Finding Free Schedules for Non-uniform Loops
作者： Volodymyr Beletskyy , Krzysztof Siedlecki
DOI: 10.1007/978-3-540-45209-6_44
关键词:
摘要: An algorithm, permitting us to build free schedules for arbitrary nested non-uniform loops, is presented. The operations of each time schedule can be executed as soon their operands are available. algorithm requires exact dependence analysis. To describe and implement the carry out experiences, analysis by Pugh Wonnacott was chosen where dependences found in form tuple relations. applied both non-parameterized parameterized loops. proposed has been implemented verified means Omega project software.
springer.com
tuniv.szczecin.pl 参考文章(36)
David A. Padua, Samuel P. Midkiff, A Comparison of Four Synchronization Optimization Techniques. international conference on parallel processing. pp. 9- 16 ,(1991)
Tatiana Shpeisman, David Wonnacott, William Pugh, Vadim Maslov, Wayne Kelly, Evan Rosser, The Omega Library interface guide University of Maryland at College Park. ,(1995)
William Pugh, Wayne Kelly, A framework for unifying reordering transformations University of Maryland at College Park. ,(1993)
Allen, Optimizing Compilers for Modern Architectures ,(2004)
Jingling Xue, Loop tiling for parallelism ,(2000)
Alain Darte, Frédéric Vivien, Optimal fine and medium grain parallelism detection in polyhedral reduced dependence graphs International Journal of Parallel Programming. ,vol. 25, pp. 447- 496 ,(1997) , 10.1023/A:1025168022993
Alain Darte, Yves Robert, Frederic Vivien, Scheduling and automatic parallelization ,(2000)
Fabien Quilleré, Sanjay Rajopadhye, Doran Wilde, Generation of Efficient Nested Loops from Polyhedra International Journal of Parallel Programming. ,vol. 28, pp. 469- 498 ,(2000) , 10.1023/A:1007554627716
Ding-Kai Chen, Compiler optimizations for parallel loops with fine-grained synchronization University of Illinois at Urbana-Champaign. ,(1994)
William Pugh, David Wonnacott, An Exact Method for Analysis of Value-based Array Data Dependences languages and compilers for parallel computing. pp. 546- 566 ,(1993) , 10.1007/3-540-57659-2_31