'freshness' of a state sequence using Gaussian mixture models (GMMs) and dynamic expectation-maximization (DynEM). We also prioritize states with high potential of revealing crashes by estimating the local sensitivity of target models over states.

MDPFuzz is evaluated on five state-of-the-art models for solving MDPs, including supervised DNN, RL, IL, and multi-agent RL. Our evaluation includes scenarios of autonomous driving, aircraft collision avoidance, and two games that are often used to benchmark RL. During a 12-hour run, we find over 80 crash-triggering state sequences on each model. We show inspiring findings that crash-triggering states, though look normal, induce distinct neuron activation patterns compared with normal states. We further develop an abnormal behavior detector to harden all the evaluated models and repair them with the findings of MDPFuzz to significantly enhance their robustness without sacrificing accuracy.

RapidFuzz: Accelerating fuzzing via Generative Adversarial Networks (Neurocomputing 2021)

Paper

Abstract: We implement a Generative Adversarial Network (GAN) based fuzzer called RapidFuzz to generate synthetic testcase, which can precisely catch the data structure feature in a relatively shorter time than the state-of-art fuzzers. RapidFuzz provides potential seeds generated by GAN. i.e., The generated seeds with similar but different numerical distributions accelerate the mutation process. An algorithm is elaborately designed to locate the hot-points generated by GAN. The generated testcases make structural features easier to be identified, which makes the whole process faster. In our experiment, RapidFuzz considerably improves the performance of American Fuzzy Lop(AFL) in speed, coverage, and mapsize. We select 9 open-sourced programs with different highly-structured inputs to demonstrate the effectiveness of RapidFuzz. As a result, code coverage is significantly improved. For tiff2pdf and tiffdump, coverage increase exceeds over 20%. We also observe that RapidFuzz achieves the same coverage with less time than AFL. Furthermore, AFL absorbs 21% of generated seed files in tiff2pdf with an average absorption rate around 15% in other programs.

CoCoFuzzing: Testing Neural Code Models with Coverage-Guided Fuzzing (2021)

Paper

Abstract: Deep learning-based code processing models have shown good performance for tasks such as predicting method names, summarizing programs, and comment generation. However, despite the tremendous progress, deep learning models are often prone to adversarial attacks, which can significantly threaten the robustness and generalizability of these models by leading them to misclassification with unexpected inputs. To address the above issue, many deep learning testing approaches have been proposed, however, these approaches mainly focus on testing deep learning applications in the domains of image, audio, and text analysis, etc., which cannot be directly applied to neural models for code due to the unique properties of programs. In this paper, we propose a coverage-based fuzzing framework, CoCoFuzzing, for testing deep learning-based code processing models. In particular, we first propose ten mutation operators to automatically generate valid and semantically preserving source code examples as tests; then we propose a neuron coverage-based approach to guide the generation of tests. We investigate the performance of CoCoFuzzing on three state-of-the-art neural code models, i.e., NeuralCodeSum, CODE2SEQ, and CODE2VEC. Our experiment results demonstrate that CoCoFuzzing can generate valid and semantically preserving source code examples for testing the robustness and generalizability of these models and improve the neuron coverage. Moreover, these tests can be used to improve the performance of the target neural code models through adversarial retraining.

Fuzz Testing based Data Augmentation to Improve Robustness of Deep Neural Networks (ICSE 2020)

Paper

Abstract: Deep neural networks (DNN) have been shown to be notoriously brittle to small perturbations in their input data. This problem is analogous to the over-fitting problem in test-based program synthesis and automatic program repair, which is a consequence of the incomplete specification, the limited tests or training examples, that the program synthesis or repair algorithm has to learn from. Recently, test generation techniques have been successfully employed to augment existing specifications of intended program behavior, to improve the generalizability of program synthesis and repair. Inspired by these approaches, in this paper, we propose a technique that re-purposes software testing methods, specifically mutation-based fuzzing, to augment the training data of DNNs, with the objective of enhancing their robustness. Our technique casts the DNN data augmentation problem as an optimization problem. It uses genetic search to generate the most suitable variant of an input data to use for training the DNN, while simultaneously identifying opportunities to accelerate training by skipping augmentation in many instances. We instantiate this technique in two tools, SENSEI and SENSEI-SA, and evaluate them on 15 DNN models spanning 5 popular image data-sets. Our evaluation shows that SENSEI can improve the robust accuracy of the DNN, compared to the state of the art, on each of the 15 models, by upto 11.9% and 5.5% on average. Further, SENSEI-SA can reduce the average DNN training time by 25%, while still improving robust accuracy.

Coverage Guided Differential Adversarial Testing of Deep Learning Systems (TNSE 2020)

Paper

Abstract: Deep learning is increasingly applied to safety-critical application domains such as autonomous cars and medical devices. It is of significant importance to ensure their reliability and robustness. In this paper, we propose DLFuzz, the coverage guided differential adversarial testing framework to guide deep learing systems exposing incorrect behaviors. DLFuzz keeps minutely mutating the input to maximize the neuron coverage and the prediction difference between the original input and the mutated input, without manual labeling effort or cross-referencing oracles from other systems with the same functionality. We also design multiple novel strategies for neuron selection to improve the neuron coverage. The incorrect behaviors obtained by DLFuzz are then exploited for retraining and improving the dependability of the models. We present empirical evaluations on two well-known datasets to demonstrate its effectiveness. Compared with DeepXplore, the state-of-the-art deep learning white-box testing framework, DLFuzz does not require extra efforts to find similar functional deep learning systems for cross-referencing check. But DLFuzz could generate 338.59% more adversarial inputs with 89.82% smaller perturbations, while maintaining the identities of the original inputs. DLFuzz also managed to averagely obtain 2.86% higher neuron coverage, and save 20.11% time consumption with respect to DeepXplore. We then evaluate the effectiveness of strategies for neuron selection, and demonstrated that all these strategies perform better than DeepXplore. Finally, DLFuzz proved to be able to improve the accuracy of deep learning systems by incorporating these adversarial inputs to retrain.

CAGFuzz: Coverage-Guided Adversarial Generative Fuzzing Testing of Deep Learning Systems (2019)

Paper

Abstract: Deep Learning systems (DL) based on Deep Neural Networks (DNNs) are more and more used in various aspects of our life, including unmanned vehicles, speech processing, and robotics. However, due to the limited dataset and the dependence on manual labeling data, DNNs often fail to detect their erroneous behaviors, which may lead to serious problems. Several approaches have been proposed to enhance the input examples for testing DL systems. However, they have the following limitations. First, they design and generate adversarial examples from the perspective of model, which may cause low generalization ability when they are applied to other models. Second, they only use surface feature constraints to judge the difference between the adversarial example generated and the original example. The deep feature constraints, which contain high-level semantic information, such as image object category and scene semantics are completely neglected. To address these two problems, in this paper, we propose CAGFuzz, a Coverage-guided Adversarial Generative Fuzzing testing approach, which generates adversarial examples for a targeted DNN to discover its potential defects. First, we train an adversarial case generator (AEG) from the perspective of general data set. Second, we extract the depth features of the original and adversarial examples, and constrain the adversarial examples by cosine similarity to ensure that the semantic information of adversarial examples remains unchanged. Finally, we retrain effective adversarial examples to improve neuron testing coverage rate. Based on several popular data sets, we design a set of dedicated experiments to evaluate CAGFuzz. The experimental results show that CAGFuzz can improve the neuron coverage rate, detect hidden errors, and also improve the accuracy of the target DNN.

DeepHunter: A Coverage-Guided Fuzz Testing Framework for Deep Neural Networks (ISSTA 2019)

Paper

Abstract: In company with the data explosion over the past decade, deep neural network (DNN) based software has experienced unprecedented leap and is becoming the key driving force of many novel industrial applications, including many safety-critical scenarios such as autonomous driving. Despite great success achieved in various human intelligence tasks, similar to traditional software, DNNs could also exhibit incorrect behaviors caused by hidden defects causing severe accidents and losses. In this paper, we propose DeepHunter, an automated fuzz testing framework for hunting potential defects of general-purpose DNNs. DeepHunter performs metamorphic mutation to generate new semantically preserved tests, and leverages multiple plugable coverage criteria as feedback to guide the test generation from different perspectives. To be scalable towards practical-sized DNNs, DeepHunter maintains multiple tests in a batch, and prioritizes the tests selection based on active feedback. The effectiveness of DeepHunter is extensively investigated on 3 popular datasets (MNIST, CIFAR-10, ImageNet) and 7 DNNs with diverse complexities, under a large set of 6 coverage criteria as feedback. The large-scale experiments demonstrate that DeepHunter can (1) significantly boost the coverage with guidance; (2) generate useful tests to detect erroneous behaviors and facilitate the DNN model quality evaluation; (3) accurately capture potential defects during DNN quantization for platform migration

TensorFuzz: Debugging Neural Networks with Coverage-Guided Fuzzing (ICML 2019)

Abstract: Machine learning models are notoriously difficult to interpret and debug. This is particularly true of neural networks. In this work, we introduce automated software testing techniques for neural networks that are well-suited to discovering errors which occur only for rare inputs. Specifically, we develop coverage-guided fuzzing (CGF) methods for neural networks. In CGF, random mutations of inputs to a neural network are guided by a coverage metric toward the goal of satisfying user-specified constraints. We describe how fast approximate nearest neighbor algorithms can provide this coverage metric. We then discuss the application of CGF to the following goals: finding numerical errors in trained neural networks, generating disagreements between neural networks and quantized versions of those networks, and surfacing undesirable behavior in character level language models. Finally, we release an open source library called TensorFuzz that implements the described techniques.

DLFuzz: Differential Fuzzing Testing of Deep Learning Systems (FSE 2018)

Abstract: Deep learning (DL) systems are increasingly applied to safety-critical domains such as autonomous driving cars. It is of significant importance to ensure the reliability and robustness of DL systems. Existing testing methodologies always fail to include rare inputs in the testing dataset and exhibit low neuron coverage. In this paper, we propose DLFuzz, the frst differential fuzzing testing framework to guide DL systems exposing incorrect behaviors. DLFuzz keeps minutely mutating the input to maximize the neuron coverage and the prediction difference between the original input and the mutated input, without manual labeling effort or cross-referencing oracles from other DL systems with the same functionality. We present empirical evaluations on two well-known datasets to demonstrate its efficiency. Compared with DeepXplore, the state-of-the-art DL whitebox testing framework, DLFuzz does not require extra efforts to find similar functional DL systems for cross-referencing check, but could generate 338.59% more adversarial inputs with 89.82% smaller perturbations, averagely obtain 2.86% higher neuron coverage, and save 20.11% time consumption.

Data Flow Sensitive Fuzzing

PATA: Fuzzing with Path Aware Taint Analysis (S&P 2022)

Paper

Abstract: Taint analysis assists fuzzers in solving complex fuzzing constraints by inferring the influencing input bytes. Execution paths in real-world programs often reach loops, where constraints in these loops can be visited and recorded multiple times. Conventional taint analysis techniques experience difficulties when distinguishing between multiple occurrences of the same constraint. In this paper, we propose PATA, a fuzzer that implements path-aware taint analysis, i.e. one that distinguishes between multiple occurrences of the same variable based on the execution path information. PATA does so using the following steps. First, PATA identifies variables used in constraints and constructs the Representative Variable Sequence (RVS), consisting of occurrences of all representative constraint variables and their values. Next, PATA perturbs the input, matches its RVS with that of the original input, and looks for value changes to identify the influencing input bytes for each entry in the RVS. Finally, PATA mutates the corresponding input bytes to solve constraints in the given path. To demonstrate the effectiveness of PATA over conventional taint analysis methods, we evaluated its performance on the benchmarks Google's fuzzer-test-suite and LAVA-M against AFL, MOPT, TortoriseFuzz, VUzzer, Angora, Redqueen, and Greyone. On Google's fuzzer-test-suite, PATA outperformed these state-of-the-art fuzzers by 29%-1830% and 7%-87% in the number of unique paths found and basic blocks covered, respectively. More importantly, it found more bugs than the comparison fuzzers, including 17 unlisted ones. On LAVA-M, PATA performed the best out of all evaluated fuzzers and found 2602 bugs. On open-source projects, PATA found 40 previously unknown bugs, with 12 of them confirmed as CVEs.

datAFLow: Towards a Data-Flow-Guided Fuzzer (NDSS 2022)

Code
Paper

Abstract: Coverage-guided greybox fuzzers rely on feedback derived from control-flow coverage to explore a target program and uncover bugs. This is despite control-flow feedback offering only a coarse-grained approximation of program behavior. Data flow intuitively more-accurately characterizes program behavior. Despite this advantage, fuzzers driven by data-flow coverage have received comparatively little attention, appearing mainly when heavyweight program analyses (e.g., taint analysis, symbolic execution) are used. Unfortunately, these more accurate analyses incur a high run-time penalty, impeding fuzzer throughput. Lightweight data-flow alternatives to control-flow fuzzing remain unexplored.

We present DATAFLOW, a greybox fuzzer driven by lightweight data-flow profiling. Whereas control-flow edges represent the order of operations in a program, data-flow edges capture the dependencies between operations that produce data values and the operations that consume them: indeed, there may be no control dependence between those operations. As such, data-flow coverage captures behaviors not visible as control flow and intuitively discovers more or different bugs. Moreover, we establish a framework for reasoning about data-flow coverage, allowing the computational cost of exploration to be balanced with precision.

We perform a preliminary evaluation of DATAFLOW, comparing fuzzers driven by control flow, taint analysis (both approximate and exact), and data flow. Our initial results suggest that, so far, pure coverage remains the best coverage metric for uncovering bugs in most targets we fuzzed (72 % of them). However, data-flow coverage does show promise in targets where control flow is decoupled from semantics (e.g., parsers). Further evaluation and analysis on a wider range of targets is required.

ovAFLow: Detecting Memory Corruption Bugs with Fuzzing-based Taint Inference (Journal of Computer Science and Technology 2021)

Paper

Abstract: Grey-box fuzzing is an effective technology to detect software vulnerabilities, such as memory corruption. Previous fuzzers in detecting memory corruption bugs either use heavy-weight analysis, or use techniques which are not customized for memory corruption detection. In this paper, we propose a novel memory bug guided fuzzer, ovAFLow. To begin with, we broaden the memory corruption targets where we frequently identify bugs. Next, ovAFLow utilizes light-weight and effective methods to build connections between the fuzzing inputs and these corruption targets. Based on the connection results, ovAFLow uses customized techniques to direct the fuzzing process closer to memory corruption. We evaluate ovAFLow against state-of-the-art fuzzers, including AFL (american fuzzy lop), AFLFast, FairFuzz, QSYM, Angora, TIFF, and TortoiseFuzz. The evaluation results show better vulnerability detection ability of ovAFLow, and the performance overhead is acceptable. Moreover, we identify 12 new memory corruption bugs and two CVEs (common vulnerability exposures) with the help of ovAFLow.

DIAR: Removing Uninteresting Bytes from Seeds in Software Fuzzing (2021)

Paper

Abstract: Software fuzzing mutates bytes in the test seeds to explore different behaviors of the program under test. Initial seeds can have great impact on the performance of a fuzzing campaign. Mutating a lot of uninteresting bytes in a large seed wastes the fuzzing resources. In this paper, we present the preliminary results of our approach that aims to improve the performance of fuzzers through identifying and removing uninteresting bytes in the seeds. In particular, we present DIAR, a technique that reduces the size of the seeds based on their coverage. Our preliminary results suggest fuzzing campaigns that start with reduced seeds, find new paths faster, and can produce higher coverage overall.

GREYONE: Data Flow Sensitive Fuzzing (USENIX Security2020)

Abstract: Data flow analysis (e.g., dynamic taint analysis) has proven to be useful for guiding fuzzers to explore hard-to-reach code and find vulnerabilities. However, traditional taint analysis is labor-intensive, inaccurate and slow, affecting the fuzzing efficiency. Apart from taint, few data flow features are utilized. In this paper, we proposed a data flow sensitive fuzzing solution GREYONE. We first utilize the classic feature taint to guide fuzzing. A lightweight and sound fuzzing-driven taint inference (FTI) is adopted to infer taint of variables, by monitoring their value changes while mutating input bytes during fuzzing. With the taint, we propose a novel input prioritization model to determine which branch to explore, which bytes to mutate and how to mutate. Further, we use another data flow feature constraint conformance, i.e., the distance of tainted variables to values expected in untouched branches, to tune the evolution direction of fuzzing.

We implemented a prototype of GREYONE and evaluated it on the LAVA data set and 19 real-world programs. The results showed that it outperforms various state-of-the-art fuzzers in terms of both code coverage and vulnerability discovery. In the LAVA data set, GREYONE found all listed bugs and 336 more unlisted. In real-world programs, GREYONE on average found 2.12X unique program paths and 3.09X unique bugs than state-of-the-art evolutionary fuzzers, including AFL, VUzzer, CollAFL, Angora and Honggfuzz, Moreover, GREYONE on average found 1.2X unique program paths and 1.52X unique bugs than a state-of-the-art symbolic execution assisted fuzzer QSYM. In total, it found 105 new security bugs, of which 41 are confirmed by CVE.

Binary Fuzzing

spotFuzzer: Static Instrument and Fuzzing Windows COTs (2022)

Paper

Abstract: The security research on Windows has received little attention in the academic circle. Most of the new methods are usually designed for Linux system, and are difficult to transplant to Windows. Fuzzing for Windows programs always suffering from its closed source. Therefore, we need to find an appropriate way to achieve feedback from Windows programs. To our knowledge, there are no stable and scalable static instrumentation tools for Windows yet, and dynamic tools, such as DynamoRIO, have been criticized for their performance. To make matters worse, dynamic instrumentation tools have very limited usage scenarios and are impotent for many system services or large commercial software. In this paper, we proposed spotInstr, a novel static tool for instrumenting Windows binaries. It is lightweight and can instrument most Windows PE programs in a very short time. At the same time, spotInstr provides a set of filters, which can be used to select instrumentation points or restrict the target regions. Based on these filters, we propose a novel memory-sensitive instrumentation method which can speed up both instrumentation and fuzzing. After that, we design a system called spotFuzzer, which leverage the ability of spotInstr and can fuzz most Windows binaries. We tested spotInstr and spotFuzzer in multiple dimensions to show their superior performance and stability.

