LibAFL, the fuzzer library.

Advanced Fuzzing Library - Slot your own fuzzers together and extend their features using Rust.

LibAFL is a collection of reusable pieces of fuzzers, written in Rust, it gives you many of the benefits of an off-the-shelf fuzzer, while being completely customizable. Some highlight features currently include:

fast: We do everything we can at compile time, keeping runtime overhead minimal. Users reach 120k execs/sec in frida-mode on a phone (using all cores).
scalable: Low Level Message Passing, LLMP for short, allows LibAFL to scale almost linearly over cores, and via TCP to multiple machines.
adaptable: You can replace each part of LibAFL. For example, BytesInput is just one potential form input: feel free to add an AST-based input for structured fuzzing, and more.
multi platform: LibAFL was confirmed to work on Windows, MacOS, Linux, and Android on x86_64 and aarch64. LibAFL can be built in no_std mode to inject LibAFL into obscure targets like embedded devices and hypervisors.
bring your own target: We support binary-only modes, like Frida-Mode, as well as multiple compilation passes for sourced-based instrumentation. Of course it's easy to add custom instrumentation backends.

Core concepts

LibAFL is fast, multi-platform, no_std compatible, and scales over cores and machines. It offers a main crate that provide building blocks for custom fuzzers, libafl, a library containing common code that can be used for targets instrumentation, libafl_targets, and a library providing facilities to wrap compilers, libafl_cc. It offers integrations with popular instrumentation frameworks. At the moment, the supported backends are:

SanitizerCoverage, in libafl_targets
Frida, in libafl_frida
QEMU user-mode and system mode, including hooks for emulation, in libafl_qemu
TinyInst, in libafl_tinyinst by elbiazo

Building and installing

Install the Dependencies

The Rust development language
- We highly recommend not to use e.g. your Linux distribution package as this is likely outdated. So rather install Rust directly, instructions can be found here.
LLVM tools
- The LLVM tools (including clang, clang++) are needed (newer than LLVM 15.0.0 up to LLVM 18.1.3) If you are using Debian/Ubuntu, again, we highly recommmend that you install the package from here
- (In libafl_concolic, we only support LLVM version newer than 18)
Cargo-make:
- We use cargo-make to build the fuzzers in fuzzers/ directory. You can install it with cargo install cargo-make

Clone the LibAFL repository with

git clone https://github.com/AFLplusplus/LibAFL

Build the library using

cargo build --release

Build the API documentation with

cargo doc

Browse the LibAFL book (WIP!) with (requires mdbook)

cd docs && mdbook serve

Getting started

We collect all example fuzzers in ./fuzzers. Be sure to read their documentation (and source), this is the natural way to get started!

cargo make run

You can run each example fuzzer with this following command, as long as the fuzzer directory has Makefile.toml file. The best-tested fuzzer is ./fuzzers/inprocess/libfuzzer_libpng, a multicore libfuzzer-like fuzzer using LibAFL for a libpng harness.

Resources

Installation guide
Online API documentation
The LibAFL book (WIP) online or in the repo
Our research paper
Our RC3 talk explaining the core concepts
Our Fuzzcon Europe talk with a (a bit but not so much outdated) step-by-step discussion on how to build some example fuzzers
The Fuzzing101 solutions & series of blog posts by epi
Blogpost on binary-only fuzzing lib libaf_qemu, Hacking TMNF - Fuzzing the game server, by RickdeJager.
A LibAFL Introductory Workshop, by Jordan Whitehead

Contributors

LibAFL is written and maintained by

Please check out CONTRIBUTING.md for the contributing guideline.

Cite

If you use LibAFL for your academic work, please cite the following paper:

@inproceedings{libafl,
 author       = {Andrea Fioraldi and Dominik Maier and Dongjia Zhang and Davide Balzarotti},
 title        = {{LibAFL: A Framework to Build Modular and Reusable Fuzzers}},
 booktitle    = {Proceedings of the 29th ACM conference on Computer and communications security (CCS)},
 series       = {CCS '22},
 year         = {2022},
 month        = {November},
 location     = {Los Angeles, U.S.A.},
 publisher    = {ACM},
}

License

^{Licensed under either of Apache License, Version
2.0 or MIT license at your option.}

_{Unless you explicitly state otherwise, any contribution intentionally submitted
for inclusion in this crate by you, as defined in the Apache-2.0 license, shall
be dual licensed as above, without any additional terms or conditions.}

AFLplusplus / LibAFL

readme