Program transformation and runtime support for threaded MPI execution on shared-memory machines

Parallel programs written in MPI have been widely used for developing high- performance applications on various platforms. Because of a restriction of the MPI computation model, conventional MPI implementations on shared-memor y machines map each MPI, node to an OS process, which can suffer serious pe rformance degradation in the presence of multiprogramming, This paper studi es compile-time and runtime techniques for enhancing performance portabilit y of MPI code running on multiprogrammed shared-memory machines. The propos ed techniques allow MPI nodes to be executed safely and efficiently as thre ads. Compile-time transformation eliminates global and static variables in C code using node-specific data. The runtime support includes an efficient and provably correct communication protocol that uses lock-free data struct ure and takes advantage of address space sharing among threads. The experim ents on SGI Origin 2000 show that our MPI prototype called TMPI using the p roposed techniques is competitive with SGI's native MPI implementation in a dedicated environment, and that it has significant performance advantages in a multiprogrammed environment.