Emmutaler: Fuzzing the iOS Boot Loader (Thesis 2021)

Code
Paper

Abstract: With the rise of smartphones and their importance in everyday life, they have also become an increasingly larger target of cyberattacks. iPhones in particular claim to be very secure and also privacy focused. This starts at the secure boot loader, which forms the beginning of a strong chain of trust. It stands to reason, that breaking this chain of trust would be of particular interest to adversaries.

In this thesis, we present emmutaler, a set of tools to enable fuzzing of the iPhone boot loader. After intensive static analysis of the boot loader, emmutaler allows us to create a normal Linux ELF containing it. This allows us to fuzz specific parts efficiently, using conventional Linux fuzzers. We present the design and implementation of emmutaler, as well as multiple fuzzing experiments ran using it, such as fuzzing the USB and firmware parsing stack. Furthermore, we introduce FETA, a novel, thread-safe allocator tailored towards fuzzing, that enables the discovery of heap vulnerabilities hidden by other allocators. In particular, we managed to find 'checkm8' - a previously known UAF vulnerability - only using FETA and hence showed that it is still present on A13 chips. Lastly, we discuss some innovative mitigations in use by iPhone boot loaders, that we found during the development of emmutaler.

Same Coverage, Less Bloat: Accelerating Binary-only Fuzzing with Coverage-preserving Coverage-guided Tracing (CCS 2021)

Paper

Abstract: Coverage-guided fuzzing's aggressive, high-volume testing has helped reveal tens of thousands of software security flaws. While executing billions of test cases mandates fast code coverage tracing, the nature of binary-only targets leads to reduced tracing performance. A recent advancement in binary fuzzing performance is Coverage-guided Tracing (CGT), which brings orders-of-magnitude gains in throughput by restricting the expense of coverage tracing to only when new coverage is guaranteed. Unfortunately, CGT suits only a basic block coverage granularity - yet most fuzzers require finer-grain coverage metrics: edge coverage and hit counts. It is this limitation which prohibits nearly all of today's state-of-the-art fuzzers from attaining the performance benefits of CGT.

This paper tackles the challenges of adapting CGT to fuzzing's most ubiquitous coverage metrics. We introduce and implement a suite of enhancements that expand CGT's introspection to fuzzing's most common code coverage metrics, while maintaining its orders-of-magnitude speedup over conventional always-on coverage tracing. We evaluate their trade-offs with respect to fuzzing performance and effectiveness across 12 diverse real-world binaries (8 open- and 4 closed-source). On average, our coverage-preserving CGT attains near-identical speed to the present block-coverageonly CGT, UnTracer; and outperforms leading binary- and sourcelevel coverage tracers QEMU, Dyninst, RetroWrite, and AFL-Clang by 2-24x, finding more bugs in less time.

Scalable Fuzzing of Program Binaries with E9AFL (ASE 2021)

Abstract: Greybox fuzzing is an effective method for software testing. Greybox fuzzers, e.g., AFL, use instrumentation to collect path coverage information to guide the test generation. The instrumentation is usually inserted by a modified compiler toolchain, meaning that the program must be recompiled in order to be compatible with greybox fuzzing. When source code is unavailable, or for projects with complex build systems, recompilation is not always feasible. In this paper, we present E9AFL, a fast and scalable tool that inserts AFL instrumentation to program binaries. E9AFL is built on top of the E9Patch static binary rewriting tool. To combat the overhead caused by binary instrumentation, E9AFL develops a set of optimization strategies. Evaluation results show that E9AFL outperforms existing binary instrumentation tools and achieves comparable performance with the compile time instrumentation.

STOCHFUZZ: Sound and Cost-effective Fuzzing of Stripped Binaries by Incremental and Stochastic Rewriting (S&P 2021)

Abstract: Fuzzing stripped binaries poses many hard challenges as fuzzers require instrumenting binaries to collect runtime feedback for guiding input mutation. However, due to the lack of symbol information, correct instrumentation is difficult on stripped binaries. Existing techniques either rely on hardware and expensive dynamic binary translation engines such as QEMU, or makes impractical assumptions such as binaries do not have inlined data. We observe that fuzzing is a highly repetitive procedure providing a large number of trial-and-error opportunities. As such, we propose a novel incremental and stochastic rewriting technique StochFuzz that piggy-backs on the fuzzing procedure. It generates many different versions of rewritten binaries whose validity can be approved/disapproved by numerous fuzzing runs. Probabilistic analysis is used to aggregate evidence collected through the sample runs and improve rewriting. The process eventually converges on a correctly rewritten binary. We evaluate StochFuzz on two sets of real-world programs and compare with five other baselines. The results show that StochFuzz outperforms state-of-the-art binary-only fuzzers (e.g., e9patch, ddisasm, and RetroWrite) in terms of soundness and cost-effectiveness and achieves performance comparable to source-based fuzzers. StochFuzz is publicly available at https://github.com/ZhangZhuoSJTU/StochFuzz.

Coverage-guided binary fuzzing with rev.ng and llvm libfuzzer

Paper

Abstract: Software vulnerabilities remain one of the most significant threats challenging information security. Memory-unsafe programming languages such as C/C++ allow bugs that could be maliciously exploited in order to tamper with the system, exposing users to risk. Fuzzing is a widespread and effective software testing technique for discovering unknown security vulnerabilities, which consists in executing a program continuously with a large number of random inputs to cause a crash. Most state-of-the-art fuzzers are coverage-based, i.e., they leverage the code coverage to favour inputs that contribute at exploring new paths. The coverage is collected by instrumenting the program source code at compile time; and, intuitively, a higher code coverage can lead fuzzers to find more bugs. However, when it comes to fuzzing binary-only software (source-unavailable), compiler instrumentation is not possible, subjecting the field to further research.

rev.ng is a reverse engineering framework based on LLVM. It features a static binary translation tool that translates machine code into LLVM IR, a higher-level intermediate representation, independent from the input architecture, and suitable to perform a variety of analyses and transformations. From a security standpoint, it allows to instrument programs in order to analyze their behavior.

In this work, we propose a novel framework based on rev.ng and libFuzzer, the LLVM fuzz testing library, to perform coverage-guided binary fuzzing of executable programs. We show that our approach is fast, semantic-preserving and simply requires to implement the so-called fuzz target, a dedicated function that the fuzzer calls to pass its inputs to the interesting functions to fuzz, just as occurs for programs with source code available. We evaluate our framework on a real-world vulnerability and we show we manage to reproduce the bug successfully

Breaking Through Binaries: Compiler-quality Instrumentation for Better Binary-only Fuzzing (USENIX Security2021)

Paper

Abstract: Coverage-guided fuzzing is one of the most effective software security testing techniques. Fuzzing takes on one of two forms: compiler-based or binary-only, depending on the availability of source code. While the fuzzing community has improved compiler-based fuzzing with performanceand feedback-enhancing program transformations, binaryonly fuzzing lags behind due to the semantic and performance limitations of instrumenting code at the binary level. Many fuzzing use cases are binary-only (i.e., closed source). Thus, applying fuzzing-enhancing program transformations to binary-only fuzzing - without sacrificing performance - remains a compelling challenge.

This paper examines the properties required to achieve compiler-quality binary-only fuzzing instrumentation. Based on our findings, we design FIBRE: a platform for applying fuzzing-enhancing program transformation to binary-only targets - maintaining compiler-level performance. We showcase FIBRE's capabilities in an implementation for the popular fuzzer AFL, including five compiler-style fuzzing-enhancing transformations, and evaluate it against the leading binaryonly fuzzing instrumenters AFL-QEMU and AFL-Dyninst. Across LAVA-M and real-world targets, FIBRE improves crash-finding by 26-96% and 37-131%; and throughput by 48-78% and 159-203% compared to AFL-Dyninst and AFLQEMU, respectively - while maintaining compiler-level of overhead of 27%. We also show that FIBRE supports realworld open- and closed-source software of varying size (10K-100MB), complexity (100-1M basic blocks), platform (Linux and Windows), and format (e.g., stripped and PIC).

WEIZZ: Automatic Grey-box Fuzzing for Structured Binary Formats

Video
Paper

Abstract: Fuzzing technologies have evolved at a fast pace in recent years, revealing bugs in programs with ever increasing depth and speed. Applications working with complex formats are however more difficult to take on, as inputs need to meet certain format-specific characteristics to get through the initial parsing stage and reach deeper behaviors of the program.

Unlike prior proposals based on manually written format specifications, we propose a technique to automatically generate and mutate inputs for unknown chunk-based binary formats. We identify dependencies between input bytes and comparison instructions, and use them to assign tags that characterize the processing logic of the program. Tags become the building block for structure-aware mutations involving chunks and fields of the input.

Our technique can perform comparably to structure-aware fuzzing proposals that require human assistance. Our prototype implementation WEIZZ revealed 16 unknown bugs in widely used programs.

Ptrix: Efficient Hardware-Assisted Fuzzing for COTS Binary (ASIACCS 2019)

Paper

Abstract: Despite its effectiveness in uncovering software defects, American Fuzzy Lop (AFL), one of the best grey-box fuzzers, is inefficient when fuzz-testing source-unavailable programs. AFL's binary-only fuzzing mode, QEMU-AFL, is typically 2-5X slower than its source-available fuzzing mode. The slowdown is largely caused by the heavy dynamic instrumentation. Recent fuzzing techniques use Intel Processor Tracing (PT), a light-weight tracing feature supported by recent Intel CPUs, to remove the need of dynamic instrumentation. However, we found that these PT-based fuzzing techniques are even slower than QEMU-AFL when fuzzing real-world programs, making them less effective than QEMU-AFL. This poor performance is caused by the slow extraction of code coverage information from highly compressed PT traces. In this work, we present the design and implementation of PTrix, which fully unleashes the benefits of PT for fuzzing via three novel techniques. First, PTrix introduces a scheme to highly parallel the processing of PT trace and target program execution. Second, it directly takes decoded PT trace as feedback for fuzzing, avoiding the expensive reconstruction of code coverage information. Third, PTrix maintains the new feedback with stronger feedback than edge-based code coverage, which helps reach new code space and defects that AFL may not. We evaluated PTrix by comparing its performance with the state-of-the-art fuzzers. Our results show that, given the same amount of time, PTrix achieves a significantly higher fuzzing speed and reaches into code regions missed by the other fuzzers. In addition, PTrix identifies 35 new vulnerabilities in a set of previously well-fuzzed binaries, showing its ability to complement existing fuzzers.

Steelix: Program-State Based Binary Fuzzing (FSE 2017)

Abstract: Coverage-based fuzzing is one of the most effective techniques to find vulnerabilities, bugs or crashes. However, existing techniques suffer from the difficulty in exercising the paths that are protected by magic bytes comparisons (e.g., string equality comparisons). Several approaches have been proposed to use heavy-weight program analysis to break through magic bytes comparisons, and hence are less scalable. In this paper, we propose a program-state based binary fuzzing approach, named Steelix, which improves the penetration power of a fuzzer at the cost of an acceptable slow down of the execution speed. In particular, we use light-weight static analysis and binary instrumentation to provide not only coverage information but also comparison progress information to a fuzzer. Such program state information informs a fuzzer about where the magic bytes are located in the test input and how to perform mutations to match the magic bytes efficiently. We have implemented Steelix and evaluated it on three datasets: LAVA-M dataset, DARPA CGC sample binaries and five real-life programs. The results show that Steelix has better code coverage and bug detection capability than the state-of-the-art fuzzers. Moreover, we found one CVE and nine new bugs.

In-memory fuzzing for binary code similarity analysis (ASE 2017)

Paper

Abstract: Detecting similar functions in binary executables serves as a foundation for much binary code analysis and reuse tasks. By far, recognizing similar components in binary code remains a challenge. Existing research employs either static or dynamic approaches to capture program syntax or semantics level features for comparison. However, there exist multiple design limitations in previous work, which result in relatively high cost, low accuracy, and scalability, and thus severely impede their practical use.

In this paper, we present a novel method that leverages in-memory fuzzing for binary code similarity analysis. Our prototype tool IMF-SIM applies in-memory fuzzing to launch analysis towards every function and collect traces of different kinds of program behaviors. The similarity score of two behavior traces is computed according to their longest common subsequence. To compare two functions, a feature vector is generated, whose elements are the similarity scores of the behavior trace-level comparisons. We train a machine learning model through labeled feature vectors; later, for a given feature vector by comparing two functions, the trained model gives a final score, representing the similarity score of the two functions. We evaluate IMF-SIM against binaries compiled by different compilers, optimizations, and commonly-used obfuscation methods, in total over one thousand binary executables. Our evaluation shows that IMFSIM notably outperforms existing tools with higher accuracy and broader application scopes.

Smart Contracts

WASAI: Uncovering Vulnerabilities in Wasm Smart Contracts (ISSTA 2022)

Paper

Abstract: WebAssembly (Wasm) smart contracts have shown growing popularity across blockchains (e.g., EOSIO) recently. Similar to Ethereum smart contracts, Wasm smart contracts suffer from various attacks exploiting their vulnerabilities. Even worse, few developers released the source code of their Wasm smart contracts for security review, raising the bar for uncovering vulnerable contracts. Although a few approaches have been proposed to detect vulnerable Wasm smart contracts, they have several major limitations, e.g., low code coverage, low accuracy and lack of scalability, unable to produce exploit payloads, etc. To fill the gap, in this paper, we design and develop WASAI, a new concolic fuzzer for uncovering vulnerabilities in Wasm smart contract after tackling several challenging issues. We conduct extensive experiments to evaluate WASAI, and the results show that it outperforms the state-of-the-art methods. For example, it achieves 2x code coverage than the baselines and surpasses them in detection accuracy, with an F1-measure of 99.2%. Moreover, WASAI can handle complicated contracts (e.g., contracts with obfuscation and sophisticated verification). Applying WASAI to 991 deployed smart contracts in the wild, we find that over 70% of smart contracts are vulnerable. By the time of this study, over 300 vulnerable contracts have not been patched and are still operating on the EOSIO Mainnet. One fake EOS vulnerability reported to the EOSIO ecosystem was recently assigned a CVE identifier (CVE-2022-27134).

Smart Contract Fuzzing for Enterprises: The Language Agnostic Way (COMSNETS 2022)

Paper

Abstract: Blockchain based applications backed by smart contracts are becoming increasingly popular in various domains. Smart contracts are vulnerable to attacks due to bugs in them and such attacks resulted in huge monetary losses, disruption in operation and so on in the past. Fuzz testing is one of the prominent methods used for identifying bugs in blockchain smart contracts. Multiple fuzzers are used for fuzzing smart contracts written in different programming languages. However, maintenance of multiple fuzzers become prohibitively difficult in an enterprise DevOps setup, in terms of skills, time, and efforts required in patching and keeping them up to date. Hence, we propose a novel vulnerability detection framework which uses a single fuzzer to fuzz smart contracts written in different programming languages, using LLVM IR. In this paper, we validated the proposed framework by testing Hyperledger Fabric smart contracts using fuzzing tools AFL++ and Honggfuzz and presented the results.

looking for lacunae in bitcoin cores fuzzing efforts (2022)

Paper

Abstract: Bitcoin is one of the most prominent distributed software systems in the world, and a key part of a potentially revolutionary new form of financial tool, cryptocurrency. At heart, Bitcoin exists as a set of nodes running an implementation of the Bitcoin protocol. This paper describes an effort to investigate and enhance the effectiveness of the Bitcoin Core implementation fuzzing effort. The effort initially began as a query about how to escape saturation in the fuzzing effort, but developed into a more general exploration once it was determined that saturation was largely illusory, a byproduct of the (then) fuzzing configuration. This paper reports the process and outcomes of the two-week focused effort that emerged from that initial contact between Chaincode Labs and academic researchers. That effort found no smoking guns indicating major test/fuzz weaknesses. However, it produced a large number of additional fuzz corpus entries to add to the Bitcoin QA assets, clarified some long-standing problems in OSS-Fuzz triage, increased the set of documented fuzzers used in Bitcoin Core testing, and ran the first (to our knowledge) mutation analysis of Bitcoin Core's fuzz targets, revealing opportunities for further improvement. We contrast the Bitcoin Core transaction verification testing with that for other popular cryptocurrencies. This paper provides an overview of the challenges involved in improving testing infrastructure, processes, and documentation for a highly visible open source target system, from both the state-of-the-art research perspective and the practical engineering perspective. One major conclusion is that for well-designed fuzzing efforts, improvements to the oracle side of testing, increasing invariant checks and assertions, may be the best route to getting more out of fuzzing.

Machine Learning Guided Cross-Contract Fuzzing (2021)

Paper

Abstract: Smart contract transactions are increasingly interleaved by cross-contract calls. While many tools have been developed to identify a common set of vulnerabilities to guard smart contracts, the cross-contract vulnerability is however overlooked by existing tools. Cross-contract vulnerabilities are exploitable bugs that manifest in the presence of more than two interacting contracts. Existing methods are however limited to analyze a maximum of two contracts at the same time. Detecting cross-contract vulnerabilities is highly non-trivial. With multiple interacting contracts, the search space is much larger than that of a single contract. To address this problem, we present xFuzz, a machine learning guided smart contract fuzzing framework. The machine learning models are trained with novel features (e.g., word vectors and instructions) and are used to filter likely benign program paths. Comparing with existing static tools, machine learning model is proven to be more robust, avoiding directly adopting manually-defined rules in specific tools. We compare xFuzz with three state-of-the-art tools on 7,391 contracts. xFuzz detects 18 exploitable cross-contract vulnerabilities, of which 15 vulnerabilities are exposed for the first time. Furthermore, our approach is shown to be efficient in detecting non-cross-contract vulnerabilities as well-using less than 20% time as that of other fuzzing tools, xFuzz detects twice as many vulnerabilities.

