Skip to content


Switch branches/tags

Name already in use

A tag already exists with the provided branch name. Many Git commands accept both tag and branch names, so creating this branch may cause unexpected behavior. Are you sure you want to create this branch?


Failed to load latest commit information.
Latest commit message
Commit time

alt text

Build status Build Status


B2R2 is a collection of useful algorithms, functions, and tools for binary analysis, written purely in F# (in .NET lingo, it is purely managed code). B2R2 has been named after R2-D2, a famous fictional robot appeared in the Star Wars. In fact, B2R2's original name was B2-R2, but we decided to use the name B2R2 instead, because .NET does not allow dash (-) characters in identifiers (or namespaces). The name essentially represents "binary" or "two": "binary" itself means "two" states anyways. "B" and "2" mean "binary", and "R" indicates reversing.


  1. B2R2 is analysis-friendly: it is written in F#, which provides all the syntactic goodies for writing program analyzers, such as pattern matching, algebraic data types, and etc.

  2. B2R2 is fast: it has a fast and efficient front-end engine for binary analysis, which is written in a functional-first way. Therefore, it naturally supports pure parallelism for binary disassembling, lifting and IR optimization.

  3. B2R2 is easy to play with: there is absolutely no dependency hell for B2R2 because it is a fully-managed library. All you need to do is to install .NET Core SDK, and you are ready to go! Native IntelliSense support is another plus!

  4. B2R2 is OS-Independent: it works on Linux, Mac, Windows, and etc. as long as .NET core supports it.

  5. B2R2 is interoperable: it is not bound to a specific language. Theoretically, you can use B2R2 APIs with any CLI supported languages.


B2R2 supports instruction parsing, binary disassembly, assembly, control-flow recovery, and many more. B2R2 also comes with several user-level command-line tools that are similar to readelf and objdump, although our tools are platform-agnostic. B2R2 currently supports four binary file formats: ELF, PE, Mach-O, and WebAssembly.

Below is a list of features that we currently support. Some of them are work in progress, but we look forward to your contributions! Feel free to write a PR (Pull Request) while making sure that you have read our contribution guideline.

Feature x86 x86-64 ARMv7 ARMv8 MIPS32 MIPS64 EVM TMS320C600 AVR PPC SPARC64 SH4 RISC-V
Instruction Parsing 🌕 🌕 🌕 🌕 🌕 🌕 🌕 🌕 🌕 🌕 🌑 🌕 🌑
Disassembly 🌕 🌕 🌕 🌕 🌕 🌕 🌕 🌕 🌕 🌕 🌑 🌕 🌑
Lifting 🌕 🌕 🌕 🌕 🌕 🌕 🌕 🌑 🌕 🌑 🌑 🌑 🌑
CFG Recovery 🌕 🌕 🌓 🌓 🌓 🌓 🌕 🌑 🌑 🌑 🌑 🌑 🌑
Data-Flow 🌕 🌕 🌕 🌕 🌕 🌕 🌕 🌑 🌑 🌑 🌑 🌑 🌑
Instruction Emulation 🌕 🌕 🌕 🌕 🌕 🌕 🌑 🌑 🌑 🌑 🌑 🌑 🌑
Assembly 🌕 🌕 🌑 🌑 🌑 🌑 🌑 🌑 🌑 🌑 🌑 🌑 🌑
REPL 🌕 🌕 🌑 🌑 🌑 🌑 🌑 🌑 🌑 🌑 🌑 🌑 🌑
ROP Compilation 🌕 🌑 🌑 🌑 🌑 🌑 🌑 🌑 🌑 🌑 🌑 🌑 🌑


B2R2 relies on a tiny set of external .NET libraries, and our design principle is to use a minimum number of libraries. Below is a list of libraries that we leverage.

Note about our middle-end

We have made significant changes in our middle-end (CFG recovery, and function identification, etc.) engines for this version, and we are still improving it. The current version is stable enough, but we are actively changing the implementation while doing some internal research, which is hoding us back for open-sourcing it. Hence, we decided to partly publicize our tool (everything but the middle-end engine). We always welcome PRs for our front-end modules 😄

We may bump few more versions before making everything public, but we will eventually open-source everything. So please stay tuned!

API Documentation

We currently use fsdocs to generate our documentation:


Let's try to use B2R2 APIs.

  1. First we create an empty directory DIRNAME:

    mkdir DIRNAME
    cd DIRNAME
  2. We then create an empty console project with dotnet command line:

    $ dotnet new console -lang F#
  3. Add our nuget package B2R2.FrontEnd to the project:

    $ dotnet add package B2R2.FrontEnd.BinInterface
  4. Modify the Program.fs file with your favorite editor as follows:

    open B2R2
    open B2R2.FrontEnd
    let main argv =
      let isa = ISA.OfString "amd64"
      let bytes = [| 0x65uy; 0xffuy; 0x15uy; 0x10uy; 0x00uy; 0x00uy; 0x00uy |]
      let hdl = BinHandle.Init (isa, bytes)
      let ins = BinHandle.ParseInstr hdl 0UL
      ins.Translate hdl.TranslationContext |> printfn "%A"
  5. We then just run it by typing: dotnet run. You will be able see lifted IR statements from your console. That's it! You just lifted an Intel instruction with only few lines of F# code!


Building B2R2 is fun and easy. All you need to do is to install .NET 6 SDK or above. Yea, that's it!

  • To build B2R2 in release mode, type make release or dotnet build -c Release in the source root.

  • To build B2R2 in debug mode, type make, or dotnet build in the source root.

For your information, please visit the official web site of F# to get more tips about installing the development environment for F#:


Members in SoftSec Lab. @ KAIST developed B2R2 in collaboration with Cyber Security Research Center (CSRC) at KAIST. See Authors for the full list.


If you plan to use B2R2 in your own research. Please consider citing our paper:

  author = {Minkyu Jung and Soomin Kim and HyungSeok Han and Jaeseung Choi and Sang Kil Cha},
  title = {{B2R2}: Building an Efficient Front-End for Binary Analysis},
  booktitle = {Proceedings of the NDSS Workshop on Binary Analysis Research},
  year = 2019


Here are papers using our work. Please create a PR if you want to add yours.

  • How'd Security Benefit Reverse Engineers? The Implication of Intel CET on Function Identification, In Proceedings of the International Conference on Dependable Systems Networks, 2022 (PDF)
  • Smartian: Enhancing Smart Contract Fuzzing with Static and Dynamic Data-Flow Analyses, In Proceedings of the IEEE/ACM International Conference on Automated Software Engineering, 2021 (PDF)
  • NTFuzz: Enabling Type-Aware Kernel Fuzzing on Windows with Static Binary Analysis, In Proceedings of the IEEE Symposium on Security and Privacy, 2021 (PDF)