Skip to main content

lightweight wrapper around basic LLVM functionality

Project description

# llvmlite

A lightweight LLVM python binding for writing JIT compilers

Old [llvmpy](https://github.com/llvmpy/llvmpy) binding exposes a lot of LLVM but the mapping of C++ style memory management to python is error prone. Numba and many JIT compiler does not need a full LLVM API. Only the IR builder, optimizer, and JIT compiler APIs are necessary.

llvmlite is a project originally tailored for Numba's needs, using the following approach:

- A small C wrapper around the parts of the LLVM C++ API we need that are
not already exposed by the LLVM C API.
- A ctypes Python wrapper around the C API.
- A pure Python implementation of the subset of the LLVM IR builder that we
need for Numba.

## Key Benefits

- llvmlite IR builder is pure python
- llvmlite uses LLVM assembly parser which provide better error messages (no more segfault due to bad IR construction)
- Most of llvmlite uses the LLVM C API which is small but very stable and backward compatible (low maintenance when changing LLVM version)
- The binding is not a python C-extension (no need to wrestle with Python's compiler requirement and C++11 compatibility)
- llvmlite binding memory management is explicit obj.close() or with-context
- llvmlite is quite faster in early benchmarks (the Numba test suite is twice faster)

## LLVMPY Compatibility Layer

`llvmlite.llvmpy.*` provides a minimal LLVMPY compatibility layer.

## Dependencies

You need Python 2.6 or greater (including Python 3.3 or greater).

### Build dependencies

- LLVM 3.5
- A C compiler, and possibly other tools (see below)

### Runtime dependencies

- For Python versions before 3.4, the [enum34](https://pypi.python.org/pypi/enum34) package is required

## Build

Run `python setup.py build`. This will build the llvmlite C wrapper, which will contain a statically-linked copy of the required subset of LLVM.

If your LLVM is installed in a non-standard location, first point the `LLVM_CONFIG` environment variable to the path of the corresponding `llvm-config` executable.

### Unix requirements

You must have a LLVM 3.5 build (libraries and header files) available somewhere. If it is not installed in a standard location, you may have to tweak the build script. Under Ubuntu 14.10 and Debian unstable, you can install `llvm-3.5-dev`. Versions shipped with earlier versions (Ubuntu 14.04, Debian stable) may not be good enough.

When building on Ubuntu, the linker may report an error if the development version of ``libedit`` is not installed. Install ``libedit-dev`` if you run into this problem.

### Windows requirements

You must have Visual Studio 2012 or later (the free Express edition is ok). In addition, you must have [cmake](http://www.cmake.org/) installed, and LLVM should have been built using cmake, in Release mode. Be careful to use the right bitness (32- or 64-bit) for your Python installation.

## Run tests

Run `python runtests.py` or `python -m llvmlite.tests`.

## Install

Run `python setup.py install`.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

llvmlite-0.4.0.tar.gz (69.9 kB view details)

Uploaded Source

File details

Details for the file llvmlite-0.4.0.tar.gz.

File metadata

  • Download URL: llvmlite-0.4.0.tar.gz
  • Upload date:
  • Size: 69.9 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No

File hashes

Hashes for llvmlite-0.4.0.tar.gz
Algorithm Hash digest
SHA256 8c1af4290db494563228cf6be8731f5069b9906e88712645a03a545822e27b16
MD5 7d2b6e9edfe16df476385e3492eeac39
BLAKE2b-256 501035919d99aba3da81953c05b41acd46ef8ae0b9cba22cad66954b69d4930b

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page