SMARTIAN : Enhancing Smart Contract Fuzzing with Static and Dynamic Data-Flow Analyses (ASE 2021)

Paper

Abstract: Unlike traditional software, smart contracts have the unique organization in which a sequence of transactions shares persistent states. Unfortunately, such a characteristic makes it difficult for existing fuzzers to find out critical transaction sequences. To tackle this challenge, we employ both static and dynamic analyses for fuzzing smart contracts. First, we statically analyze smart contract bytecodes to predict which transaction sequences will lead to effective testing, and figure out if there is a certain constraint that each transaction should satisfy. Such information is then passed to the fuzzing phase and used to construct an initial seed corpus. During a fuzzing campaign, we perform a lightweight dynamic data-flow analysis to collect data-flow-based feedback to effectively guide fuzzing. We implement our ideas on a practical open-source fuzzer, named SMARTIAN. SMARTIAN can discover bugs in real-world smart contracts without the need for the source code. Our experimental results show that SMARTIAN is more effective than existing state-of-the-art tools in finding known CVEs from real-world contracts. SMARTIAN also outperforms other tools in terms of code coverage.

HFContractFuzzer: Fuzzing Hyperledger Fabric Smart Contracts for Vulnerability Detection (EASE 2021)

Paper

Abstract: With its unique advantages such as decentralization and immutability, blockchain technology has been widely used in various fields in recent years. The smart contract running on the blockchain is also playing an increasingly important role in decentralized application scenarios. Therefore, the automatic detection of security vulnerabilities in smart contracts has become an urgent problem in the application of blockchain technology. Hyperledger Fabric is a smart contract platform based on enterprise-level licensed distributed ledger technology. However, the research on the vulnerability detection technology of Hyperledger Fabric smart contracts is still in its infancy. In this paper, we propose HFContractFuzzer, a method based on Fuzzing technology to detect Hyperledger Fabric smart contracts, which combines a Fuzzing tool for golang named go-fuzz and smart contracts written by golang. We use HFContractFuzzer to detect vulnerabilities in five contracts from typical sources and discover that four of them have security vulnerabilities, proving the effectiveness of the proposed method.

sFuzz: An Efficient Adaptive Fuzzer for Solidity Smart Contracts (ICSE 2020)

Abstract: Smart contracts are Turing-complete programs that execute on the infrastructure of the blockchain, which often manage valuable digital assets. Solidity is one of the most popular programming languages for writing smart contracts on the Ethereum platform.Like traditional programs, smart contracts may contain vulnerabilities. Unlike traditional programs, smart contracts cannot be easily patched once they are deployed. It is thus important that smart contracts are tested thoroughly before deployment. In this work, we present an adaptive fuzzer for smart contracts on the Ethereum platform called sFuzz. Compared to existing Solidity fuzzers, sFuzz combines the strategy in the AFL fuzzer and an efficient lightweight multi-objective adaptive strategy targeting those hard-to-cover branches. sFuzz has been applied to more than 4 thousand smart contracts and the experimental results show that (1) sFuzz is efficient, e.g., two order of magnitudes faster than state-of-the-art tools; (2) sFuzz is effective in achieving high code coverage and discovering vulnerabilities; and (3) the different fuzzing strategies in sFuzz complement each other.

Targeted Greybox Fuzzing with Static Lookahead Analysis (ICSE 2020)

Paper

Abstract: Automatic test generation typically aims to generate inputs that explore new paths in the program under test in order to find bugs. Existing work has, therefore, focused on guiding the exploration toward program parts that are more likely to contain bugs by using an offline static analysis.

In this paper, we introduce a novel technique for targeted greybox fuzzing using an online static analysis that guides the fuzzer toward a set of target locations, for instance, located in recently modified parts of the program. This is achieved by first semantically analyzing each program path that is explored by an input in the fuzzer's test suite. The results of this analysis are then used to control the fuzzer's specialized power schedule, which determines how often to fuzz inputs from the test suite. We implemented our technique by extending a state-of-the-art, industrial fuzzer for Ethereum smart contracts and evaluate its effectiveness on 27 real-world benchmarks. Using an online analysis is particularly suitable for the domain of smart contracts since it does not require any code instrumentation - adding instrumentation to contracts changes their semantics. Our experiments show that targeted fuzzing significantly outperforms standard greybox fuzzing for reaching 83% of the challenging target locations (up to 14x of median speed-up).

Learning to Fuzz from Symbolic Execution with Application to Smart Contracts (CCS 2019)

Abstract: Fuzzing and symbolic execution are two complementary techniques for discovering software vulnerabilities. Fuzzing is fast and scalable, but can be ineffective when it fails to randomly select the right inputs. Symbolic execution is thorough but slow and often does not scale to deep program paths with complex path conditions.

In this work, we propose to learn an effective and fast fuzzer from symbolic execution, by phrasing the learning task in the framework of imitation learning. During learning, a symbolic execution expert generates a large number of quality inputs improving coverage on thousands of programs. Then, a fuzzing policy, represented with a suitable architecture of neural networks, is trained on the generated dataset. The learned policy can then be used to fuzz new programs.

We instantiate our approach to the problem of fuzzing smart contracts, a domain where contracts often implement similar functionality (facilitating learning) and security is of utmost importance. We present an end-to-end system, ILF (for Imitation Learning based Fuzzer), and an extensive evaluation over >18K contracts. Our results show that ILF is effective: (i) it is fast, generating 148 transactions per second, (ii) it outperforms existing fuzzers (e.g., achieving 33% more coverage), and (iii) it detects more vulnerabilities than existing fuzzing and symbolic execution tools for Ethereum.

ContractFuzzer: Fuzzing Smart Contracts for Vulnerability Detection (ASE 2018)

Abstract: Decentralized cryptocurrencies feature the use of blockchain to transfer values among peers on networks without central agency. Smart contracts are programs running on top of the blockchain consensus protocol to enable people make agreements while minimizing trusts. Millions of smart contracts have been deployed in various decentralized applications. The security vulnerabilities within those smart contracts pose significant threats to their applications. Indeed, many critical security vulnerabilities within smart contracts on Ethereum platform have caused huge financial losses to their users. In this work, we present ContractFuzzer, a novel fuzzer to test Ethereum smart contracts for security vulnerabilities. ContractFuzzer generates fuzzing inputs based on the ABI specifications of smart contracts, defines test oracles to detect security vulnerabilities, instruments the EVM to log smart contracts runtime behaviors, and analyzes these logs to report security vulnerabilities. Our fuzzing of 6991 smart contracts has flagged more than 459 vulnerabilities with high precision. In particular, our fuzzing tool successfully detects the vulnerability of the DAO contract that leads to USD 60 million loss and the vulnerabilities of Parity Wallet that have led to the loss of USD 30 million and the freezing of USD 150 million worth of Ether.

Constraint Solving

JIGSAW: Efficient and Scalable Path Constraints Fuzzing (S&P 2022)

Paper
Code

Abstract: Coverage-guided testing has shown to be an effective way to find bugs. If we model coverage-guided testing as a search problem (i.e., finding inputs that can cover more branches), then its efficiency mainly depends on two factors: (1) the accuracy of the searching algorithm and (2) the number of inputs that can be evaluated per unit time. Therefore, improving the search throughput has shown to be an effective way to improve the performance of coverage-guided testing.

In this work, we present a novel design to improve the search throughput: by evaluating newly generated inputs with JIT-compiled path constraints. This approach allows us to significantly improve the single thread throughput as well as scaling to multiple cores. We also developed several optimization techniques to eliminate major bottlenecks during this process. Evaluation of our prototype JIGSAW shows that our approach can achieve three orders of magnitude higher search throughput than existing fuzzers and can scale to multiple cores. We also find that with such high throughput, a simple gradient-guided search heuristic can solve path constraints collected from a large set of real-world programs faster than SMT solvers with much more sophisticated search heuristics. Evaluation of end-to-end coverage-guided testing also shows that our JIGSAW-powered hybrid fuzzer can outperform state-of-the-art testing tools.

Fuzzing Symbolic Expressions (ICSE 2021)

Paper
Code

Abstract: Recent years have witnessed a wide array of results in software testing, exploring different approaches and methodologies ranging from fuzzers to symbolic engines, with a full spectrum of instances in between such as concolic execution and hybrid fuzzing. A key ingredient of many of these tools is Satisfiability Modulo Theories (SMT) solvers, which are used to reason over symbolic expressions collected during the analysis. In this paper, we investigate whether techniques borrowed from the fuzzing domain can be applied to check whether symbolic formulas are satisfiable in the context of concolic and hybrid fuzzing engines, providing a viable alternative to classic SMT solving techniques. We devise a new approximate solver, \fuzzysat, and show that it is both competitive with and complementary to state-of-the-art solvers such as Z3 with respect to handling queries generated by hybrid fuzzers.

Just Fuzz It: Solving Floating-Point Constraints Using Coverage-guided Fuzzing (FSE 2019)

Paper

Abstract: We investigate the use of coverage-guided fuzzing as a means of proving satisfiability of SMT formulas over finite variable domains, with specific application to floating-point constraints.We showhow an SMT formula can be encoded as a program containing a location that is reachable if and only if the program's input corresponds to a satisfying assignment to the formula. A coverage-guided fuzzer can then be used to search for an input that reaches the location, yielding a satisfying assignment. We have implemented this idea in a tool, Just Fuzz-it Solver (JFS), and we present a large experimental evaluation showing that JFS is both competitive with and complementary to state-of-the-art SMT solvers with respect to solving floating-point constraints, and that the coverage-guided approach of JFS provides significant benefit over naive fuzzing in the floating-point domain. Applied in a portfolio manner, the JFS approach thus has the potential to complement traditional SMT solvers for program analysis tasks that involve reasoning about floating-point constraints.

Side-Channel Detection

QFuzz: Quantitative Fuzzing for Side Channels (ISSTA 2021)

Paper
Video

Abstract: Side channels pose a significant threat to the confidentiality of software systems. Such vulnerabilities are challenging to detect and evaluate because they arise from non-functional properties of software such as execution times and require reasoning on multiple execution traces. Recently, noninterference notions have been adapted in static analysis, symbolic execution, and greybox fuzzing techniques. However, noninterference is a strict notion and may reject security even if the strength of information leaks are weak. A quantitative notion of security allows for the relaxation of noninterference and tolerates small (unavoidable) leaks. Despite progress in recent years, the existing quantitative approaches have scalability limitations in practice.

In this work, we present QFuzz, a greybox fuzzing technique to quantitatively evaluate the strength of side channels with a focus on min entropy. Min entropy is a measure based on the number of distinguishable observations (partitions) to assess the resulting threat from an attacker who tries to compromise secrets in one try. We develop a novel greybox fuzzing equipped with two partitioning algorithms that try to maximize the number of distinguishable observations and the cost differences between them.

We evaluate QFuzz on a large set of benchmarks from existing work and real-world libraries (with a total of 70 subjects). QFuzz compares favorably to three state-of-the-art detection techniques. QFuzz provides quantitative information about leaks beyond the capabilities of all three techniques. Crucially, we compare QFuzz to a state-of-the-art quantification tool and find that QFuzz significantly outperforms the tool in scalability while maintaining similar precision. Overall, we find that our approach scales well for real-world applications and provides useful information to evaluate resulting threats. Additionally, QFuzz identifies a zero-day side-channel vulnerability in a security critical Java library that has since been confirmed and fixed by the developers.

JVM Fuzzing for JIT-Induced Side-Channel Detection (ICSE 2020)

Paper

Abstract: Timing side channels arise in software when a program's execution time can be correlated with security-sensitive program input. Recent results on software side-channel detection focus on analysis of program's source code. However, runtime behavior, in particular optimizations introduced during just-in-time (JIT) compilation, can impact or even introduce timing side channels in programs. In this paper, we present a technique for automatically detecting such JIT-induced timing side channels in Java programs. We first introduce patterns to detect partitions of secret input potentially separable by side channels. Then we present an automated approach for exploring behaviors of the Java Virtual Machine (JVM) to identify states where timing channels separating these partitions arise. We evaluate our technique on three datasets used in recent work on side-channel detection. We find that many code variants labeled ``safe'' with respect to side-channel vulnerabilities are in fact vulnerable to JIT-induced timing side channels. Our results directly contradict the conclusions of four separate state-of-the-art program analysis tools for side-channel detection and demonstrate that JIT-induced side channels are prevalent and can be detected automatically

ct-fuzz: Fuzzing for Timing Leaks (ICST 2020)

Paper

Abstract: Testing-based methodologies like fuzzing are able to analyze complex software which is not amenable to traditional formal approaches like verification, model checking, and abstract interpretation. Despite enormous success at exposing countless security vulnerabilities in many popular software projects, applications of testing-based approaches have mainly targeted checking traditional safety properties like memory safety. While unquestionably important, this class of properties does not precisely characterize other important security aspects such as information leakage, e.g., through side channels. In this work we extend testing-based software analysis methodologies to two-safety properties, which enables the precise discovery of information leaks in complex software. In particular, we present the ct-fuzz tool, which lends coverage-guided greybox fuzzers the ability to detect two-safety property violations. Our approach is capable of exposing violations to any two-safety property expressed as equality between two program traces. Empirically, we demonstrate that ct-fuzz swiftly reveals timing leaks in popular cryptographic implementations.

Concurrency Fuzzing

Context-Sensitive and Directional Concurrency Fuzzing for Data-Race Detection (NDSS 2022)

Paper
Video

Abstract: Fuzzing is popular for bug detection and vulnerability discovery nowadays. To adopt fuzzing for concurrency problems like data races, several recent concurrency fuzzing approaches consider concurrency information of program execution, and explore thread interleavings by affecting threads scheduling at runtime. However, these approaches are still limited in data-race detection. On the one hand, they fail to consider the execution contexts of thread interleavings, which can miss real data races in specific runtime contexts. On the other hand, they perform random thread-interleaving exploration, which frequently repeats already covered thread interleavings and misses many infrequent thread interleavings.

In this paper, we develop a novel concurrency fuzzing framework named CONZZER, to effectively explore thread interleavings and detect hard-to-find data races. The core of CONZZER is a context-sensitive and directional concurrency fuzzing approach for thread-interleaving exploration, with two new techniques. First, to ensure context sensitivity, we propose a new concurrencycoverage metric, concurrent call pair, to describe thread interleavings with runtime calling contexts. Second, to directionally explore thread interleavings, we propose an adjacency-directed mutation to generate new possible thread interleavings with already covered thread interleavings and then use a breakpoint-control method to attempt to actually cover them at runtime. With these two techniques, this concurrency fuzzing approach can effectively cover infrequent thread interleavings with concrete context information, to help discover hard-to-find data races. We have evaluated CONZZER on 8 user-level applications and 4 kernel-level filesystems, and found 95 real data races. We identify 75 of these data races to be harmful and send them to related developers, and 44 have been confirmed. We also compare CONZZER to existing fuzzing tools, and CONZZER continuously explores more thread interleavings and finds many real data races missed by these tools.

Fuzzing with automatically controlled interleavings to detect concurrency bugs (The Journal of Systems & Software 2022)

Paper

Concurrency vulnerabilities are an irresistible threat to security, and detecting them is challenging. Triggering the concurrency vulnerabilities requires a specific thread interleaving and a bug-inducing input. Existing methods have focused on one of these but not both or have experimented with small programs, which raises scalability issues.

This paper introduces AutoInter-fuzzing, a fuzzer controlling thread interleavings elaborately and providing an interleaving-aware power schedule to detect vulnerabilities in a multi-threaded program. AutoInter-fuzzing consists of static analysis and dynamic fuzzing. At the static analysis, the fuzzer extracts and optimizes the interleaving space to be explored and adds instrumentation to control thread interleavings. We apply the power schedule in the dynamic fuzzing to focus on the seeds that reveal the new interleaving space. The fuzzer records the interleaving information in a log when a crash occurs and uses it to reproduce and validate the crash.

Experiments with 13 real-world multi-threaded programs show that the interleaving-aware power schedule effectively enlarges the untested interleaving space, and AutoInter-fuzzing outperforms AFL and ConAFL in detecting interleaving-relevant vulnerabilities. AutoInter-fuzzing has detected six interleaving-relevant vulnerabilities, including two new vulnerabilities and four interleaving-irrelevant vulnerabilities.

Controlled Concurrency Testing via Periodical Scheduling (ICSE 2022)

Paper
Code

Abstract: Controlled Concurrency testing (CCT) techniques have been shown promising for concurrency bugs detection. They often have a mechanism to control the order in which threads get executed, and attempt to explore the space of possible interleavings of a concurrent program to detect bugs. Various CCT techniques have been proposed, but they are generally ineffective and ad-hoc. In this paper, we introduce a novel CCT technique PERIOD. Unlike previous works, PERIOD models the execution of concurrent programs as periodical execution, and systematically explores the space of possible interleavings, where the exploration is guided by periodical scheduling and influenced by previously tested interleavings. We have evaluated PERIOD on 10 real-world CVEs and 36 widely-used benchmark programs. Our experimental results show that PERIOD demonstrates superiority over other CCT techniques in both effectiveness and performance overhead. Moreover, we have discovered 5 previously unknown concurrency bugs in real-world programs.

Fuzzing Channel-Based Concurrency Runtimes using Types and Effects Slides (OOPSLA 2021)

Paper
Video

Abstract: Modern programming languages support concurrent programming based on channels and processes. Channels enable synchronous and asynchronous message-passing between independent light-weight processes making it easy to express common concurrency patterns. The implementation of channels and processes in compilers and language runtimes is a difficult task that relies heavily on traditional and error-prone low-level concurrency primitives, raising concerns about correctness and reliability. In this paper, we present an automatic program generation technique to test such programming language implementations. We define a type and effect system for programs that communicate over channels and where every execution is guaranteed to eventually terminate. We can generate and run such programs, and if a program fails to terminate, we have found a bug in the programming language implementation. We implement such an automatic program generator and apply it to Go, Kotlin, Crystal, and Flix. We find two new bugs in Flix, and reproduce two bugs; one in Crystal and one in Kotlin.

MUZZ: Thread-aware Grey-box Fuzzing for Effective Bug Hunting in Multithreaded Programs (USENIX Security2020)

Abstract: Grey-box fuzz testing has revealed thousands of vulnerabilities in real-world software owing to its lightweight instrumentation, fast coverage feedback, and dynamic adjusting strategies. However, directly applying grey-box fuzzing to input-dependent multithreaded programs can be extremely inefficient. In practice, multithreading-relevant bugs are usually buried in the sophisticated program flows. Meanwhile, existing grey-box fuzzing techniques do not stress thread-interleavings that affect execution states in multithreaded programs. Therefore, mainstream grey-box fuzzers cannot adequately test problematic segments in multithreaded software, although they might obtain high code coverage statistics.

