Features

Par4All is an automatic parallelizing and optimizing compiler (workbench) for C and Fortran sequential programs.

The purpose of this source-to-source compiler is to adapt existing applications to various hardware targets such as multicore systems, high performance computers and GPUs. It creates a new source code and thus allows the original source code of the application to remain unchanged.

The main benefits of this source-to-source approach are:

• to keep the original source code of the application free from modifications,
• to obtain generated parallelized sources for various hardware platforms,
• to rely on vendor’s optimized target tools,
• to be able to optimize manually the generated source code.

The generated sources just need to be processed through the usual compilers:

• optimized compilers for a given processor,
• vendor compilers for embedded processors,
• CUDA,
• OpenCL,
• OpenMP,
• linkable with MPI and other libraries.

Par4All current version

The current 1.*x* version can generate CUDA and OpenCL code from C code and OpenMP from C and Fortran 77 code with a simple easy-to-use high-level script p4a. With this script, you can get a parallelized version of your sources or even call the backend compiler to generate executable binaries with gcc, nvcc or icc for example.

On the benchmarks page, there are some performance results with Par4All on multicores and GPU.

Currently there is no support for Windows. Mac OS X may work by compiling from the sources but is not supported. But you can use a virtual machine with Ubuntu or Debian 64-bit x86 on these systems to generate parallel versions of your programs.

What is going on?

The main development of the 1.4 branch is almost stopped since we are focusing our developments on the 2.0 version based on Clang that takes most of our time.

Warning: since the project is no longer supported by SILKAN, most of the developments are frozen, such as the Clang/LLVM/SoSlang for 2.x version. :-(

• Switching to Clang as the base framework for Par4All 2.*x*
• Scilab/Xcos and MATLAB/Simulink to OpenMP/CUDA/OpenCL with Wild Cruncher
• Python compilation & parallelization to OpenMP/CUDA/OpenCL
• Code generation for more embedded systems (Tilera, Kalray MPPA, ST P2012/STORM)
• More user-friendly interfaces (Eclipse...)
• Improving vector code generation (x86 SSE & AVX, ARM Neon, CUDA and OpenCL vectors)
• Better CUDA and OpenCL generation (loop fusion, shared memory...)
• Improving the OpenMP output
• Automatic instrumentation for loop parameters extraction at runtime
• Java compilation & parallelization to OpenMP/CUDA/OpenCL
• Finish the Fortran 95 support with the gcc/gfc front-end

Roadmap

• Par4All 0.1 and 0.2 went out to provide Fortran 77 to OpenMP parallelization to modernize legacy code and C to OpenMP parallelization. There were first releases to test the integration process and were not really distributed as packages or with high level compilation scripts;
• Par4All 1.0 (07/2010) parallelizes Fortran and C to OpenMP and C to CUDA and is the first easy-to-use public version;
• Par4All 1.1 (03/2011) deals with C99 and introduces basic support for Fortran 95 to OpenMP;
• Par4All 1.2 (07/2011) loop-fusion and communication optimizations for CUDA;
• Par4All 1.3.1 (01/2012) generates OpenCL;
• Par4All 1.4.3 (09/2012) deal with Spear-DE output;
• Par4All 2.0 : new version based on Clang/LLVM. The developments are on hold...

Internals

Internally, Par4All is currently composed of different components:

• the PIPS source-to-source compiler that began at MINES ParisTech in 1988 and is currently developed also in many other places: SILKAN, Institut TÉLÉCOM/TÉLÉCOM Bretagne, IT SudParis, RPI (Rensselaer Polytechnic Institute).
• the PolyLib used by PIPS,
• GCC/GFC for the Fortran95 parser,
• and of course own tools and scripts to make all these components and the global infrastructure usable.

Par4All is an open source project that merges various open source developments. More info on the community here.