https://github.com/ChrisCummins/ProGraML

A Graph-based Program Representation for Data Flow Analysis and Compiler Optimizations
https://github.com/ChrisCummins/ProGraML

compiler-irs control-flow data-flow graph-neural-networks graph-representation llvm llvm-ir machine-learning programming-languages

Last synced: 4 days ago
JSON representation

A Graph-based Program Representation for Data Flow Analysis and Compiler Optimizations

• Host: GitHub
• URL: https://github.com/ChrisCummins/ProGraML
• Owner: ChrisCummins
• License: other
• Created: 2019-11-25T02:49:35.000Z (over 4 years ago)
• Default Branch: development
• Last Pushed: 2024-05-22T23:14:58.000Z (about 1 month ago)
• Last Synced: 2024-06-23T10:01:18.451Z (7 days ago)
• Topics: compiler-irs, control-flow, data-flow, graph-neural-networks, graph-representation, llvm, llvm-ir, machine-learning, programming-languages
• Language: C++
• Homepage:
• Size: 51.1 MB
• Stars: 288
• Watchers: 10
• Forks: 63
• Open Issues: 23
• Metadata Files:
  • Readme: README.md
  • Changelog: CHANGELOG.md
  • Contributing: CONTRIBUTING.md
  • License: LICENSE

Lists

• awesome-machine-learning-in-compilers - programl - LLVM and XLA IR program representation for machine learning ([paper](https://arxiv.org/pdf/2003.10536.pdf)). (Software / Memory/Cache Modeling/Analysis)
• awesome-mlir - ProGraML

README

        
ProGraML: Program Graphs for Machine Learning




  

  

      

  

  

  

      

  

  

  

      

  

  

  

    

  

  

  

    

  

  

  

    

  





  An expressive, language-independent representation of programs.





  

    Check the website

    for more information.

  



## Introduction

ProGraML is a representation for programs as input to a machine learning model.

The key features are:

1. **Simple:** Everything is available through a `pip install`, no compilation

   required. Supports several programming languages (*C, C++, LLVM-IR, XLA*) and

   several graph formats (*NetworkX, DGL, Graphviz, JSON*) out of the box.

2. **Expressive:** Captures every control, data, and call relation across entire

   programs. The representation is independent of the source language. Features

   and labels can be added at any granularity to support whole-program,

   per-instruction, or per-relation reasoning tasks.

3. **Fast:** The core graph construction is implemented in C++ with a low

   overhead interface to Python. Every API method supports simple and efficient

   parallelization through an `executor` parameter.

To get stuck in and play around with our graph representation, visit:



  



Or if papers are more your ☕, have a read of ours:



  



## Supported Programming Languages

The following programming languages and compiler IRs are supported

out-of-the-box:

  

    Language

    API Calls

    Supported Versions

  

  

    C

    

      programl.from_cpp(),

      programl.from_clang()

    

    Up to ISO C 2017

  

  

    C++

    

      programl.from_cpp(),

      programl.from_clang()

    

    Up to ISO C++ 2020 DIS

  

  

    LLVM-IR

    

      programl.from_llvm_ir()

    

    3.8.0, 6.0.0, 10.0.0

  

  

    XLA

    

      programl.from_xla_hlo_proto()

    

    2.0.0

  

Is your favorite language not supported here? Submit a [feature

request](https://github.com/ChrisCummins/ProGraML/issues/new/choose)!

## Getting Started

Install the latest release of the Python package using:

```

pip install -U programl

```

The API is very simple, comprising graph *creation* ops, graph *transform* ops,

and graph *serialization* ops. Here is a quick demo of each:

```py

>>> import programl as pg

# Construct a program graph from C++:

>>> G = pg.from_cpp("""

... #include 

...

... int main(int argc, char** argv) {

...   std::cout << "Hello, world!" << std::endl;

...   return 0;

... }

... """)

# A program graph is a protocol buffer:

>>> type(G).__name__

'ProgramGraph'

# Convert the graph to NetworkX:

>>> pg.to_networkx(G)

# Save the graph for later:

>>> pg.save_graphs('file.data', [G])

```

For further details check out the [API

reference](https://chriscummins.cc/ProGraML/api/python.html).

## Contributing

Patches, bug reports, feature requests are welcome! Please use the

[issue tracker](https://github.com/ChrisCummins/ProGraML/issues) to file a

bug report or question. If you would like to help out with the code, please

read [this document](CONTRIBUTING.md).

## Citation

If you use ProGraML in any of your work, please cite [this

paper](https://chriscummins.cc/pub/2021-icml.pdf):

```

@inproceedings{cummins2021a,

  title={{ProGraML: A Graph-based Program Representation for Data Flow Analysis and Compiler Optimizations}},

  author={Cummins, Chris and Fisches, Zacharias and Ben-Nun, Tal and Hoefler, Torsten and O'Boyle, Michael and Leather, Hugh},

  booktitle = {Thirty-eighth International Conference on Machine Learning (ICML)},

  year={2021}

}

```