To this end, we propose Muzz, a new grey-box fuzzing technique that hunts for bugs in multithreaded programs. Muzz owns three novel thread-aware instrumentations, namely coverage-oriented instrumentation, thread-context instrumentation, and schedule-intervention instrumentation. During fuzzing, these instrumentations engender runtime feedback to accentuate execution states caused by thread interleavings. By leveraging such feedback in the dynamic seed selection and execution strategies, Muzz preserves more valuable seeds that expose bugs under a multithreading context.

We evaluate Muzz on twelve real-world multithreaded programs. Experiments show that Muzz outperforms AFL in both multithreading-relevant seed generation and concurrency-vulnerability detection. Further, by replaying the target programs against the generated seeds, Muzz also reveals more concurrency-bugs (e.g., data-races, thread-leaks) than AFL. In total, Muzz detected eight new concurrency-vulnerabilities and nineteen new concurrency-bugs. At the time of writing, four reported issues have received CVE IDs.

KRace: Data Race Fuzzing for Kernel File Systems (S&P 2020)

Abstract: Data races occur when two threads fail to use proper synchronization when accessing shared data. In kernel file systems, which are highly concurrent by design, data races are common mistakes and often wreak havoc on the users, causing inconsistent states or data losses. Prior fuzzing practices on file systems have been effective in uncovering hundreds of bugs, but they mostly focus on the sequential aspect of file system execution and do not comprehensively explore the concurrency dimension and hence, forgo the opportunity to catch data races.

In this paper, we bring coverage-guided fuzzing to the concurrency dimension with three new constructs: 1) a new coverage tracking metric, alias coverage, specially designed to capture the exploration progress in the concurrency dimension; 2) an evolution algorithm for generating, mutating, and merging multithreaded syscall sequences as inputs for concurrency fuzzing; and 3) a comprehensive lockset and happens-before modeling for kernel synchronization primitives for precise data race detection. These components are integrated into KRACE, an end-to-end fuzzing framework that has discovered 23 data races in ext4, btrfs, and the VFS layer so far, and 9 are confirmed to be harmful.

ConFuzz: A Concurrency Fuzzer (2019)

Paper

Abstract: Concurrency bugs are as equally vulnerable as the bugs found in the single-threaded programs and these bugs can be exploited using concurrency attacks. Unfortunately, there is not much literature available in detecting various kinds of concurrency issues in a multi-threaded program due to its complexity and uncertainty. In this paper, we aim at detecting concurrency bugs by using directed evolutionary fuzzing with the help of static analysis of the source code. Concurrency bug detection involves two main entities: an input and a particular thread execution order. The evolutionary part of fuzzing will prefer inputs that involve memory access patterns across threads (data flow interleaving) and thread ordering that disturb the data dependence more and direct them to trigger concurrency bugs. This paper suggests the idea of a concurrency fuzzer, which is first of its kind. We use a combination of LLVM, Thread Sanitizer and fuzzing techniques to detect various concurrency issues in an application. The source code of the application is statically analyzed for various paths, from the different thread related function calls to the main function. Every basic block in these paths are assigned a unique ID and a weight based on the distance of the basic block from the thread function calls. These basic blocks are instrumented to print their ID and weight upon execution. The knowledge about the basic blocks in the sliced paths are used to generate new sets of inputs from the old ones, thus covering even more basic blocks in the path and thereby increasing the chances of hitting a concurrency warning. We use Thread Sanitizer present in the LLVM compiler infrastructure to detect the concurrency bug warnings while executing each input. The inputs are directed to discover even new address locations with possible concurrency issues. The system was tested on three simple multi-threaded applications pigz, pbzip2, and pixz. The results show a quicker detection of unique addresses in the application with possible concurrency issues.

Fuzz Testing of Multithreaded Applications Based on Waiting (MICSECS 2019)

Paper

Abstract: In our days, it is hard to imagine big software products written without the use of multithreading. However, they not only use multithreading, but are also complicated by being distributed. On one hand, it gives performance advantages, but on the other hand, it becomes much more difficult to find bugs and test such applications. When developing programs that use multithreading, we can find the following types of errors: priority inversions, deadlock, livelock, ABA problem, and others. Such errors can lead to large financial losses, for example, in the banking infrastructure, or losses of human lives in aircraft engineering, civil engineering, medical devices and other areas. Special tools such as Valgrind, Google TSAN and others are used to find such bugs. Until recently, such tools were not able to fuzzing testing multithreaded applications, but now Google TSAN has a special module. The main limitation of the testing fuzzing module is that it is not able to handle waiting on nonatomic variables. The results presented in this paper allow us to carry fuzzing testing of threads and at the same time correctly handle the situation with waiting on variables that are not atomic, as well as examples on which the improved algorithm successfully copes with handling such waiting.

A Heuristic Framework to Detect Concurrency Vulnerabilities (ACSAC 2018)

Abstract: With a growing demand of concurrent software to exploit multi-core hardware capability, concurrency vulnerabilities have become an inevitable threat to the security of today's IT industry. Existing concurrent program detection schemes focus mainly on detecting concurrency errors such as data races, atomicity violation, etc., with little attention paid to detect concurrency vulnerabilities that may be exploited to infringe security. In this paper, we propose a heuristic framework that combines both static analysis and fuzz testing to detect targeted concurrency vulnerabilities such as concurrency buffer overflow, double free, and use-after-free. The static analysis locates sensitive concurrent operations in a concurrent program, categorizes each finding into a potential type of concurrency vulnerability, and determines the execution order of the sensitive operations in each finding that would trigger the suspected concurrency vulnerability. The results are then plugged into the fuzzer with the execution order fixed by the static analysis in order to trigger the suspected concurrency vulnerabilities.

In order to introduce more variance which increases possibility that the concurrency errors can be triggered, we also propose manipulation of thread scheduling priority to enable a fuzzer such as AFL to effectively explore thread interleavings in testing a concurrent program. To the best of our knowledge, this is the first fuzzer that is capable of effectively exploring concurrency errors.

In evaluating the proposed heuristic framework with a benchmark suit of six real-world concurrent C programs, the framework detected two concurrency vulnerabilities for the proposed concurrency vulnerability detection, both being confirmed to be true positives, and produced three new crashes for the proposed interleaving exploring fuzzer that existing fuzzers could not produce. These results demonstrate the power and effectiveness of the proposed heuristic framework in detecting concurrency errors and vulnerabilities.

API Testing/Fuzzing

Assessing the Impact of Interface Vulnerabilities in Compartmentalized Software (NDSS 2023)

Paper

Abstract: Least-privilege separation decomposes applications into compartments limited to accessing only what they need. When compartmentalizing existing software, many approaches neglect securing the new inter-compartment interfaces, although what used to be a function call from/to a trusted component is now potentially a targeted attack from a malicious compartment. This results in an entire class of security bugs: Compartment Interface Vulnerabilities (CIVs).

This paper provides an in-depth study of CIVs. We taxonomize these issues and show that they affect all known compartmentalization approaches. We propose ConfFuzz, an in-memory fuzzer specialized to detect CIVs at possible compartment boundaries. We apply ConfFuzz to a set of 25 popular applications and 36 possible compartment APIs, to uncover a wide data-set of 629 vulnerabilities. We systematically study these issues, and extract numerous insights on the prevalence of CIVs, their causes, impact, and the complexity to address them. We stress the critical importance of CIVs in compartmentalization approaches, demonstrating an attack to extract isolated keys in OpenSSL and uncovering a decade-old vulnerability in sudo. We show, among others, that not all interfaces are affected in the same way, that API size is uncorrelated with CIV prevalence, and that addressing interface vulnerabilities goes beyond writing simple checks. We conclude the paper with guidelines for CIV-aware compartment interface design, and appeal for more research towards systematic CIV detection and mitigation.

Minerva: Browser API Fuzzing with Dynamic Mod-Ref Analysis (FSE 2022)

Paper

Abstract: Browser APIs are essential to the modern web experience. Due to their large number and complexity, they vastly expand the attack surface of browsers. To detect vulnerabilities in these APIs, fuzzers generate test cases with a large amount of random API invocations. However, the massive search space formed by arbitrary API combinations hinders their effectiveness: since randomly-picked API invocations unlikely interfere with each other (i.e., compute on partially shared data), few interesting API interactions are explored. Consequently, reducing the search space by revealing inter-API relations is a major challenge in browser fuzzing.

We propose Minerva, an efficient browser fuzzer for browser API bug detection. The key idea is to leverage API interference relations to reduce redundancy and improve coverage. Minerva consists of two modules: dynamic mod-ref analysis and guided code generation. Before fuzzing starts, the dynamic mod-ref analysis module builds an API interference graph. It first automatically identifies individual browser APIs from the browser's code base. Next, it instruments the browser to dynamically collect mod-ref relations between APIs. During fuzzing, the guided code generation module synthesizes highly-relevant API invocations guided by the mod-ref relations. We evaluate Minerva on three mainstream browsers, i.e. Safari, FireFox, and Chromium. Compared to state-of-the-art fuzzers, Minerva improves edge coverage by $19.63%$ to $229.62%$ and finds 2x to 3x more unique bugs. Besides, Minerva has discovered 35 previously-unknown bugs out of which 20 have been fixed with 5 CVEs assigned and acknowledged by browser vendors.

Combinatorial Testing of RESTful APIs (ICSE 2022)

Paper

Abstract: This paper presents RestCT, a systematic and fully automatic approach that adopts Combinatorial Testing (CT) to test RESTful APIs. RestCT is systematic in that it covers and tests not only the interactions of a certain number of operations in RESTful APIs, but also the interactions of particular input-parameters in every single operation. This is realised by a novel two-phase test case generation approach, which first generates a constrained sequence covering array to determine the execution orders of available operations, and then applies an adaptive strategy to generate and refine several constrained covering arrays to concretise input-parameters of each operation. RestCT is also automatic in that its application relies on only a given Swagger specification of RESTful APIs. The creation of CT test models (especially, the inferring of dependency relationships in operation and input-parameter levels), and the generation and execution of test cases are performed without any human intervention. Experimental results on 11 real-world RESTful APIs demonstrate the effectiveness and efficiency of RestCT. In particular, RestCT can find eight new bugs, where only one of them can be triggered by the state-of-the-art testing tool of RESTful APIs.

GraphFuzz: Library API Fuzzing with Lifetime-aware Dataflow Graphs (ICSE 2022)

Paper

Abstract: We present the design and implementation of GraphFuzz, a new structure-, coverage- and object lifetime-aware fuzzer capable of automatically testing low-level Library APIs. Unlike other fuzzers, GraphFuzz models sequences of executed functions as a dataflow graph, thus enabling it to perform graph-based mutations both at the data and at the execution trace level. GraphFuzz comes with an automated specification generator to minimize the developer integration effort.

We use GraphFuzz to analyze Skia - the rigorously tested Google Chrome graphics library - and benchmark GraphFuzz-generated fuzzing harnesses against hand-optimized, painstakingly written libFuzzer harnesses. We find that GraphFuzz generates test cases that achieve 2-3x more code coverage on average with minimal development effort, and also uncovered previous unknown defects in the process. We demonstrate GraphFuzz's applicability on low-level APIs by analyzing four additional open-source libraries and finding dozens of previously unknown defects. All security relevant findings have already been reported and fixed by the developers.

Last, we open-source GraphFuzz under a permissible license and provide code to reproduce all results in this paper.

MOREST: Model-based RESTful API Testing with Execution Feedback (ICSE 2022)

Paper

Abstract: RESTful APIs are arguably the most popular endpoints for accessing Web services. Blackbox testing is one of the emerging techniques for ensuring the reliability of RESTful APIs. The major challenge in testing RESTful APIs is the need for correct sequences of API operation calls for in-depth testing. To build meaningful operation call sequences, researchers have proposed techniques to learn and utilize the API dependencies based on OpenAPI specifications. However, these techniques either lack the overall awareness of how all the APIs are connected or the flexibility of adaptively fixing the learned knowledge.

In this paper, we propose MOREST, a model-based RESTful API testing technique that builds and maintains a dynamically updating RESTful-service Property Graph (RPG) to model the behaviors of RESTful-services and guide the call sequence generation. We empirically evaluated MOREST and the results demonstrate that \textsc{Morest} can successfully request an average of 152.66%-232.45% more API operations, cover 26.16%-103.24% more lines of code, and detect 40.64%-215.94% more bugs than state-of-the-art techniques.

In total, we applied MOREST to 6 real-world projects and found 44 bugs (13 of them cannot be detected by existing approaches). Specifically, 2 of the confirmed bugs are from Bitbucket, a famous code management service with more than 6 million users.

Automated Testing of Software that Uses Machine Learning APIs (ICSE 2022)

Paper

Abstract: An increasing number of software applications incorporate machine learning (ML) solutions for cognitive tasks that statistically mimic human behaviors. To test such software, tremendous human effort is needed to design image/text/audio inputs that are relevant to the software, and to judge whether the software is processing these inputs as most human beings do. Even when misbehavior is exposed, it is often unclear whether the culprit is inside the cognitive ML API or the code using the API.

This paper presents Keeper, a new testing tool for software that uses cognitive ML APIs. Keeper designs a pseudo-inverse function for each ML API that reverses the corresponding cognitive task in an empirical way (e.g., an image search engine pseudo-reverses the image-classification API), and incorporates these pseudo-inverse functions into a symbolic execution engine to automatically generate relevant image/text/audio inputs and judge output correctness. Once misbehavior is exposed, Keeper attempts to change how ML APIs are used in software to alleviate the misbehavior. Our evaluation on a variety of open-source applications shows that Keeper greatly improves the branch coverage, while identifying many previously unknown bugs. Our user study with Python developers and application users further demonstrates the efficacy of Keeper.

Nessie: Automatically Testing JavaScript APIs with Asynchronous Callbacks (ICSE 2022)

Paper

Abstract: Previous algorithms for feedback-directed unit test generation iteratively create sequences of API calls by executing partial tests and by adding new API calls at the end of the test. These algorithms are challenged by a popular class of APIs: higher-order functions that receive callback arguments, which often are invoked asynchronously. Existing test generators cannot effectively test such APIs because they only sequence API calls, but do not nest one call into the callback function of another. This paper presents Nessie, the first feedback-directed unit test generator that supports nesting of API calls and that tests asynchronous callbacks. Nesting API calls enables a test to use values produced by an API that are available only once a callback has been invoked, and is often necessary to ensure that methods are invoked in a specific order. The core contributions of our approach are a tree-based representation of unit tests with callbacks and a novel algorithm to iteratively generate such tests in a feedback-directed manner. We evaluate our approach on ten popular JavaScript libraries with both asynchronous and synchronous callbacks. The results show that, in a comparison with LambdaTester, a state of the art test generation technique that only considers sequencing of method calls, Nessie finds more behavioral differences and achieves slightly higher coverage. Notably, Nessie needs to generate significantly fewer tests to achieve and exceed the coverage achieved by the state of the art.

REST API Fuzzing by Coverage Level Guided Blackbox Testing (2021)

Code
Paper

Abstract: With the growth of web applications, REST APIs have become the primary communication method between services. In order to ensure system reliability and security, software quality can be assured by effective testing methods. Black box fuzz testing is one of the effective methods to perform tests on a large scale. However, conventional black box fuzz testing generates random data without judging the quality of the input. We implement a black box fuzz testing method for REST APIs. It resolves the issues of blind mutations without knowing the effectiveness by Test Coverage Level feedback. We also enhance the mutation strategies by reducing the testing complexity for REST APIs, generating more appropriate test cases to cover possible paths. We evaluate our method by testing two large open-source projects and 89 bugs are reported and confirmed. In addition, we find 351 bugs from 64 remote API services in APIs.guru.

FuzzGen: Automatic Fuzzer Generation (USENIX Security2020)

Abstract: Fuzzing is a testing technique to discover unknown vulnerabilities in software. When applying fuzzing to libraries, the core idea of supplying random input remains unchanged, yet it is non-trivial to achieve good code coverage. Libraries cannot run as standalone programs, but instead are invoked through another application. Triggering code deep in a library remains challenging as specific sequences of API calls are required to build up the necessary state. Libraries are diverse and have unique interfaces that require unique fuzzers, so far written by a human analyst.

To address this issue, we present FuzzGen, a tool for automatically synthesizing fuzzers for complex libraries in a given environment. FuzzGen leverages a whole system analysis to infer the library's interface and synthesizes fuzzers specifically for that library. FuzzGen requires no human interaction and can be applied to a wide range of libraries. Furthermore, the generated fuzzers leverage LibFuzzer to achieve better code coverage and expose bugs that reside deep in the library.

FuzzGen was evaluated on Debian and the Android Open Source Project (AOSP) selecting 7 libraries to generate fuzzers. So far, we have found 17 previously unpatched vulnerabilities with 6 assigned CVEs. The generated fuzzers achieve an average of 54.94% code coverage; an improvement of 6.94% when compared to manually written fuzzers, demonstrating the effectiveness and generality of FuzzGen.

Intelligent REST API Data Fuzzing (FSE 2020)

Abstract: The cloud runs on REST APIs. In this paper, we study how to intelligently generate data payloads embedded in REST API requests in order to find data-processing bugs in cloud services. We discuss how to leverage REST API specifications, which, by definition, contain data schemas for API request bodies. We then propose and evaluate a range of data fuzzing techniques, including structural schema fuzzing rules, various rule combinations, search heuristics, extracting data values from examples included in REST API specifications, and learning data values on-the-fly from previous service responses. After evaluating these techniques, we identify the top-performing combination and use this algorithm to fuzz several Microsoft Azure cloud services. During our experiments, we found 100s of 'Internal Server Error'? service crashes, which we triaged into 17 unique bugs and reported to Azure developers. All these bugs are reproducible, confirmed, and fixed or in the process of being fixed.

