Fuzzing tool (TFuzz): a fuzzing tool based on program transformation

Related tags

Deep LearningT-Fuzz
Overview

T-Fuzz

T-Fuzz consists of 2 components:

  • Fuzzing tool (TFuzz): a fuzzing tool based on program transformation
  • Crash Analyzer (CrashAnalyzer): a tool that verifies whether crashes found transformed programs are true bugs in the original program or not (coming soon).

OS support

The current version is tested only on Ubuntu-16.04, while trying to run the code, please use our tested OS.

Prerequisite

T-Fuzz system is built on several opensource tools.

Installing radare2

$ git clone https://github.com/radare/radare2.git
$ cd radare2
$ ./sys/install.sh

Installing python libraries

installing some dependent libraries

Note: to use apt-get build-dep, you need to uncomment the deb-src lines in your apt source file (/etc/apt/sources.list) and run apt-get update.

$ sudo apt-get install build-essential gcc-multilib libtool automake autoconf bison debootstrap debian-archive-keyring
$ sudo apt-get build-dep qemu-system
$ sudo apt-get install libacl1-dev

installing pip and setting up virtualenv & wrapper

$ sudo apt-get install python-pip python-virtualenv
$ pip install virtualenvwrapper

Add the following lines to your shell rc file (~/.bashrc or ~/.zshrc).

export WORKON_HOME=$HOME/.virtual_envs
source /usr/local/bin/virtualenvwrapper.sh

Creating a python virtual environment

$ mkvirtualenv tfuzz-env

Installing dependent libraries

This command will install all the dependent python libraries for you.

$ workon tfuzz-env
$ pip install -r req.txt

Fuzzing target programs with T-Fuzz

$ ./TFuzz  --program  
   
     --work_dir 
    
      --target_opts

Where

  • : the path to the target program to fuzz
  • : the directory to save the results
  • : the options to pass to the target program, like AFL, use @@ as placeholder for files to mutate.

Examples

  1. Fuzzing base64 with T-Fuzz
$ ./TFuzz  --program  target_programs/base64  --work_dir workdir_base64 --target_opts "-d @@"
  1. Fuzzing uniq with T-Fuzz
$ ./TFuzz  --program  target_programs/uniq  --work_dir workdir_uniq --target_opts "@@"
  1. Fuzzing md5sum with T-Fuzz
$ ./TFuzz  --program  target_programs/md5sum  --work_dir workdir_md5sum --target_opts "-c @@"
  1. Fuzzing who with T-Fuzz
$ ./TFuzz  --program  target_programs/who  --work_dir workdir_who --target_opts "@@"

Using CrashAnalyzer to verify crashes

T-Fuzz CrashAnalyzer has been put in a docker image, however, it is still not working in all binaries we tested, we are still investigating it the cause.

Here is how:

Run the following command to run our docker image

$ [sudo] docker pull tfuzz/tfuzz-test
$ [sudo] docker run  --security-opt seccomp:unconfined -it tfuzz/tfuzz-test  /usr/bin/zsh

In the container:

There are 3 directories:

  • release: contains code the built lava binaries
  • results: contains some results we found in lava-m dataset
  • radare2: it is a program used by T-Fuzz.

Currently, T-Fuzz may not work, because the tracer crashes accidentally. And the CrashAnalyzer can not work on all results. But some cases can be recovered.

For example:

To verify bugs in base64, first goto release and checkout ca_base64:

$ cd release
$ git checkout ca_base64

Then we use a transformed program to recover the crash in the original program:

