Abstract
An approach to performing interprocedural analysis for program restructuring and parallelization is presented. Compared to previous approaches, it provides more information needed for most data-dependence test schemes. It is quite effective in loop parallelization, loop restructuring, as well as array alias recognition (without array linearization). The issue of recursive calls in program parallelization is also discussed. It is shown that the proposed approach can handle recursive calls quite effectively.
Original language | English (US) |
---|---|
Pages (from-to) | 221-228 |
Number of pages | 8 |
Journal | Proceedings of the International Conference on Parallel Processing |
Volume | 2 |
State | Published - Dec 1 1988 |