FUDGE: Fuzz Driver Generation at Scale (FSE 2019)

Paper

Abstract: At Google we have found tens of thousands of security and robustness bugs by fuzzing C and C++ libraries. To fuzz a library, a fuzzer requires a fuzz driver - which exercises some library code - to which it can pass inputs. Unfortunately, writing fuzz drivers remains a primarily manual exercise, a major hindrance to the widespread adoption of fuzzing. In this paper, we address this major hindrance by introducing the Fudge system for automated fuzz driver generation. Fudge automatically generates fuzz driver candidates for libraries based on existing client code. We have used Fudge to generate thousands of new drivers for a wide variety of libraries. Each generated driver includes a synthesized C/C++ program and a corresponding build script, and is automatically analyzed for quality. Developers have integrated over 200 of these generated drivers into continuous fuzzing services and have committed to address reported security bugs. Further, several of these fuzz drivers have been upstreamed to open source projects and integrated into the OSS-Fuzz fuzzing infrastructure. Running these fuzz drivers has resulted in over 150 bug fixes, including the elimination of numerous exploitable security vulnerabilities.

REST-ler: Stateful REST API Fuzzing (ICSE 2019)

Abstract: This paper introduces REST-ler, the first stateful REST API fuzzer. REST-ler analyzes the API specification of a cloud service and generates sequences of requests that automatically test the service through its API. REST-ler generates test sequences by (1) inferring producer-consumer dependencies among request types declared in the specification (eg inferring that 'a request B should be executed after request A' because B takes as an input a resource-id x produced by A) and by (2) analyzing dynamic feedback from responses observed during prior test executions in order to generate new tests (eg learning that 'a request C after a request sequence A;B is refused by the service' and therefore avoiding this combination in the future). We present experimental results showing that these two techniques are necessary to thoroughly exercise a service under test while pruning the large search space of possible request sequences. We used REST-ler to test GitLab, a large open-source self-hosted Git service, as well as several Microsoft Azure and Office365 cloud services. REST-ler found 28 bugs in Gitlab and several bugs in each of the Azure and Office365 cloud services tested so far. These bugs have been confirmed by the service owners, and are either in the process of being fixed or have already been fixed.

Fuzzing JavaScript Environment APIs with Interdependent Function Calls (IFM 2019)

Paper

Abstract: The prevalence of the JavaScript programming language makes the correctness and security of its execution environments highly important. The most exposed and vulnerable parts of these environments are the APIs published to the executed untrusted JavaScript programs. This paper revisits the fuzzing technique that generates JavaScript environment API calls using random walks on so-called prototype graphs to uncover potentially security-related failures. We show the limits of generating independent call expressions, the approach of prior work, and give an extension to enable the generation of interdependent API calls that re-use each other's results. We demonstrate with an experiment that this enhancement allows our approach to exercise JavaScript environment APIs in ways that were not possible with the previous approach, and that it can also trigger more issues in a real target.

Systematic Fuzzing and Testing of TLS Libraries (CCS 2016)

Paper

Abstract: We present TLS-Attacker, an open source framework for evaluating the security of TLS libraries. TLS-Attacker allows security engineers to create custom TLS message flows and arbitrarily modify message contents using a simple interface in order to test the behavior of their libraries. Based on TLS-Attacker, we present a two-stage fuzzing approach to evaluate TLS server behavior. Our approach automatically searches for cryptographic failures and boundary violation vulnerabilities. It allowed us to find unusual padding oracle vulnerabilities and overflows/overreads in widely used TLS libraries, including OpenSSL, Botan, and MatrixSSL.

Our findings motivate developers to create comprehensive test suites, including positive as well as negative tests, for the evaluation of TLS libraries. We use TLS-Attacker to create such a test suite framework which finds further problems in Botan.

Hardware Fuzzing

CPU Fuzzing for Discovering Hardware-caused Information Leakage (2022)

Video
Slides

Abstract: In the last years, many microarchitectural attacks, such as cache attacks, Spectre, or Meltdown, have been discovered in CPUs. However, discovering microarchitectural attacks is still a tedious manual process in many cases. Hence, CPUs are still full of vulnerabilities and side channels.

This talk presents our research on the automated discovery of microarchitectural attack vectors. We discuss how we developed Osiris, a fuzzing-based framework that discovers timing side channels in the CPU microarchitecture. In addition to side channels, we present Transynther, an automated approach for finding Meltdown-type transient-execution attacks and generating effective exploitation code for these attacks. Starting from a model of the Meltdown mechanism, Transynther uses a fuzzing-based approach for finding new variants based on known Meltdown-type attacks. With this approach, we did not only re-discover new ways to exploit known attacks but also discovered a new ZombieLoad variant named "Medusa". Our tools are available as open-source software and can be used to check any x86 CPU for potentially new vulnerabilities. We present different case studies based on our findings to demonstrate that the discovered vulnerabilities are indeed relevant.

Fuzzing Hardware Like Software (USENIX SEC 2022)

Paper

Abstract: Hardware flaws are permanent and potent: hardware cannot be patched once fabricated, and any flaws may undermine even formally verified software executing on top. Consequently, verification time dominates implementation time. The gold standard in hardware Design Verification (DV) is concentrated at two extremes: random dynamic verification and formal verification. Both techniques struggle to root out the subtle flaws in complex hardware that often manifest as security vulnerabilities. The root problem with random verification is its undirected nature, making it inefficient, while formal verification is constrained by the state-space explosion problem, making it infeasible to apply to complex designs. What is needed is a solution that is directed, yet under-constrained.

Instead of making incremental improvements to existing hardware verification approaches, we leverage the observation that existing software fuzzers already provide such a solution; we adapt it for hardware verification, thus leveraging existing - more advanced - software verification tools. Specifically, we translate RTL hardware to a software model and fuzz that model. The central challenge we address is how best to mitigate the differences between the hardware execution model and software execution model. This includes: 1) how to represent test cases, 2) what is the hardware equivalent of a crash, 3) what is an appropriate coverage metric, and 4) how to create a general-purpose fuzzing harness for hardware.

To evaluate our approach, we design, implement, and opensource a Hardware Fuzzing Pipeline that enables fuzzing hardware at scale, using only open-source tools. Using our pipeline, we fuzz four IP blocks from Google's OpenTitan Root-of-Trust chip. Our experiments reveal a two orders-of-magnitude reduction in run time to achieve Finite State Machine (FSM) coverage over traditional dynamic verification schemes. Moreover, with our design-agnostic harness, we achieve over 88% HDL line coverage in three out of four of our designs - even without any initial seeds.

PeriScope: An Effective Probing and Fuzzing Framework for the Hardware-OS Boundary (NDSS2019)

Paper

Abstract: The OS kernel is an attractive target for remote attackers. If compromised, the kernel gives adversaries full system access, including the ability to install rootkits, extract sensitive information, and perform other malicious actions, all while evading detection. Most of the kernel's attack surface is situated along the system call boundary. Ongoing kernel protection efforts have focused primarily on securing this boundary; several capable analysis and fuzzing frameworks have been developed for this purpose. However, there are additional paths to kernel compromise that do not involve system calls, as demonstrated by several recent exploits. For example, by compromising the firmware of a peripheral device such as a Wi-Fi chipset and subsequently sending malicious inputs from the Wi-Fi chipset to the Wi-Fi driver, adversaries have been able to gain control over the kernel without invoking a single system call. Unfortunately, there are currently no practical probing and fuzzing frameworks that can help developers find and fix such vulnerabilities occurring along the hardware-OS boundary. We present PeriScope, a Linux kernel based probing framework that enables fine-grained analysis of device-driver interactions. PeriScope hooks into the kernel's page fault handling mechanism to either passively monitor and log traffic between device drivers and their corresponding hardware, or mutate the data stream on-the-fly using a fuzzing component, PeriFuzz, thus mimicking an active adversarial attack. PeriFuzz accurately models the capabilities of an attacker on peripheral devices, to expose different classes of bugs including, but not limited to, memory corruption bugs and double-fetch bugs. To demonstrate the risk that peripheral devices pose, as well as the value of our framework, we have evaluated PeriFuzz on the Wi-Fi drivers of two popular chipset vendors, where we discovered 15 unique vulnerabilities, 9 of which were previously unknown.

Other Interesting Fuzzing

LOKI: State-Aware Fuzzing Framework for the Implementation of Blockchain Consensus Protocols (NDSS 2023)

Paper

Abstract: Blockchain consensus protocols are responsible for coordinating the nodes to make agreements on the transaction results. Their implementation bugs, including memory-related and consensus logic vulnerabilities, may pose serious threats. Fuzzing is a promising technique for protocol vulnerability detection. However, existing fuzzers cannot deal with complex consensus states of distributed nodes, thus generating a large number of useless packets, inhibiting their effectiveness in reaching the deep logic of consensus protocols.

In this work, we propose LOKI, a blockchain consensus protocol fuzzing framework that detects the consensus memory-related and logic bugs. LOKI senses consensus states in real-time by masquerading as a node. First, LOKI dynamically builds a state model that records the state transition of each node. After that, LOKI adaptively generates the input targets, types, and contents according to the state model. With a bug analyzer, LOKI detects the consensus protocol implementation bugs with well-defined oracles. We implemented and evaluated LOKI on four widely used commercial blockchain systems, including Go-Ethereum, Facebook Diem, IBM Fabric, and WeBank FISCO-BCOS. LOKI has detected 20 serious previously unknown vulnerabilities with 9 CVEs assigned. 14 of them are memory-related bugs, and 6 are consensus logic bugs. Compared with state-of-the-art tools such as Peach, Fluffy, and Twins, LOKI improves the branch coverage by an average of 43.21%, 182.05%, and 291.58%.

DriveFuzz: Discovering Autonomous Driving Bugs through Driving Quality-Guided Fuzzing (CCS 2022)

Abstract: Autonomous driving has become real; semi-autonomous driving vehicles in an affordable price range are already on the streets, and major automotive vendors are actively developing full self-driving systems to deploy them in this decade. Before rolling the products out to the end-users, it is critical to test and ensure the safety of the autonomous driving systems, consisting of multiple layers intertwined in a complicated way. However, while safety-critical bugs may exist in any layer and even across layers, relatively little attention has been given to testing the entire driving system across all the layers. Prior work mainly focuses on white-box testing of individual layers and preventing attacks on each layer.

In this paper, we aim at holistic testing of autonomous driving systems that have a whole stack of layers integrated in their entirety. Instead of looking into the individual layers, we focus on the vehicle states that the system continuously changes in the driving environment. This allows us to design DriveFuzz, a new systematic fuzzing framework that can uncover potential vulnerabilities regardless of their locations. DriveFuzz automatically generates and mutates driving scenarios based on diverse factors leveraging a high-fidelity driving simulator. We build novel driving test oracles based on the real-world traffic rules to detect safety-critical misbehaviors, and guide the fuzzer towards such misbehaviors through driving quality metrics referring to the physical states of the vehicle.

DriveFuzz has discovered 30 new bugs in various layers of two autonomous driving systems (Autoware and CARLA Behavior Agent) and three additional bugs in the CARLA simulator. We further analyze the impact of these bugs and how an adversary may exploit them as security vulnerabilities to cause critical accidents in the real world.

LibAFL: A Framework to Build Modular and Reusable Fuzzers (CCS 2022)

Paper

Abstract: The release of AFL marked an important milestone in the area of software security testing, revitalizing fuzzing as a major research topic and spurring a large number of research studies that attempted to improve and evaluate the different aspects of the fuzzing pipeline.

Many of these studies implemented their techniques by forking the AFL codebase. While this choice might seem appropriate at first, combining multiple forks into a single fuzzer requires a high engineering overhead, which hinders progress in the area and prevents fair and objective evaluations of different techniques. The highly fragmented landscape of the fuzzing ecosystem also prevents researchers from combining orthogonal techniques and makes it difficult for end users to adopt new prototype solutions.

To tackle this problem, in this paper we propose LibAFL, a framework to build modular and reusable fuzzers. We discuss the different components generally used in fuzzing and map them to an extensible framework. LibAFL allows researchers and engineers to extend the core fuzzer pipeline and share their new components for further evaluations. As part of LibAFL, we integrated techniques from more than 20 previous works and conduct extensive experiments to show the benefit of our framework to combine and evaluate different approaches. We hope this can help to shed light on current advancements in fuzzing and provide a solid base for comparative and extensible research in the future.

RoboFuzz: Fuzzing Robotic Systems over Robot Operating System (ROS) for Finding Correctness Bugs (FSE 2022)

Paper

Abstract: Robotic systems are becoming an integral part of human lives. Responding to the increased demands for robot productions, Robot Operating System (ROS), an open-source middleware suite for robotic development, is gaining traction by providing practical tools and libraries for quickly developing robots. In this paper, we are concerned with a relatively less-tested class of bugs in ROS and ROS-based robotic systems, called semantic correctness bugs, including the violation of specification, violation of physical laws, and cyber-physical discrepancy. These bugs often stem from the cyber-physical nature of robotic systems, in which noisy hardware components are intertwined with software components, and thus cannot be detected by existing fuzzing approaches that mostly focus on finding memory-safety bugs.

We propose RoboFuzz, a feedback-driven fuzzing framework that integrates with ROS and enables testing of the correctness bugs. RoboFuzz features (1) data type-aware mutation for effectively stressing data-driven ROS systems, (2) hybrid execution for acquiring robotic states from both real-world and a simulator, capturing unforeseen cyber-physical discrepancies, (3) an oracle handler that identifies correctness bugs by checking the execution states against predefined correctness oracles, and (4) a semantic feedback engine for providing augmented guidance to the input mutator, complementing classic code coverage-based feedback, which is less effective for distributed, data-driven robots. By encoding the correctness invariants of ROS and four ROS-compatible robotic systems into specialized oracles, RoboFuzz detected 30 previously unknown bugs, of which 25 are acknowledged and six have been fixed.

Almost Correct Invariants: Synthesizing Inductive Invariants by Fuzzing Proofs (ISSTA 2022)

Paper

Abstract: Real-life programs contain multiple operations whose semantics are unavailable to verification engines, like third-party library calls, inline assembly and SIMD instructions, special compiler-provided primitives, and queries to uninterpretable machine learning models. Even with the exceptional success story of program verification, synthesis of inductive invariants for such open" programs has remained a challenge. Currently, this problem is handled by manuallyclosing" the program by providing hand-written \textit{stubs} that attempt to capture the behavior of the unmodelled operations; writing stubs is not only difficult and tedious, but the stubs are often incorrect raising serious questions on the whole endeavor.

In this work, we propose \textit{Almost Correct Invariants} as an automated strategy for synthesizing inductive invariants for such ``open" programs. We adopt an active learning strategy where a data-driven learner proposes candidate invariants. In deviation from prior work that attempt to \textit{verify} invariants, we attempt to \textit{falsify} the invariants: we reduce the falsification problem to a set of reachability checks on non-deterministic programs; we ride on the success of modern fuzzers to answer these reachability queries. Our tool, $\textit{ACHAR}$, automatically synthesizes inductive invariants that are sufficient to prove the correctness of the target programs. We compare $\textit{ACHAR}$ with a state-of-the-art invariant synthesis tool that employs theorem proving on formulae built over the program source. Though $\textit{ACHAR}$ is without strong soundness guarantees, our experiments show that even when we provide almost no access to the program source, $\textit{ACHAR}$ outperforms the state-of-the-art invariant generator that has complete access to the source. We also evaluate $\textit{ACHAR}$ on programs that current invariant synthesis engines cannot handle programs that invoke external library calls, inline assembly, and queries to convolution neural networks; $\textit{ACHAR}$ successfully infers the necessary inductive invariants within a reasonable time.

��AFL: Non-intrusive Feedback-driven Fuzzing for Microcontroller Firmware (ICSE 2022)

Paper

Abstract: Fuzzing is one of the most effective approaches to find software flaws. However, applying it to microcontroller firmware incurs many challenges. For example, rehosting-based solutions cannot accurately model peripheral behaviors and thus cannot be used to fuzz the corresponding driver code. In this work, we present μAFL, a hardware-assisted approach to fuzzing firmware on real devices. It leverages debugging tools in existing embedded system development to construct an AFL-compatible fuzzing framework. Specifically, we use the debug dongle to bridge the fuzzing environment on the PC and the target firmware on the microcontroller device. To collect code coverage information without costly code instrumentation, μAFL relies on the ARM ETM hardware debugging feature, which transparently collects the instruction trace and streams the results to the PC. However, the raw ETM data is obscure and needs enormous computing resources to recover the actual instruction flow. We therefore propose an alternative representation of code coverage, which retains the same path sensitivity as the original AFL algorithm, but can directly work on the raw ETM data without matching them with disassembled instructions. To further reduce the workload, we use the DWT hardware feature to selectively collect runtime information of interest. We evaluated μAFL on two real evaluation boards from two major vendors: NXP and STMicroelectronics. With our prototype, we discovered ten 0-day bugs in the driver code shipped with the SDK of STMicroelectronics and three 0-day bugs in the SDK of NXP. Considering the wide adoption of chip-vendor-provided SDK code in real products, our results are alarming

CorbFuzz: Checking Browser Security Policies with Fuzzing (ASE 2021)

Paper

Abstract: Browsers use security policies to block malicious behaviors. Cross-Origin Read Blocking (CORB) is a browser security policy for preventing side-channel attacks such as Spectre. We propose a web browser security policy fuzzer called CorbFuzz for checking CORB and similar policies. In implementing a security policy, the browser only has access to HTTP requests and responses, and takes policy actions based solely on those interactions. In checking the browser security policies, CorbFuzz uses a policy oracle that tracks the web application behavior and infers the desired policy action based on the web application state. By comparing the policy oracle with the browser behavior, CorbFuzz detects weaknesses in browser security policies. CorbFuzz checks the web browser policy by fuzzing a set of web applications where the state-related queries are symbolically evaluated for increased coverage and automation. CorbFuzz collects type information from database queries and branch conditions in order to prevent the generation of inconsistent data values during fuzzing. We evaluated CorbFuzz on CORB implementations of Chromium and Webkit, and Opaque Response Blocking (ORB) policy implementation of Firefox using web applications collected from GitHub. We found three classes of weaknesses in Chromium's implementation of CORB.