  1. Choose a transformed program and run it on the input found by a fuzzer:
$ cd ~
$./results/ca_base64/554/base64_tfuzz_28/base64_tfuzz_28 -d ./results/ca_base64/554/crashing_inputs_from/results_saved_0_from 
[1]    131 segmentation fault (core dumped)  ./results/ca_base64/554/base64_tfuzz_28/base64_tfuzz_28 -d
  1. Recover an input from this transformed program and crashing input
). Re-hooking. WARNING | 2018-12-04 04:28:23,228 | angr.project | Address is already hooked, during hook(0x90dd000, ). Re-hooking. WARNING | 2018-12-04 04:28:23,229 | angr.simos.linux | Tracer has been heavily tested only for CGC. If you find it buggy for Linux binaries, we are sorry! Adding = 65) && (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_0_0_8 <= 90)), ((file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_0_0_8 >= 97) && (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_0_0_8 <= 122)), ((file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_0_0_8 >= 48) && (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_0_0_8 <= 57)), (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_0_0_8 == 43), (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_0_0_8 == 47))> Adding = 65) && (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_1_1_8 <= 90)), ((file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_1_1_8 >= 97) && (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_1_1_8 <= 122)), ((file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_1_1_8 >= 48) && (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_1_1_8 <= 57)), (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_1_1_8 == 43), (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_1_1_8 == 47))> Adding = 65) && (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_2_2_8 <= 90)), ((file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_2_2_8 >= 97) && (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_2_2_8 <= 122)), ((file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_2_2_8 >= 48) && (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_2_2_8 <= 57)), (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_2_2_8 == 43), (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_2_2_8 == 47))> Adding = 65) && (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_3_3_8 <= 90)), ((file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_3_3_8 >= 97) && (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_3_3_8 <= 122)), ((file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_3_3_8 >= 48) && (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_3_3_8 <= 57)), (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_3_3_8 == 43), (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_3_3_8 == 47))> results saved to /root/base64_result/recover_0 "> 
$ ./release/CrashAnalyzer  --tprogram ./results/ca_base64/554/base64_tfuzz_28/base64_tfuzz_28 --target_opts "-d @@" --crash_input ./results/ca_base64/554/crashing_inputs_from/results_saved_0_from --result_dir base64_result --save_to recover
WARNING | 2018-12-04 04:28:22,350 | angr.analyses.disassembly_utils | Your verison of capstone does not support MIPS instruction groups.
Trying /root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from
WARNING | 2018-12-04 04:28:23,228 | angr.project | Address is already hooked, during hook(0x9021cd0, 
        
         ). Re-hooking.
WARNING | 2018-12-04 04:28:23,228 | angr.project | Address is already hooked, during hook(0x90dd000, 
         
          ). Re-hooking.
WARNING | 2018-12-04 04:28:23,229 | angr.simos.linux | Tracer has been heavily tested only for CGC. If you find it buggy for Linux binaries, we are sorry!
Adding 
          
           = 65) && (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_0_0_8 <= 90)), ((file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_0_0_8 >= 97) && (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_0_0_8 <= 122)), ((file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_0_0_8 >= 48) && (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_0_0_8 <= 57)), (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_0_0_8 == 43), (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_0_0_8 == 47))>
Adding 
           
            = 65) && (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_1_1_8 <= 90)), ((file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_1_1_8 >= 97) && (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_1_1_8 <= 122)), ((file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_1_1_8 >= 48) && (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_1_1_8 <= 57)), (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_1_1_8 == 43), (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_1_1_8 == 47))>
Adding 
            
             = 65) && (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_2_2_8 <= 90)), ((file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_2_2_8 >= 97) && (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_2_2_8 <= 122)), ((file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_2_2_8 >= 48) && (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_2_2_8 <= 57)), (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_2_2_8 == 43), (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_2_2_8 == 47))> Adding 
             
              = 65) && (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_3_3_8 <= 90)), ((file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_3_3_8 >= 97) && (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_3_3_8 <= 122)), ((file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_3_3_8 >= 48) && (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_3_3_8 <= 57)), (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_3_3_8 == 43), (file_/root/results/ca_base64/554/crashing_inputs_from/results_saved_0_from_9_3_3_8 == 47))> results saved to /root/base64_result/recover_0

Then /root/base64_result/recover_0 is generated, we can use it to trigger a crash in the original program.

  1. verify the input by running the generated input on the original program
$ ./results/base64 -d base64_result/recover_0 
Successfully triggered bug 554, crashing now!
Successfully triggered bug 554, crashing now!
Successfully triggered bug 554, crashing now!
[1]    177 segmentation fault (core dumped)  ./results/base64 -d base64_result/recover_0
Owner
HexHive
Enforcing memory safety guarantees and type safety guarantees at the compiler and runtime level
HexHive
GitHub Repository
Official implementation of CATs: Cost Aggregation Transformers for Visual Correspondence NeurIPS'21

CATs: Cost Aggregation Transformers for Visual Correspondence NeurIPS'21 For more information, check out the paper on [arXiv]. Training with different

Sunghwan Hong 120 Jan 04, 2023
Official Implementation of DE-CondDETR and DELA-CondDETR in "Towards Data-Efficient Detection Transformers"

DE-DETRs By Wen Wang, Jing Zhang, Yang Cao, Yongliang Shen, and Dacheng Tao This repository is an official implementation of DE-CondDETR and DELA-Cond

Wen Wang 41 Dec 12, 2022
Pytorch implementation of our paper under review -- 1xN Pattern for Pruning Convolutional Neural Networks

1xN Pattern for Pruning Convolutional Neural Networks (paper) . This is Pytorch re-implementation of "1xN Pattern for Pruning Convolutional Neural Net

