Compiling data-parallel programs to a distributed runtime environment with thread isomigration

Gabriel Antoniu, Luc Bougé, Raymond Namyst, and Christian Perez

Traditionally, the compilation of data-parallel languages is targeted to low-level runtime environments: abstract processors are mapped onto static system processes, which directly address the low-level IPC library. Alternatively, we propose to map each HPF abstract processor onto a ``lightweight process'' (thread) which can be freely migrated between nodes together with the data it manages, under the supervision of some external scheduler. We discuss the pros and cons of such an approach and the facilities which must be provided by the multithreaded runtime. We describe a prototype HPF compiler built along these lines, based on the Adaptor HPF compiler and the PM2 multithreaded runtime environment.

Reference

Gabriel Antoniu, Luc Bougé, Raymond Namyst, and Christian Perez. Compiling data-parallel programs to a distributed runtime environment with thread isomigration. In The 1999 Intl Conf. on Parallel and Distributed Processing Techniques and Applications (PDPTA '99), Technical Session on parallel and distributed languages: mechanisms, implementations, and tools, volume 4, pages 1756-1762, Las Vegas, NV, June 1999. Invited paper.
Postscript copy.