Fully Automated Functional Fuzzing of Android Apps for Detecting Non-crashing Logic Bugs (OOPSLA 2021)

Paper

Abstract: Android apps are GUI-based event-driven software and have become ubiquitous in recent years. Obviously, functional correctness is critical for an app's success. However, in addition to crash bugs, non-crashing functional bugs (in short as 'non-crashing bugs' in this work) like inadvertent function failures, silent user data lost and incorrect display information are prevalent, even in popular, well-tested apps. These non-crashing functional bugs are usually caused by program logic errors and manifest themselves on the graphic user interfaces (GUIs). In practice, such bugs pose significant challenges in effectively detecting them because (1) current practices heavily rely on expensive, small-scale manual validation (the lack of automation); and (2) modern fully automated testing has been limited to crash bugs (the lack of test oracles).

This paper fills this gap by introducing independent view fuzzing, a novel, fully automated approach for detecting non-crashing functional bugs in Android apps. Inspired by metamorphic testing, our key insight is to leverage the commonly-held independent view property of Android apps to manufacture property-preserving mutant tests from a set of seed tests that validate certain app properties. The mutated tests help exercise the tested apps under additional, adverse conditions. Any property violations indicate likely functional bugs for further manual confirmation. We have realized our approach as an automated, end-to-end functional fuzzing tool, Genie. Given an app, (1) Genie automatically detects non-crashing bugs without requiring human-provided tests and oracles (thus fully automated); and (2) the detected non-crashing bugs are diverse (thus general and not limited to specific functional properties), which set Genie apart from prior work.

We have evaluated Genie on 12 real-world Android apps and successfully uncovered 34 previously unknown non-crashing bugs in their latest releases - all have been confirmed, and 22 have already been fixed. Most of the detected bugs are nontrivial and have escaped developer (and user) testing for at least one year and affected many app releases, thus clearly demonstrating Genie's effectiveness. According to our analysis, Genie achieves a reasonable true positive rate of 40.9%, while these 34 non-crashing bugs could not be detected by prior fully automated GUI testing tools (as our evaluation confirms). Thus, our work complements and enhances existing manual testing and fully automated testing for crash bugs.

webFuzz: Grey-Box Fuzzing for Web Applications (ESORICS 2021)

Paper

Abstract: Fuzzing is significantly evolved in analysing native code, but web applications, invariably, have received limited attention until now. This paper designs, implements and evaluates webFuzz, a gray-box fuzzing prototype for discovering vulnerabilities in web applications.

webFuzz is successful in leveraging instrumentation for detecting cross-site scripting (XSS) vulnerabilities, as well as covering more code faster than black-box fuzzers. In particular, webFuzz has discovered one zero-day vulnerability in WordPress, a leading CMS platform, and five in an online commerce application named CE-Phoenix.

Moreover, in order to systematically evaluate webFuzz, and similar tools, we provide the first attempt for automatically synthesizing reflective cross-site scripting (RXSS) vulnerabilities in vanilla web applications.

Vulnerability-Oriented Fuzz Testing for Connected Autonomous Vehicle Systems ( IEEE Transactions on Reliability 2021)

Paper

Abstract: In an era of connectivity and automation, the vehicle industry is adopting numerous technologies to transform driver-centric vehicles into intelligent mechanical devices driven by software components. Software integration and network connectivity inherit numerous security issues that open the door for malicious attacks. Software security testing is a scalable and practical approach to identify systems' weaknesses and vulnerabilities at an early stage and throughout their life-cycle. Security specialists recommend fuzz testing to identify vulnerabilities within vehicle software systems. Nevertheless, the randomness and blindness of fuzzing hinder it from becoming a reliable security tool. This article presents a vulnerability-oriented fuzz (VulFuzz) testing framework that utilizes security vulnerability metrics designed particularly for connected and autonomous vehicles to direct and prioritize the fuzz testing toward the most vulnerable components. While most gray-box fuzzing techniques aim solely to expand code coverage, the proposed approach assigns weights to ensure a thorough examination of the most vulnerable components. Moreover, we employ an input structure-aware mutation technique that can bypass vehicle software systems' input formats to boost test performance and avoid dropped test cases. Such a testing technique will contribute to the quality assurance of vehicle software engineering. We implemented the proposed approach on OpenPilot, a driver assistance system, and compared our results to American fuzzy lop (AFL) and an unguided mutation-based fuzzer. Within 16.8 h, VulFuzz exposed 335 crashes, 41 times more than AFL and two times more than an unguided mutation-based fuzzer. VulFuzz is explicitly efficient for automotive systems, reaching the same code coverage as AFL but with more exposed crashes and fewer dropped messages.

Hardware Support to Improve Fuzzing Performance and Precision (CCS 2021)

Paper

Abstract: Coverage-guided fuzzing is considered one of the most efficient bug-finding techniques, given its number of bugs reported. However, coverage tracing provided by existing software-based approaches, such as source instrumentation and dynamic binary translation, can incur large overhead. Hindered by the significantly lowered execution speed, it also becomes less beneficial to improve coverage feedback by incorporating additional execution states. In this paper, we propose SNAP, a customized hardware platform that implements hardware primitives to enhance the performance and precision of coverage-guided fuzzing. By sitting at the bottom of the computer stack, SNAP leverages the existing CPU pipeline and micro-architectural features to provide coverage tracing and rich execution semantics with near-zero cost regardless of source code availability. Prototyped as a synthesized RISC-V BOOM processor on FPGA, SNAP incurs a barely 3.1% tracing overhead on the SPEC benchmarks while achieving a 228× higher fuzzing throughput than the existing software-based solution. Posing only a 4.8% area and 6.5% power overhead, SNAP is highly practical and can be adopted by existing CPU architectures with minimal changes.

Fuzzing: Cyberphysical System Testing for Security and Dependability (Computer 2021)

Paper

Abstract: Fuzz testing enables automated remote testing of cyberphysical and Internet of Things systems. The selection of effective test vector sequences is critical and a natural domain for the employment of artificial intelligence and machine learning techniques that incorporate successful selection criteria.

ESRFuzzer: an enhanced fuzzing framework for physical SOHO router devices to discover multi-Type vulnerabilities (Cybersecurity 2021)

Paper

Abstract: SOHO (small office/home office) routers provide services for end devices to connect to the Internet, playing an important role in cyberspace. Unfortunately, security vulnerabilities pervasively exist in these routers, especially in the web server modules, greatly endangering end users. To discover these vulnerabilities, fuzzing web server modules of SOHO routers is the most popular solution. However, its effectiveness is limited due to the lack of input specification, lack of routers' internal running states, and lack of testing environment recovery mechanisms. Moreover, existing works for device fuzzing are more likely to detect memory corruption vulnerabilities.In this paper, we propose a solution ESRFuzzer to address these issues. It is a fully automated fuzzing framework for testing physical SOHO devices. It continuously and effectively generates test cases by leveraging two input semantic models, i.e., KEY-VALUE data model and CONF-READ communication model, and automatically recovers the testing environment with power management. It also coordinates diversified mutation rules with multiple monitoring mechanisms to trigger multi-type vulnerabilities. With the guidance of the two semantic models, ESRFuzzer can work in two ways: general mode fuzzing and D-CONF mode fuzzing. General mode fuzzing can discover both issues which occur in the CONF and READ operation, while D-CONF mode fuzzing focus on the READ-op issues especially missed by general mode fuzzing.We ran ESRFuzzer on 10 popular routers across five vendors. In total, it discovered 136 unique issues, 120 of which have been confirmed as 0-day vulnerabilities we found. As an improvement of SRFuzzer, ESRFuzzer have discovered 35 previous undiscovered READ-op issues that belong to three vulnerability types, and 23 of them have been confirmed as 0-day vulnerabilities by vendors. The experimental results show that ESRFuzzer outperforms state-of-the-art solutions in terms of types and number of vulnerabilities found.

Estimating Residual Risk in Greybox Fuzzing (FSE 2021)

Paper
Code

Abstract: For any errorless fuzzing campaign, no matter how long, there is always some residual risk that a software error would be discovered if only the campaign was run for just a bit longer. Recently, greybox fuzzing tools have found widespread adoption. Yet, practitioners can only guess when the residual risk of a greybox fuzzing campaign falls below a specific, maximum allowable threshold.

In this paper, we explain why residual risk cannot be directly estimated for greybox campaigns, argue that the discovery probability (i.e., the probability that the next generated input increases code coverage) provides an excellent upper bound, and explore sound statistical methods to estimate the discovery probability in an ongoing greybox campaign. We find that estimators for blackbox fuzzing systematically and substantially \emph{under}-estimate the true risk. An engineer - who stops the campaign when the estimators purport a risk below the maximum allowable risk - is vastly misled. She might need execute a campaign that is orders of magnitude longer to achieve the allowable risk. Hence, the \emph{key challenge} we address in this paper is \emph{adaptive bias}: The probability to discover a specific error actually increases over time. We provide the first probabilistic analysis of adaptive bias, and introduce two novel classes of estimators that tackle adaptive bias. With our estimators, the engineer can decide with confidence when to abort the campaign.

HeteroFuzz: Fuzz Testing to Detect Platform Dependent Divergence for Heterogeneous Applications (FSE 2021)

Paper
Code

Abstract: As specialized hardware accelerators like FPGAs become a prominent part of the current computing landscape, software applications are increasingly constructed to leverage heterogeneous architectures. Such a trend is already happening in the domain of machine learning and Internet-of-Things (IoT) systems built on edge devices. Yet, debugging and testing methods for heterogeneous applications are currently lacking. These applications may look similar to regular C/C++ code but include hardware synthesis details in terms of preprocessor directives. Therefore, their behavior under heterogeneous architectures may diverge significantly from CPU due to hardware synthesis details. Further, the compilation and hardware simulation cycle takes an enormous amount of time, prohibiting frequent invocations required for fuzz testing.

We propose a novel fuzz testing technique, called HeteroFuzz, designed to specifically target heterogeneous applications and to detect platform-dependent divergence. The key essence of HeteroFuzz is that it uses a three-pronged approach to reduce the long latency of repetitively invoking a hardware simulator on a heterogeneous application. First, in addition to monitoring code coverage as a fuzzing guidance mechanism, we analyze synthesis pragmas in kernel code and monitor accelerator-relevant value spectra. Second, we design dynamic probabilistic mutations to increase the chance of hitting divergent behavior under different platforms. Third, we memorize the boundaries of seen kernel inputs and skip HLS simulator invocation if it can expose only redundant divergent behavior. We evaluate HeteroFuzz on seven real-world heterogeneous applications with FPGA kernels. HeteroFuzz is 754X faster in exposing the same set of distinct divergence symptoms than naive fuzzing. Probabilistic mutations contribute to 17.5X speed up than the one without. Selective invocation of HLS simulation contributes to 8.8X speed up than the one without.

DirectFuzz: Automated Test Generation for RTL Designs using Directed Graybox Fuzzing (DAC 2021)

Paper

Abstract: A critical challenge in RTL verification is to generate effective test inputs. Recently, RFUZZ proposed to use an automated software testing technique, namely Graybox Fuzzing, to effectively generate test inputs to maximize the coverage of the whole hardware design. For a scenario where a tiny fraction of a large hardware design needs to be tested, the RFUZZ approach is extremely time consuming. In this work, we present DirectFuzz, a directed test generation mechanism. DirectFuzz uses Directed Graybox Fuzzing to generate test inputs targeted towards a module instance, which enables targeted testing. Our experimental results show that DirectFuzz covers the target sites up to 17.5× faster (2.23×on average) than RFUZZ on a variety of RTL designs.

PMFuzz: Test Case Generation for Persistent Memory Programs (ASPLOS 2021)

Paper
Code

Abstract: This work provides PMFuzz, a fuzzer that automatically generates high-value test cases to detect crash consistency bugs in PM programs. The source of PMFuzz is publicly available at: https://github.com/Systems-ShiftLab/PMFuzz. We implement PMFuzz on top of a well-known fuzzer, AFL++ by incorporating our key insights, and evaluate PMFuzz in a real PM system with common PM programs, including keyvalue stores and databases.

PGFUZZ: Policy-Guided Fuzzing for Robotic Vehicles (NDSS 2021)

Paper

Abstract: Robotic vehicles (RVs) are becoming essential tools of modern systems, including autonomous delivery services, public transportation, and environment monitoring. Despite their diverse deployment, safety and security issues with RVs limit their wide adoption. Most attempts to date in RV security aim to propose defenses that harden their control program against syntactic bugs, input validation bugs, and external sensor spoofing attacks. In this paper, we introduce PGFUZZ, a policy-guided fuzzing framework, which validates whether an RV adheres to identified safety and functional policies that cover user commands, configuration parameters, and physical states. PGFUZZ expresses desired policies through temporal logic formulas with time constraints as a guide to fuzz the analyzed system. Specifically, it generates fuzzing inputs that minimize a distance metric measuring how close'' the RV current state is to a policy violation. In addition, it uses static and dynamic analysis to focus the fuzzing effort only on those commands, parameters, and environmental factors that influence the truth value'' of any of the exercised policies. The combination of these two techniques allows PGFUZZ to increase the efficiency of the fuzzing process significantly. We validate PGFUZZ on three RV control programs, ArduPilot, PX4, and Paparazzi, with 56 unique policies. PGFUZZ discovered 156 previously unknown bugs, 106 of which have been acknowledged by their developers.

IntelliGen: Automatic Driver Synthesis for Fuzz Testing (ICSE 2021 SEIP)

Abstract: Fuzzing is a technique widely used in vulnerability detection. The process usually involves writing effective fuzz driver programs, which, when done manually, can be extremely labor intensive. Previous attempts at automation leave much to be desired, in either degree of automation or quality of output. In this paper, we propose IntelliGen, a framework that constructs valid fuzz drivers automatically. First, IntelliGen determines a set of entry functions and evaluates their respective chance of exhibiting a vulnerability. Then, IntelliGen generates fuzz drivers for the entry functions through hierarchical parameter replacement and type inference. We implemented IntelliGen and evaluated its effectiveness on real-world programs selected from the Android Open-Source Project, Google's fuzzer-testsuite and industrial collaborators. IntelliGen covered on average 1.08×-2.03× more basic blocks and 1.36×-2.06× more paths over state-of-the-art fuzz driver synthesizers FUDGE and FuzzGen. IntelliGen performed on par with manually written drivers and found 10 more bugs.

WINNIE: Fuzzing Windows Applications with Harness Synthesis and Fast Cloning (NDSS 2021)

Paper

Abstract: Fuzzing is an emerging technique to automatically validate programs and uncover bugs. It has been widely used to test many programs and has found thousands of security vulnerabilities. However, existing fuzzing efforts are mainly centered around Unix-like systems, as Windows imposes unique challenges for fuzzing: a closed-source ecosystem, the heavy use of graphical interfaces and the lack of fast process cloning machinery.

In this paper, we propose two solutions to address the challenges Windows fuzzing faces. Our system, WINNIE, first tries to synthesize a harness for the application, a simple program that directly invokes target functions, based on sample executions. It then tests the harness, instead of the original complicated program, using an efficient implementation of fork on Windows. Using these techniques, WINNIE can bypass irrelevant GUI code to test logic deep within the application. We used WINNIE to fuzz 59 closed-source Windows binaries, and it successfully generated valid fuzzing harnesses for all of them. In our evaluation, WINNIE can support 2.2× more programs than existing Windows fuzzers could, and identified 3.9× more program states and achieved 26.6× faster execution. In total, WINNIE found 61 unique bugs in 32 Windows binaries.

Industry Practice of Coverage-Guided Enterprise-Level DBMS Fuzzing (ICSE 2021)

Paper

Abstract: As an infrastructure for data persistence and analysis, Database Management Systems (DBMSs) are the cornerstones of modern enterprise software. To improve their correctness, the industry has been applying blackbox fuzzing for decades. Recently, the research community achieved impressive fuzzing gains using coverage guidance. However, due to the complexity and distributed nature of enterprise-level DBMSs, seldom are these researches applied to the industry.

In this paper, we apply coverage-guided fuzzing to enterprise-level DBMSs from Huawei and Bloomberg LP. In our practice of testing GaussDB and Comdb2, we found major challenges in all three testing stages. The challenges are collecting precise coverage, optimizing fuzzing performance, and analyzing root causes. In search of a general method to overcome these challenges, we propose RATEL, a coverage-guided fuzzer for enterprise-level DBMSs. With its industry-oriented design, RATEL improves the feedback precision, enhances the robustness of input generation, and performs an on-line investigation on the root cause of bugs. As a result, RATEL outperformed other fuzzers in terms of coverage and bugs. Compared to industrial black box fuzzers SQLsmith and SQLancer, as well as coverage-guided academic fuzzer Squirrel, RATEL covered 38.38%, 106.14%, 583.05% more basic blocks than the best results of other three fuzzers in GaussDB, PostgreSQL, and Comdb2, respectively. More importantly, RATEL has discovered 32, 42, and 5 unknown bugs in GaussDB, Comdb2, and PostgreSQL.

Refined Grey-Box Fuzzing with Sivo (arXiv 2021)

Paper

Abstract: We design and implement from scratch a new fuzzer called Sivo that refines multiple stages of grey-box fuzzing. First, Sivo refines dataflow fuzzing in two ways: (a) it provides a new taint inference engine that requires only logarithmic in the input size number of tests to infer the dependency of all program branches on the input bytes, and (b) it deploys a novel method for inverting branches by solving directly and efficiently systems of inequalities. Second, our fuzzer refines accurate tracking and detection of code coverage with simple and easily implementable methods. Finally, Sivo refines selection of parameters and strategies by parameterizing all stages of fuzzing and then dynamically selecting optimal values during fuzzing. Thus the fuzzer can easily adapt to a target program and rapidly increase coverage. We compare our fuzzer to 11 other state-of-the-art greybox fuzzers on 27 popular benchmarks. Our evaluation shows that Sivo scores the highest both in terms of code coverage and in terms of number of found vulnerabilities.

Fuzzing Technique in Web Applications and Beyond (MCTE 2020)

Paper

Abstract: Web security is an old yet important topic. With the emergence of new applications and new threats, we need new techniques to fight against these threats. Fuzzing technique has been proven as an effective tool and has the potential of evolving with the new threats. In this paper, we discuss the applications of fuzzing technique in both web and other areas. We also present a case study by implementing a fuzzer in this study.

