GithubHelp home page GithubHelp logo

lucabongiorni / afl-crash-analyzer Goto Github PK

View Code? Open in Web Editor NEW

This project forked from floyd-fuh/afl-crash-analyzer

0.0 2.0 0.0 144 KB

Another crash analyzer tool for the American Fuzzy Lop (AFL) fuzzer

License: GNU General Public License v3.0

Python 89.46% Shell 10.54%

afl-crash-analyzer's Introduction

afl crash analyzer: Another crash analysis and triage tool for American Fuzzy Lop (AFL) results.

At the moment very alpha.

Goals of this project:
- Simple, flexible python scripts for crash analysis
- Everything file based
- There is no command line argument parsing planned, but you should modify the main python script (or run it more or less in default configuration)

Installation:
0. Install AFL (to your $PATH)
1. Prepare your linux virtual box, do whatever makes sense for a fuzzing machine, for example:
 a) sudo apt-get remove whoopsie
2. Install exploitable gdb script if you want to use it
3. Compile binaries (one plain, one ASAN, one instrumented is recommended) if you haven't already (see test-cases folder for examples)
4. Start with reading AflCrashAnalyzer.py and configure the CrashAnalysisConfig object in there (examples included)
5. Start AflCrashAnalyzer.py with python 2.7

A couple of notes: 
- program assume you have all AFL binaries in your $PATH! (aka make install)
- The target binary is started as: /opt/binary-instrumented args_before input-file args_after
- You might want to update the code for all your binaries to the newest developer version before compilation (even if you fuzzed on older version), so you filter out crashes that were fixed by the maintainer in the meantime. Check test-cases folder for examples.
- afl instrumented binary is the minimum, recommend is to also set at least one of the following binaries
- target_binary_plain (non-instrumented, with symbols, no ASAN)
- target_binary_asan (non-instrumented, with symbols, with ASAN)
- To get line numbers for files when running gdb, use clang's ASAN and specifiy ASAN_SYMBOLIZER_PATH (env configuration option), see https://code.google.com/p/address-sanitizer/wiki/Flags
- When you configure the gdb script, DON'T add the "run" command at the beginning and the "quit" command at the end
- The main python file has code that shows how *I* like to run stuff, maybe you have other preferences

Bugs and todo's:
- see "TODO" in code

License:
- GPL version 3

Features for now:
- crash input file deduplication (on filesize + md5)
- running binary with each crash input file and divide the input file into folders for each signal
- For crash input files that create signals you are interested in you can:
  - Run the binary and put stdout/stderr into a .txt file next to the crash input file
    - The same for ASAN enabled binaries
    - Run the binary with gdb in batch mode and a configurable gdb script (for example including the "exploitable" gdb script)
- Minimize input crashes with afl-tmin, deduplication on the results
  

afl-crash-analyzer's People

Contributors

floyd-fuh avatar

Watchers

James Cloos avatar  avatar

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.