Pysvf: Python Bindings for SVF

News

• SVF-Python now supports MTA bindings and more pointer analysis strategies(AndersenWaveDiff, AndersenBase, and Steensgaard bindings). (Thank JoelYYoung for his help!).

1. Introduction

Pysvf is a Python binding for the SVF (Static Value-Flow), a static analysis framework built on LLVM.

2. Installation

Pysvf can be installed in two ways:

Method 1: Install via pip (Test PyPI)

Requirements

• Python 3.8 - 3.11
• OS: Linux X86-64, Linux Arm64, MacOS

Install Command

python3 -m pip install -i https://test.pypi.org/simple/ pysvf

Method 2: Build from Source

Clone the repository and run the build script:

This will build the SVF and its depedencies like LLVM and Z3, and set up the Python bindings.

3. Documentation

Documentation is available on the SVF-Python Wiki:

• Pysvf API Reference

4. Tutorial

Example notebooks are provided in the tutorial folder. Before running them, please install Jupyter Notebook:

python3 -m pip install jupyter
cd demo
jupyter notebook

Then you can click one of the following notebooks.

Example Notebooks

• PAG: Generate Program Assignment Graph (PAG / SVFIR) from LLVM BC files and manipulate PAGNode (SVFVar).
• ICFG: Generate Interprocedural Control Flow Graph (ICFG) and manipulate ICFGNode, ICFGEdge, and subclasses.
• SVFG: Generate Static Value Flow Graph (SVFG) and manipulate SVFGNode, SVFGEdge, and subclasses.