Nyx: Greybox Hypervisor Fuzzing using Fast Snapshots and Affine Types (USENIX Security2021)

Paper
Code

Abstract: A hypervisor (also know as virtual machine monitor, VMM) enforces the security boundaries between different virtual machines (VMs) running on the same physical machine. A malicious user who is able to run her own kernel on a cloud VM can interact with a large variety of attack surfaces. Exploiting a software fault in any of these surfaces leads to full access to all other VMs that are co-located on the same host. Hence, the efficient detection of hypervisor vulnerabilities is crucial for the security of the modern cloud infrastructure. Recent work showed that blind fuzzing is the most efficient approach to identify security issues in hypervisors, mainly due to an outstandingly high test throughput.

In this paper we present the design and implementation of NYX, a highly optimized, coverage-guided hypervisor fuzzer. We show how a fast snapshot restoration mechanism that allows us to reload the system under test thousands of times per second is key to performance. Furthermore, we introduce a novel mutation engine based on custom bytecode programs, encoded as directed acyclic graphs (DAG), and affine types, that enables the required flexibility to express complex interactions. Our evaluation shows that, while NYX has a lower throughput than the state-of-the-art hypervisor fuzzer, it performs competitively on simple targets: NYX typically requires only a few minutes longer to achieve the same test coverage. On complex devices, however, our approach is able to significantly outperform existing works. Moreover, we are able to uncover substantially more bugs: in total, we uncovered 44 new bugs with 22 CVEs requested. Our results demonstrate that coverage guidance is highly valuable, even if a blind fuzzer can be significantly faster.

The Use of Likely Invariants as Feedback for Fuzzers (USENIX Security2021)

Paper

Abstract: While fuzz testing proved to be a very effective technique to find software bugs, open challenges still exist. One of the its main limitations is the fact that popular coverage-guided designs are optimized to reach different parts of the program under test, but struggle when reachability alone is insufficient to trigger a vulnerability. In reality, many bugs require a specific program state that involve not only the control flow, but also the values of some of the program variables. Unfortunately, alternative exploration strategies that have been proposed in the past to capture the program state are of little help in practice, as they immediately result in a state explosion. In this paper, we propose a new feedback mechanism that augments code coverage by taking into account the usual values and relationships among program variables. For this purpose, we learn likely invariants over variables at the basic�block level, and partition the program state space accordingly. Our feedback can distinguish when an input violates one or more invariants and reward it, thus refining the program state approximation that code coverage normally offers. We implemented our technique in a prototype called INVSCOV, developed on top of LLVM and AFL++. Our ex�periments show that our approach can find more, and different, bugs with respect to fuzzers that use a pure code-coverage feedback. Furthermore, they led to the discovery of two vulnerabilities in a library tested daily on OSS-Fuzz, and still present at the time in its latest version.

FREEDOM: Engineering a State-of-the-Art DOM Fuzzer (CCS 2020)

Paper
Code

Abstract: The DOM engine of a web browser is a popular attack surface and has been thoroughly fuzzed during its development. A common approach adopted by the latest DOM fuzzers is to generate new inputs based on context-free grammars. However, such a generative approach fails to capture the data dependencies in the inputs of a DOM engine, namely, HTML documents. Meanwhile, it is unclear whether or not coverage-guided mutation, which is well-known to be effective in fuzzing numerous software, still remains to be effective against DOM engines. Worse yet, existing DOM fuzzers cannot adopt a coverage-guided approach because they are unable to fully support HTML mutation and suffer from low browser throughput. To scientifically understand the effectiveness and limitations of the two approaches, we propose FreeDom, a full-fledged cluster-friendly DOM fuzzer that works with both generative and coverage-guided modes. FreeDom relies on a context-aware intermediate representation to describe HTML documents with proper data dependencies. FreeDom also exhibits up to 3.74x higher throughput through browser self-termination. FreeDom has found 24 previously unknown bugs in commodity browsers including Safari, Firefox, and Chrome, and 10 CVEs has been assigned so far. With the context-aware generation, FreeDom finds 3x more unique crashes in WebKit than the state-of-the-art DOM fuzzer, Domato. FreeDom guided by coverage is more effective in revealing new code blocks (2.62%) and finds three complex bugs that its generative approach fails to find. However, coverage-guided mutation that bootstraps with an empty corpus triggers 3.8x fewer unique crashes than the generative approach. The newly revealed coverage, more often than not, negatively affects the effectiveness of DOM fuzzers in bug finding. Therefore, we consider context-aware generation the best practice to find more DOM engine bugs and expect further improvement on coverage-guided DOM fuzzing facilitated by FreeDom.

SQUIRREL: Testing Database Management Systems with Language Validity and Coverage Feedback (CCS 2020)

Paper

Abstract: Fuzzing is an increasingly popular technique for verifying software functionalities and finding security vulnerabilities. However, current mutation-based fuzzers cannot effectively test database management systems (DBMSs), which strictly check inputs for valid syntax and semantics. Generation-based testing can guarantee the syntax correctness of the inputs, but it does not utilize any feedback, like code coverage, to guide the path exploration.

In this paper, we develop Squirrel, a novel fuzzing framework that considers both language validity and coverage feedback to test DBMSs. We design an intermediate representation (IR) to maintain SQL queries in a structural and informative manner. To generate syntactically correct queries, we perform type-based mutations on IR, including statement insertion, deletion and replacement. To mitigate semantic errors, we analyze each IR to identify the logical dependencies between arguments, and generate queries that satisfy these dependencies. We evaluated Squirrel on four popular DBMSs: SQLite, MySQL, PostgreSQL and MariaDB. Squirrel found 51 bugs in SQLite, 7 in MySQL and 5 in MariaDB. 52 of the bugs are fixed with 12 CVEs assigned. In our experiment, Squirrel achieves 2.4x-243.9x higher semantic correctness than state-of-the-art fuzzers, and explores 2.0x-10.9x more new edges than mutation-based tools. These results show that Squirrel is effective in finding memory errors of database management systems.

BigFuzz: Efficient Fuzz Testing for Data Analytics using Framework Abstraction (ASE 2020)

Paper

Abstract: As big data analytics become increasingly popular, data-intensive scalable computing (DISC) systems help address the scalability issue of handling large data. However, there exists a lack of automated testing techniques to test such data-centric applications, because data is often incomplete, continuously evolving, and hard to know a priori. Fuzz testing has been proven to be highly effective in other domains such as security; however, it is nontrivial to apply such traditional fuzzing to big data analytics directly for three reasons: (1) the long latency of DISC systems prohibits the applicability of fuzzing: naïve fuzzing would spend 98% of the time in setting up a test environment; (2) conventional branch coverage is unlikely to scale to DISC applications because most binary code comes from the framework implementation such as Apache Spark; and (3) random bit or byte-level mutations can hardly generate meaningful data, which fails to reveal real-world application bugs.

We propose a novel coverage-guided fuzz testing tool for big data analytics, called BigFuzz. The key essence of our approach is that: (a) we focus on exercising application logic as opposed to increasing framework code coverage by abstracting the DISC frame-work using specifications. BigFuzz performs automated source to source transformations to construct an equivalent DISC application suitable for fast test generation, and (b) we design schema-aware data mutation operators based on our in-depth study of DISC application error types. BigFuzz speeds up the fuzzing time by 78X-1477X compared to random fuzzing, improves application code coverage by 20%-271%, and achieves 33%-157% improvement in detecting application errors. When compared to the state of the art that uses symbolic execution to test big data analytics, BigFuzz is applicable to twice more programs and can find 80.6% more bugs.

MoFuzz: A Fuzzer Suite for Testing Model-Driven Software Engineering Tools (ASE 2020)

Paper
Code

Abstract: Fuzzing or fuzz testing is an established technique that aims to discover unexpected program behavior (e.g. bugs, security vulnerabilities, or crashes) by feeding automatically generated data into a program under test. However, the application of fuzzing to test Model-Driven Software Engineering (MDSE) tools is still limited because of the difficulty of existing fuzzers to provide structured, well-typed inputs, namely models that conform to typing and consistency constraints induced by a given meta-model and underlying modeling framework. By drawing from recent advances on both fuzz testing and automated model generation, we present three different approaches for fuzzing MDSE tools: A graph grammar-based fuzzer and two variants of a coverage-guided mutation-based fuzzer working with different sets of model mutation operators. We have evaluated our fuzzing approaches on a set of real-world MDSE tools. Our experimental results show that all approaches can outperform both standard fuzzers and model generators w.r.t.\ their fuzzing capabilities. Moreover, we found that each of our approaches comes with its own strengths and weaknesses in terms of fault finding capabilities and the ability to cover different aspects of the system under test. Thus the approaches complement each other, forming a fuzzer suite for testing MDSE tools.

AFL++: Combining Incremental Steps of Fuzzing Research (USENIX Woot2020)

Abstract: In this paper, we present AFL++, a community-driven opensource tool that incorporates state-of-the-art fuzzing research, to make the research comparable, reproducible, combinable, and - most importantly - useable. It offers a variety of novel features, for example, its Custom Mutator API, able to extend the fuzzing process at many stages. With it, mutators for specific targets can also be written by experienced security testers. We hope for AFL++ to become a new baseline tool not only for current, but also for future research, as it allows us to test new techniques quickly, and evaluate not only the effectiveness of the single technique versus the state-of-the art, but also in combination with other techniques. The paper gives an evaluation of hand-picked fuzzing technologies - shining light on the fact that while each novel fuzzing method can increase performance in some targets - it decreases performance for other targets. This is an insight future fuzzing research should consider in their evaluations.

Active Fuzzing for Testing and Securing Cyber-Physical Systems (ISSTA 2020)

Video
Paper

Abstract: Cyber-physical systems (CPSs) in critical infrastructure face a pervasive threat from attackers, motivating research into a variety of countermeasures for securing them. Assessing the effectiveness of these countermeasures is challenging, however, as realistic benchmarks of attacks are difficult to manually construct, blindly testing is ineffective due to the enormous search spaces and resource requirements, and intelligent fuzzing approaches require impractical amounts of data and network access. In this work, we propose active fuzzing, an automatic approach for finding test suites of packet-level CPS network attacks, targeting scenarios in which attackers can observe sensors and manipulate packets, but have no existing knowledge about the payload encodings. Our approach learns regression models for predicting sensor values that will result from sampled network packets, and uses these predictions to guide a search for payload manipulations (i.e. bit flips) most likely to drive the CPS into an unsafe state. Key to our solution is the use of online active learning, which iteratively updates the models by sampling payloads that are estimated to maximally improve them. We evaluate the efficacy of active fuzzing by implementing it for a water purification plant testbed, finding it can automatically discover a test suite of flow, pressure, and over/underflow attacks, all with substantially less time, data, and network access than the most comparable approach. Finally, we demonstrate that our prediction models can also be utilised as countermeasures themselves, implementing them as anomaly detectors and early warning systems.

CrFuzz: Fuzzing Multi-purpose Programs through Input Validation (FSE 2020)

Paper

Abstract: Fuzz testing has been proved its effectiveness in discovering software vulnerabilities. Empowered its randomness nature along with a coverage-guiding feature, fuzzing has been identified a vast number of vulnerabilities in real-world programs. This paper begins with an observation that the design of the current state-of-the-art fuzzers is not well suited for a particular (but yet important) set of software programs. Specifically, current fuzzers have limitations in fuzzing programs serving multiple purposes, where each purpose is controlled by extra options. This paper proposes CrFuzz, which overcomes this limitation. CrFuzz designs a clustering analysis to automatically predict if a newly given input would be accepted or not by a target program. Exploiting this prediction capability, CrFuzz is designed to efficiently explore the programs with multiple purposes. We employed CrFuzz for three state-of-the-art fuzzers, AFL, QSYM, and MOpt, and CrFuzz-augmented versions have shown 19.3% and 5.68% better path and edge coverage on average. More importantly, during two weeks of long-running experiments, CrFuzz discovered 277 previously unknown vulnerabilities where 212 of those are already confirmed and fixed by the respected vendors. We would like to emphasize that many of these vulnerabilities were discoverd from FFMpeg, ImageMagick, and Graphicsmagick, all of which are targets of Google's OSS-Fuzz project and thus heavily fuzzed for last three years by far. Nevertheless, CrFuzz identified a remarkable number of vulnerabilities, demonstrating its effectiveness of vulnerability finding capability.

SpecFuzz: Bringing Spectre-type vulnerabilities to the surface (USENIX Security2020)

Paper
Code

Abstract: SpecFuzz is the first tool that enables dynamic testing for speculative execution vulnerabilities (e.g., Spectre). The key is a novel concept of speculation exposure: The program is instrumented to simulate speculative execution in software by forcefully executing the code paths that could be triggered due to mispredictions, thereby making the speculative memory accesses visible to integrity checkers (e.g., AddressSanitizer). Combined with the conventional fuzzing techniques, speculation exposure enables more precise identification of potential vulnerabilities compared to state-of-the-art static analyzers. Our prototype for detecting Spectre V1 vulnerabilities successfully identifies all known variations of Spectre V1 and decreases the mitigation overheads across the evaluated applications, reducing the amount of instrumented branches by up to 77% given a sufficient test coverage.

USBFuzz: A Framework for Fuzzing USB Drivers by Device Emulation (USENIX Security2020)

Abstract: The Universal Serial Bus (USB) connects external devices to a host. This interface exposes the OS kernels and device drivers to attacks by malicious devices. Unfortunately, kernels and drivers were developed under a security model that implicitly trusts connected devices. Drivers expect faulty hardware but not malicious attacks. Similarly, security testing drivers are challenging as input must cross the hardware/software barrier. Fuzzing, the most widely used bug-finding technique, relies on providing random data to programs. However, fuzzing device drivers is challenging due to the difficulty in crossing the hardware/software barrier and providing random device data to the driver under test.

We present USBFuzz, a portable, flexible, and modular framework for fuzz testing USB drivers. At its core, USBFuzz uses a software-emulated USB device to provide random device data to drivers (when they perform IO operations). As the emulated USB device works at the device level, porting it to other platforms is straight-forward. Using the USBFuzz framework, we apply (i) coverage-guided fuzzing to a broad range of USB drivers in the Linux kernel; (ii) dumb fuzzing in FreeBSD, MacOS, and Windows through cross-pollination seeded by the Linux inputs; and (iii) focused fuzzing of a USB webcam driver. USBFuzz discovered a total of 26 new bugs, including 16 memory bugs of high-security impact in various Linux subsystems (USB core, USB sound, and network), one bug in FreeBSD, three in MacOS (two resulting in an unplanned reboot and one freezing the system), and four in Windows 8 and Windows 10 (resulting in Blue Screens of Death), and one bug in the Linux USB host controller driver and another one in a USB camera driver. From the Linux bugs, we have fixed and upstreamed 11 bugs and received 10 CVEs.

Boosting Fuzzer Efficiency: An Information Theoretic Perspective (FSE 2020)

Paper

Abstract: In this paper, we take the fundamental perspective of fuzzing as a learning process. Suppose before fuzzing, we know nothing about the behaviors of a program P: What does it do? Executing the first test input, we learn how P behaves for this input. Executing the next input, we either observe the same or discover a new behavior. As such, each execution reveals 'some amount' of information about P's behaviors. A classic measure of information is Shannon's entropy. Measuring entropy allows us to quantify how much is learned from each generated test input about the behaviors of the program. Within a probabilistic model of fuzzing, we show how entropy also measures fuzzer efficiency. Specifically, it measures the general rate at which the fuzzer discovers new behaviors. Intuitively, efficient fuzzers maximize information.

From this information theoretic perspective, we develop Entropic, an entropy-based power schedule for greybox fuzzing which assigns more energy to seeds that maximize information. We implemented Entropic into the popular greybox fuzzer LibFuzzer. Our experiments with more than 250 open-source programs (60 million LoC) demonstrate a substantially improved efficiency and confirm our hypothesis that an efficient fuzzer maximizes information. Entropic has been independently evaluated and invited for integration into main-line LibFuzzer. Entropic will run on more than 25,000 machines fuzzing hundreds of security-critical software systems simultaneously and continuously.

Fuzzing Error Handling Code using Context-Sensitive Software Fault Injection (USENIX Security2020)

Abstract: Error handling code is often critical but difficult to test in reality. As a result, many hard-to-find bugs exist in error handling code and may cause serious security problems once triggered. Fuzzing has become a widely used technique for finding software bugs nowadays. Fuzzing approaches mutate and/or generate various inputs to cover infrequently-executed code. However, existing fuzzing approaches are very limited in testing error handling code, because some of this code can be only triggered by occasional errors (such as insufficient memory and network-connection failures), but not specific inputs. Therefore, existing fuzzing approaches in general cannot effectively test such error handling code.

In this paper, we propose a new fuzzing framework named FIFUZZ, to effectively test error handling code and detect bugs. The core of FIFUZZ is a context-sensitive software fault injection (SFI) approach, which can effectively cover error handling code in different calling contexts to find deep bugs hidden in error handling code with complicated contexts. We have implemented FIFUZZ and evaluated it on 9 widely-used C programs. It reports 317 alerts which are caused by 50 unique bugs in terms of the root causes. 32 of these bugs have been confirmed by related developers. We also compare FIFUZZ to existing fuzzing tools (including AFL, AFLFast, AFLSmart and FairFuzz), and find that FIFUZZ finds many bugs missed by these tools. We believe that FIFUZZ can effectively augment existing fuzzing approaches to find many real bugs that have been otherwise missed.

FANS: Fuzzing Android Native System Services via Automated Interface Analysis (USENIX Security2020)

Abstract: Android native system services provide essential supports and fundamental functionalities for user apps. Finding vulnerabilities in them is crucial for Android security. Fuzzing is one of the most popular vulnerability discovery solutions, yet faces several challenges when applied to Android native system services. First, such services are invoked via a special interprocess communication (IPC) mechanism, namely binder, via service-specific interfaces. Thus, the fuzzer has to recognize all interfaces and generate interface-specific test cases automatically. Second, effective test cases should satisfy the interface model of each interface. Third, the test cases should also satisfy the semantic requirements, including variable dependencies and interface dependencies.

