Abstract
Graph Traverse Scheduling, GTS [ALTB89], was developed as a technique suitable for parallelizing DO loops with cycles on their dependence graph, but actually, GTS is a generalization of many loop restructuring techniques. The improvements achieved with its application in front of other techniques are the reduction of synchronizations, the introduction of scalar optimizations and the better use of the memory hierarchy. The purpose of this work has been the implementation of this new technique for a particular existing shared-memory multiprocessor and its integration on a source-to-source restructuring compiler. We studied the limitations imposed by the architecture to the technique and modified it in order to adapt the generated parallel code to the architecture. Finally, we measured the improvements of the new scheduling comparing the execution time of some example loops obtained by the commercial parallelizer with the execution time obtained by GTS.
This work was supported by the Ministry of Education of Spain under contract TIC-880/92, by ESPRIT 6634 Basic Research Action (APPARC) and by the CEPBA (European Center of Parallelism of Barcelona).
This is a preview of subscription content, log in via an institution to check access.
Preview
Unable to display preview. Download preview PDF.
References
Aho A., Sheti R and Ullman J., “Principles of Compiler Design”, Reading, MA: Addison-Wesley, 1986.
Alliant Computer Systems Corp., “FX/Series Architecture Manual”, 1985.
Alliant Computer Corporation, “Product Summary”, Dec 1988.
Ayguadé E., Labarta J., Torres J. and Borensztejn P., “GTS: Parallelization and Vectorization of Tight Recurrences”, Proc. of the Supercomputing'89, pp.531–539, 1989.
Ayguadé E., Labarta J., Torres J., LLaberia J.M. and Valero M.,“Parallelism Evaluation and Partitioning of Nested Loops for Shared Memory Multiprocessors”, chap. 11 of Advances in Languages and Compilers for Parallel Processing, PITMAN, 1991.
Borensztejn P., Labarta J. and Barrado C., “Measures of Parallelism at Compile Time”, Euromicro Workshop on Parallel and Distributed Processing, Jan 1993.
Callaghan D., Cocke J., Kennedy K., “Estimating Interlock and Improving Balance for pipelined arquitectures”, Proc. of the International Conference on Parallel Processing, pp. 295–304, 1987.
Cytron R., “DOACROSS: Beyong Vectorization for Multiprocessors”, Proc. of the International Conference on Parallel Procesing, pp. 836–844, 1986.
Midkiff S., “Automatic Generation of Synchronization Instructions for Parallel Processors”, MS. Thesis, CSRD Report No. 588, Center of Supercomputing Research and Development, University of Illinois, IL, May 1986.
Midkiff S. and Padua D., “Compiler Algorithms for Synchronization”, IEEE Transactions on Computers, Vol C-36, No 12, pp.1485–1495, Dec 1987.
Padua D. and Wolfe M. “Advanced Compiler Optimizations for Supercomputers”, Communications of the ACM, vol. 29, no. 12, pp.1184–1201, December 1986.
Pointer L., “Perfect Report: 1”, CSRD Report No. 896, Center of Supercomputing Research and Development, University of Illinois, IL, July 1989.
Polychronopoulos C. and Kuck D.,“Guided Self-Scheduling: A Practical Scheduling Scheme for Parallel Supercomputres”, IEEE Transactions on Computers, vol. C–36, no. 12, December 1987.
Polychronopoulos C et al. “Parafrase-2: An Environment for Parallelizing, Partitioning, Synchronizing and Scheduling of Programs on Multiprocessors”, Proc of the International Conference on Parallel Processing, vol II, pp.39–48, 1989.
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 1994 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Barrado, C., Labarta, J., Borensztejn, P. (1994). Implementation of GTS. In: Halatsis, C., Maritsas, D., Philokyprou, G., Theodoridis, S. (eds) PARLE'94 Parallel Architectures and Languages Europe. PARLE 1994. Lecture Notes in Computer Science, vol 817. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3-540-58184-7_131
Download citation
DOI: https://doi.org/10.1007/3-540-58184-7_131
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-58184-0
Online ISBN: 978-3-540-48477-6
eBook Packages: Springer Book Archive