View Code? Open in Web Editor NEW

This project forked from eth-sri/learch

Dockerfile 0.07% CMake 3.55% LLVM 2.44% SMT 0.62% C 35.64% C++ 50.52% Makefile 0.33% Python 3.59% Shell 2.06% PHP 0.58% NASL 0.51% HTML 0.06% PowerShell 0.04%

learch's Introduction

Learch: a Learning-based Strategies for Path Exploration in Symbolic Execution

Learch is a learning-based state selection strategy for symbolic execution. It can achieve significantly more coverage and detects more security violations than existing manual heuristics. Learch is instantiated on KLEE. The directory klee contains our modified KLEE code (from this commit) and learch contains the Learch code. Learch is developed at SRI Lab, Department of Computer Science, ETH Zurich as part of the Machine Learning for Programming project. For more details, please refer to Learch CCS'21 paper.

Setup

We provide a docker file, which we recommend to start with. To set Learch up locally, one can follow the instructions in the docker file. To build and run:

$ docker build -t learch .
$ docker run -it learch

Usages

The following README files explains how to use Learch:

Obtaining the benchmarks used in the paper: benchmarks/README.md.
Training Learch: train/README.md.
Using Learch to test new programs: eval/README.md.

Citing Learch

@inproceedings{10.1145/3460120.3484813,
  author = {He, Jingxuan and Sivanrupan, Gishor and Tsankov, Petar and Vechev, Martin},
  title = {Learning to Explore Paths for Symbolic Execution},
  year = {2021},
  isbn = {9781450384544},
  publisher = {Association for Computing Machinery},
  address = {New York, NY, USA},
  url = {https://doi.org/10.1145/3460120.3484813},
  doi = {10.1145/3460120.3484813},
  booktitle = {Proceedings of the 2021 ACM SIGSAC Conference on Computer and Communications Security},
  pages = {2526–2540},
  numpages = {15},
  keywords = {fuzzing, symbolic execution, machine learning, program testing},
  location = {Virtual Event, Republic of Korea},
  series = {CCS '21}
}