Loop Fusion in High Performance Fortran Loop Fusion in High Performance Fortran

In this paper we investigate a unique problem associated with fusing loops within a High Performance Fortran (HPF) program. In particular, we discuss the issue of performing loop fusion in an HPF compiler when compiling Fortran90 array assignment statements for execution on a distributed-memory machine. During compilation of an HPF program, Fortran90 array assignment statements must be scalarized into loop nests. We show how a certain class of these loop nests, when fused, can cause problems for the compiler's distributed-memory code generator. We then present an algorithm which not only prevents the fusion of these loops, but also increases the amount of useful fusion that can be performed. 1 Introduction High-Performance Fortran (HPF))12], an extension of For-tran90, has attracted considerable attention as a promising language for writing portable parallel programs. HPF ooers a simple programming model shielding programmers from the intricacies of concurrent programming and managing distributed data. Programmers express data parallelism using Fortran90 array operations and use data layout directives to direct partitioning of the data and computation among the processors of a parallel machine. One transformation an HPF compiler must address is the scalarization of the Fortran90 array operations into serial DO-loops. Scalarization is often followed by loop fusion in an attempt to improve a program's data locality and data reuse characteristics. Unfortunately, the fusion of some of these scalarized loops can produce loops for which it is dif-cult to generate an eecient SPMD program. However, loop fusion is too important of an optimization to simply disable. To solve this dilemma, we have developed an algo