Mingbao Lin (林明宝) 29 Nov 29, 2022
Imitating Deep Learning Dynamics via Locally Elastic Stochastic Differential Equations

Imitating Deep Learning Dynamics via Locally Elastic Stochastic Differential Equations This repo contains official code for the NeurIPS 2021 paper Imi

Jiayao Zhang 2 Oct 18, 2021
YOLOv5 🚀 is a family of object detection architectures and models pretrained on the COCO dataset

YOLOv5 🚀 is a family of object detection architectures and models pretrained on the COCO dataset, and represents Ultralytics open-source research int

阿才 73 Dec 16, 2022
Code for Boundary-Aware Segmentation Network for Mobile and Web Applications

BASNet Boundary-Aware Segmentation Network for Mobile and Web Applications This repository contain implementation of BASNet in tensorflow/keras. comme

Hamid Ali 8 Nov 24, 2022
Confident Semantic Ranking Loss for Part Parsing

Confident Semantic Ranking Loss for Part Parsing

Jiachen Xu 5 Oct 22, 2022
Make differentially private training of transformers easy for everyone

private-transformers This codebase facilitates fast experimentation of differentially private training of Hugging Face transformers. What is this? Why

Xuechen Li 73 Dec 28, 2022
Fully Automatic Page Turning on Real Scores

Fully Automatic Page Turning on Real Scores This repository contains the corresponding code for our extended abstract Henkel F., Schwaiger S. and Widm

Florian Henkel 7 Jan 02, 2022
Official repository of the paper Learning to Regress 3D Face Shape and Expression from an Image without 3D Supervision

Official repository of the paper Learning to Regress 3D Face Shape and Expression from an Image without 3D Supervision

Soubhik Sanyal 689 Dec 25, 2022
Towards Fine-Grained Reasoning for Fake News Detection

FinerFact This is the PyTorch implementation for the FinerFact model in the AAAI 2022 paper Towards Fine-Grained Reasoning for Fake News Detection (Ar

Ahren_Jin 15 Dec 15, 2022
Python library to receive live stream events like comments and gifts in realtime from TikTok LIVE.

TikTokLive A python library to connect to and read events from TikTok's LIVE service A python library to receive and decode livestream events such as

Isaac Kogan 277 Dec 23, 2022
A method that utilized Generative Adversarial Network (GAN) to interpret the black-box deep image classifier models by PyTorch.

A method that utilized Generative Adversarial Network (GAN) to interpret the black-box deep image classifier models by PyTorch.

Yunxia Zhao 3 Dec 29, 2022
A PyTorch-based open-source framework that provides methods for improving the weakly annotated data and allows researchers to efficiently develop and compare their own methods.

Knodle (Knowledge-supervised Deep Learning Framework) - a new framework for weak supervision with neural networks. It provides a modularization for se

93 Nov 06, 2022
Offcial implementation of "A Hybrid Video Anomaly Detection Framework via Memory-Augmented Flow Reconstruction and Flow-Guided Frame Prediction, ICCV-2021".

HF2-VAD Offcial implementation of "A Hybrid Video Anomaly Detection Framework via Memory-Augmented Flow Reconstruction and Flow-Guided Frame Predictio

76 Dec 21, 2022
Mixup for Supervision, Semi- and Self-Supervision Learning Toolbox and Benchmark

OpenSelfSup News Downstream tasks now support more methods(Mask RCNN-FPN, RetinaNet, Keypoints RCNN) and more datasets(Cityscapes). 'GaussianBlur' is

AI Lab, Westlake University 332 Jan 03, 2023
Code of Classification Saliency-Based Rule for Visible and Infrared Image Fusion

CSF Code of Classification Saliency-Based Rule for Visible and Infrared Image Fusion Tips: For testing: CUDA_VISIBLE_DEVICES=0 python main.py For trai

Han Xu 14 Oct 31, 2022
RefineNet: Multi-Path Refinement Networks for High-Resolution Semantic Segmentation

Multipath RefineNet A MATLAB based framework for semantic image segmentation and general dense prediction tasks on images. This is the source code for

Guosheng Lin 575 Dec 06, 2022
This is the official Pytorch implementation of the paper "Diverse Motion Stylization for Multiple Style Domains via Spatial-Temporal Graph-Based Generative Model"

Diverse Motion Stylization (Official) This is the official Pytorch implementation of this paper. Diverse Motion Stylization for Multiple Style Domains

Soomin Park 28 Dec 16, 2022
Neural Logic Inductive Learning

Neural Logic Inductive Learning This is the implementation of the Neural Logic Inductive Learning model (NLIL) proposed in the ICLR 2020 paper: Learn

36 Nov 28, 2022