INSTALL.md

Option 1: Using Docker Image

To get into the container, you need:

• Download the Docker Image histfuzz.tar. You can download the compressed package (about 4G) from https://doi.org/10.5281/zenodo.7631379 or https://drive.google.com/file/d/1JN8hG8qKGoQMcn8Jnmbw3iqLpWYNhT_q/view?usp=sharing.
• Generate Image data file histfuzz.tar by using command unzip histfuzz.tar.zip.
• docker import histfuzz.tar
• docker run -it [image id] /bin/bash

Option 2: Installing From Sources

Step 1: Installing dependencies needed by HistFuzz

HistFuzz needs Python library efficient-apriori, antlr, and github3.py.

pip3 install efficient-apriori 
pip3 install antlr4-python3-runtime==4.9.2
pip3 install github3.py

Step 2: Installing subject solvers

• Install dependencies following the guidance of SMT solvers, e.g., z3 and cvc5.
• Download source, build, and install the solvers.
• To evaluate the code coverage of solvers achieved by testing tools, you shold enable gcov support when building solvers. For example, to build z3 with gcov, you can use the command:

CFLAGS=--coverage CXXFLAGS=--coverage LDFLAGS=--coverage python scripts/mk_make.py
cd build; make

For cvc5, you should use the command:

./configure.sh debug --coverage --auto-download
cd build; make

Step 3: Installing the baselines of evaluation

• Unzip the compressed files in the histfuzz/baselines directory.
• Install storm's dependencies and storm following the guidance in the website. For example:

cd storm
virtualenv --python=/usr/bin/python3 venv
source venv/bin/activate
python setup.py install
deactivate