In this paper, we propose an automated generation-based fuzzing solution FANS to find vulnerabilities in Android native system services. It first collects all interfaces in target services and uncovers deep nested multi-level interfaces to test. Then, it automatically extracts interface models, including feasible transaction code, variable names and types in the transaction data, from the abstract syntax tree (AST) of target interfaces. Further, it infers variable dependencies in transactions via the variable name and type knowledge, and infers interface dependencies via the generation and use relationship. Finally, it employs the interface models and dependency knowledge to generate sequences of transactions, which have valid formats and semantics, to test interfaces of target services. We implemented a prototype of FANS from scratch and evaluated it on six smartphones equipped with a recent version of Android, i.e., android-9.0.0_r46 , and found 30 unique vulnerabilities deduplicated from thousands of crashes, of which 20 have been confirmed by Google. Surprisingly, we also discovered 138 unique Java exceptions during fuzzing.

Fuzzing IPC with Knowledge Inference (SRDS 2019)

Paper

Abstract: Sandboxing provides a strong security guarantee for applications, by isolating untrusted code into separated compartments. Untrusted code could only use IPC (inter-process communication) to launch sensitive actions, which are implemented in trusted (and maybe privileged) code. IPC-related security bugs in trusted code could facilitate jailbreaks of sandboxing, and thus are becoming high-value targets. However, finding vulnerabilities that could be triggered by IPC is challenging, due to the fact that IPC communication is stateful and format-sensitive.

In this paper, we propose a new fuzzing solution to discover IPC bugs in IPC services without source code, by combining static analysis and dynamic analysis. We use static analysis to recognize format checks and help construct IPC messages of valid formats. We then use dynamic analysis to infer the constraints between IPC messages, and model the stateful logic with a probability matrix. Therefore, we are able to generate highquality IPC messages to test IPC services, and discover deep and complex IPC bugs. Without loss of generality, we implemented a prototype MACHFUZZER, for a specific complicated and crucial IPC service, i.e., WindowServer in macOS. This prototype helps us find 12 previously unknown vulnerabilities in WindowServer in 48 hours. Among them, three vulnerabilities are confirmed exploitable, and could be exploited to escape the sandbox and gain root privilege.

HYPER-CUBE: High-Dimensional Hypervisor Fuzzing (NDSS 2020)

Abstract: Applying modern fuzzers to novel targets is often a very lucrative venture. Hypervisors are part of a very critical code base: compromising them could allow an attacker to compromise the whole cloud infrastructure of any cloud provider. In this paper, we build a novel fuzzer that aims explicitly at testing modern hypervisors. Our high throughput fuzzer design for long running interactive targets allows us to fuzz a large number of hypervisors, both open source, and proprietary. In contrast to one-dimensional fuzzers such as AFL, HYPER-CUBE can interact with any number of interfaces in any order.

Our evaluation shows that we can find more bugs (over 2x) and coverage (as much as 2x) than state of the art hypervisor fuzzers. Additionally, in most cases, we were able to do so using multiple orders of magnitude less time than comparable fuzzers. HYPER-CUBE was also able to rediscover a set of well-known vulnerabilities for hypervisors, such as VENOM, in less than five minutes. In total, HYPER-CUBE found 54 novel bugs, and so far we obtained 37 CVEs. Our evaluation results demonstrates that next generation coverage-guided fuzzers should incorporate a higher-throughput design for long running targets such as hypervisors.

Reproducible Crashes: Fuzzing Pharo by Mutating the Test Methods (IWST20 2020)

Paper
Code

Abstract: Fuzzing (or Fuzz Testing) is a technique to verify the robustness of a program-under-test. Valid input is replaced by random values with the goal to force the program-under-test into unresponsive states. In this position paper, we propose a white box Fuzzing approach by transforming (mutating) existing test methods. We adopt the mechanisms used for test amplification to generate crash inducing tests, which developers can reproduce later. We provide anecdotal evidence that our approach towards Fuzzing reveals crashing issues in the Pharo environment.

Opening Pandoras Box through ATFuzzer: Dynamic Analysis of AT Interface for Android Smartphones (ACSAC 2019)

Paper
Code

Abstract: This paper focuses on checking the correctness and robustness ofthe AT command interface exposed by the cellular baseband processor through Bluetooth and USB. A device's application processor uses this interface for issuing high-level commands (or, AT commands) to the baseband processor for performing cellular network operations (e.g., placing a phone call). Vulnerabilities in this interface can be leveraged by malicious Bluetooth peripherals to launch pernicious attacks including DoS and privacy attacks. To identify such vulnerabilities, we propose ATFuzzer that uses a grammar-guided evolutionary fuzzing approach which mutates production rules of the AT command grammar instead of concrete AT commands. Empirical evaluation with ATFuzzer on 10 Android smartphones from 6 vendors revealed 4 invalid AT command grammars over Bluetooth and 13 over USB with implications ranging from DoS, downgrade of cellular protocol version (e.g., from 4G to 3G/2G) to severe privacy leaks. The vulnerabilities along with the invalid AT command grammars were responsibly disclosed to affected vendors and two of the reported vulnerabilities have been already assigned CVEs (CVE-2019-16400 and CVE-2019-16401).

FuzzFactory: Domain-Specific Fuzzing with Waypoints (OOPSLA 2019)

Abstract: Coverage-guided fuzz testing has gained prominence as a highly effective method of finding security vulnerabilities such as buffer overflows in programs that parse binary data. Recently, researchers have introduced various specializations to the coverage-guided fuzzing algorithm for different domain-specific testing goals, such as finding performance bottlenecks, generating valid inputs, handling magic-byte comparisons, etc. Each such solution can require non-trivial implementation effort and produces a distinct variant of a fuzzing tool. We observe that many of these domain-specific solutions follow a common solution pattern.

In this paper, we present FuzzFactory, a framework for developing domain-specific fuzzing applications without requiring changes to mutation and search heuristics. FuzzFactory allows users to specify the collection of dynamic domain-specific feedback during test execution, as well as how such feedback should be aggregated. FuzzFactory uses this information to selectively save intermediate inputs, called waypoints, to augment coverage-guided fuzzing. Such waypoints always make progress towards domain-specific multi-dimensional objectives. We instantiate six domain-specific fuzzing applications using FuzzFactory: three re-implementations of prior work and three novel solutions, and evaluate their effectiveness on benchmarks from Google's fuzzer test suite. We also show how multiple domains can be composed to perform better than the sum of their parts. For example, we combine domain-specific feedback about strict equality comparisons and dynamic memory allocations, to enable the automatic generation of LZ4 bombs and PNG bombs.

Compiler Fuzzing: How Much Does It Matter (OOPSLA2019)

Paper

Abstract: Despite much recent interest in randomised testing (fuzzing) of compilers, the practical impact of fuzzer-found compiler bugs on real-world applications has barely been assessed. We present the first quantitative and qualitative study of the tangible impact of miscompilation bugs in a mature compiler. We follow a rigorous methodology where the bug impact over the compiled application is evaluated based on (1) whether the bug appears to trigger during compilation; (2) the extent to which generated assembly code changes syntactically due to triggering of the bug; and (3) whether such changes cause regression test suite failures, or whether we can manually find application inputs that trigger execution divergence due to such changes. The study is conducted with respect to the compilation of more than 10 million lines of C/C++ code from 309 Debian packages, using 12% of the historical and now fixed miscompilation bugs found by four state-of-the-art fuzzers in the Clang/LLVM compiler, as well as 18 bugs found by human users compiling real code or as a by-product of formal verification efforts. The results show that almost half of the fuzzer-found bugs propagate to the generated binaries for at least one package, in which case only a very small part of the binary is typically affected, yet causing two failures when running the test suites of all the impacted packages. User-reported and formal verification bugs do not exhibit a higher impact, with a lower rate of triggered bugs and one test failure. The manual analysis of a selection of the syntactic changes caused by some of our bugs (fuzzer-found and non fuzzer-found) in package assembly code, shows that either these changes have no semantic impact or that they would require very specific runtime circumstances to trigger execution divergence.

RVFuzzer: Finding Input Validation Bugs in Robotic Vehicles through Control-Guided Random Testing (USENIX Security2019)

Paper
Slides

Abstract: Robotic vehicles (RVs) are being adopted in a variety of application domains. Despite their increasing deployment, many security issues with RVs have emerged, limiting their wider deployment. In this paper, we address a new type of vulnerability in RV control programs, called input validation bugs, which involve missing or incorrect validation checks on control parameter inputs. Such bugs can be exploited to cause physical disruptions to RVs which may result in mission failures and vehicle damages or crashes. Furthermore, attacks exploiting such bugs have a very small footprint: just one innocent-looking ground control command, requiring no code injection, control flow hijacking or sensor spoofing. To prevent such attacks, we propose RVFuzzer, a vetting system for finding input validation bugs in RV control programs through control-guided input mutation. The key insight behind RVFuzzer is that the RV control model, which is the generic theoretical model for a broad range of RVs, provides helpful semantic guidance to improve bug-discovery accuracy and efficiency. Specifically, RVFuzzer involves a control instability detector that detects control program misbehavior, by observing (simulated) physical operations of the RV based on the control model. In addition, RVFuzzer steers the input generation for finding input validation bugs more efficiently, by leveraging results from the control instability detector as feedback. In our evaluation of RVFuzzer on two popular RV control programs, a total of 89 input validation bugs are found, with 87 of them being zero-day bugs.

Engineering a Better Fuzzer with Synergically Integrated Optimizations (ISSRE 2019)

Paper

Abstract: State-of-the-art fuzzers implement various optimizations to enhance their performance. As the optimizations reside in different stages such as input seed selection and mutation, it is tempting to combine the optimizations in different stages. However, our initial attempts demonstrate that naive combination actually worsens the performance, which explains that most optimizations are still isolated by stages and metrics. In this paper, we present InteFuzz, the first framework that synergically integrates multiple fuzzing optimizations. We analyze the root cause for performance degradation in naive combination, and discover optimizations conflict in coverage criteria and optimization granularity. To resolve the conflicts, we propose a novel priority-based scheduling mechanism. The dynamic integration considers both branch-based and block-based coverage feedbacks that are used by most fuzzing optimizations. In our evaluation, we extract four optimizations from popular fuzzers such as AFLFast and FairFuzz and compare InteFuzz against naive combinations. The evaluation results show that InteFuzz outperforms the naive combination by 29% and 26% in path-and branch-coverage. Additionally, InteFuzz triggers 222 more unique crashes, and discovers 33 zero-day vulnerabilities in real-world projects with 12 registered as CVEs.

Fuzzing Error Handling Code in Device Drivers Based on Software Fault Injection (ISSRE 2019)

Paper

Abstract: Device drivers remain a main source of runtime failures in operating systems. To detect bugs in device drivers, fuzzing has been commonly used in practice. However, a main limitation of existing fuzzing approaches is that they cannot effectively test error handling code. Indeed, these fuzzing approaches require effective inputs to cover target code, but much error handling code in drivers is triggered by occasional errors (such as insufficient memory and hardware malfunctions) that are not related to inputs. In this paper, based on software fault injection, we propose a new fuzzing approach named FIZZER, to test error handling code in device drivers. At compile time, FIZZER uses static analysis to recommend possible error sites that can trigger error handling code. During driver execution, by analyzing runtime information, it automatically fuzzes error-site sequences for fault injection to improve code coverage. We evaluate FIZZER on 18 device drivers in Linux 4.19, and in total find 22 real bugs. The code coverage is increased by over 15% compared to normal execution without fuzzing.

What You Corrupt Is Not What You Crash: Challenges in Fuzzing Embedded Devices (NDSS 2018)

Paper
Slides

Abstract: As networked embedded systems are becoming more ubiquitous, their security is becoming critical to our daily life. While manual or automated large scale analysis of those systems regularly uncover new vulnerabilities, the way those systems are analyzed follows often the same approaches used on desktop systems. More specifically, traditional testing approaches relies on observable crashes of a program, and binary instrumentation techniques are used to improve the detection of those faulty states. In this paper, we demonstrate that memory corruptions, a common class of security vulnerabilities, often result in different behavior on embedded devices than on desktop systems. In particular, on embedded devices, effects of memory corruption are often less visible. This reduces significantly the effectiveness of traditional dynamic testing techniques in general, and fuzzing in particular. Additionally, we analyze those differences in several categories of embedded devices and show the resulting impact on firmware analysis. We further describe and evaluate relatively simple heuristics which can be applied at run time (on an execution trace or in an emulator), during the analysis of an embedded device to detect previously undetected memory corruptions.

FOT: A Versatile, Configurable, Extensible Fuzzing Framework (FSE 2018)

Paper
Slides

Abstract: Greybox fuzzing is one of the most effective approaches for detecting software vulnerabilities. Various new techniques have been continuously emerging to enhance the effectiveness and/or efficiency by incorporating novel ideas into different components of a greybox fuzzer. However, there lacks a modularized fuzzing framework that can easily plugin new techniques and hence facilitate the reuse, integration and comparison of different techniques. To address this problem, we propose a fuzzing framework, namely Fuzzing Orchestration Toolkit (FOT). FOT is designed to be versatile, configurable and extensible. With FOT and its extensions, we have found 111 new bugs from 11 projects. Among these bugs, 18 CVEs have been assigned. Video link: https://youtu.be/O6Qu7BJ8RP0

Designing New Operating Primitives to Improve Fuzzing Performance (CCS 2017)

Paper
Code

Abstract: Fuzzing is a software testing technique that finds bugs by repeatedly injecting mutated inputs to a target program. Known to be a highly practical approach, fuzzing is gaining more popularity than ever before. Current research on fuzzing has focused on producing an input that is more likely to trigger a vulnerability. In this paper, we tackle another way to improve the performance of fuzzing, which is to shorten the execution time of each iteration. We observe that AFL, a state-of-the-art fuzzer, slows down by 24x because of file system contention and the scalability of fork() system call when it runs on 120 cores in parallel. Other fuzzers are expected to suffer from the same scalability bottlenecks in that they follow a similar design pattern. To improve the fuzzing performance, we design and implement three new operating primitives specialized for fuzzing that solve these performance bottlenecks and achieve scalable performance on multi-core machines. Our experiment shows that the proposed primitives speed up AFL and LibFuzzer by 6.1 to 28.9x and 1.1 to 735.7x, respectively, on the overall number of executions per second when targeting Google's fuzzer test suite with 120 cores. In addition, the primitives improve AFL's throughput up to 7.7x with 30 cores, which is a more common setting in data centers. Our fuzzer-agnostic primitives can be easily applied to any fuzzer with fundamental performance improvement and directly benefit large-scale fuzzing and cloud-based fuzzing services.

Chizpurfle: A Gray-Box Android Fuzzer for Vendor Service Customizations (ISSRE 2017)

Paper

Abstract: Android has become the most popular mobile OS, as it enables device manufacturers to introduce customizations to compete with value-added services. However, customizations make the OS less dependable and secure, since they can introduce software flaws. Such flaws can be found by using fuzzing, a popular testing technique among security researchers.This paper presents Chizpurfle, a novel "gray-box" fuzzing tool for vendor-specific Android services. Testing these services is challenging for existing tools, since vendors do not provide source code and the services cannot be run on a device emulator. Chizpurfle has been designed to run on an unmodified Android OS on an actual device. The tool automatically discovers, fuzzes, and profiles proprietary services. This work evaluates the applicability and performance of Chizpurfle on the Samsung Galaxy S6 Edge, and discusses software bugs found in privileged vendor services

Fuzzware: Using Precise MMIO Modeling for Effective Firmware Fuzzing (USENIX Security2022)

Paper
Code

Abstract: As embedded devices are becoming more pervasive in our everyday lives, they turn into an attractive target for adversaries. Despite their high value and large attack surface, applying automated testing techniques such as fuzzing is not straightforward for such devices. As fuzz testing firmware on constrained embedded devices is inefficient, state-of-the-art approaches instead opt to run the firmware in an emulator (through a process called re-hosting). However, existing approaches either use coarse-grained static models of hardware behavior or require manual effort to re-host the firmware.

We propose a novel combination of lightweight program analysis, re-hosting, and fuzz testing to tackle these challenges. We present the design and implementation of Fuzzware, a software-only system to fuzz test unmodified monolithic firmware in a scalable way. By determining how hardware-generated values are actually used by the firmware logic, Fuzzware can automatically generate models that help focusing the fuzzing process on mutating the inputs that matter, which drastically improves its effectiveness.

We evaluate our approach on synthetic and real-world targets comprising a total of 19 hardware platforms and 77 firmware images. Compared to state-of-the-art work, Fuzzware achieves up to 3.25 times the code coverage and our modeling approach reduces the size of the input space by up to 95.5%. The synthetic samples contain 66 unit tests for various hardware interactions, and we find that our approach is the first generic re-hosting solution to automatically pass all of them. Fuzzware discovered 15 completely new bugs including bugs in targets which were previously analyzed by other works; a total of 12 CVEs were assigned.

Auto Off-Target: Enabling Thorough and Scalable Testing for Complex Software Systems (ASE 2022)

Abstract: Software systems powering OS kernels, basebands, bootloaders, firmware, IoT or automotive build the foundation of infrastructure that billions of people rely on every day. Testing these systems is crucial, especially as their complexity grows and they are often written in unsafe languages such as C/C++.

However, testing such complex systems poses significant challenges, e.g., custom hardware for which there is no emulator, or a non-trivial setup of testing and debugging on the target device. As a result, the commonly used testing techniques and tools are not always easily applicable.

An off-target (OT) testing is a promising technique which addresses these challenges: part of the code is extracted and adapted to run on a different hardware platform with better tool support, easier debugging and higher test throughput. Unfortunately, since the process of creating an OT program has been manual, the technique did not scale well and was mostly used in an ad hoc manner.

In this paper we present a novel complex systems testing approach called Auto Off-target (AoT). Based on the information extracted from the source code and from the build process, AoT can automatically generate OT programs in C. AoT goes beyond the code generation and provides mechanisms that help to recreate and discover the program state in the OT code. The generated OTs are self-contained and independent of the original build environment. As a result, pieces of complex or embedded software can be easily run, analyzed, debugged and tested on a standard x86_64 machine.

We evaluate AoT on tens of thousands of functions selected from OS kernels, a bootloader and a network stack. We demonstrate we can run fuzzing and symbolic execution on the majority of the generated OTs. We further used AoT in a bug finding campaign and discovered seven bugs in the Android redfin and oriole kernels powering Google Pixel 5 and 6 phones.

Page Traffic Analysis

License

bo-wang-bob/FuzzingPaper

About

Resources

License

Stars

Watchers

Forks

Releases

Packages 0

Packages