Aggressive loop parallelization using speculative execution mechanisms
作者： Yuguang Wu , Jin Lin
DOI:
关键词:
摘要: A system and method for aggressive loop parallelization using speculative execution is disclosed. The may include transforming code of a target application concurrent execution, which adding an instruction to create global address table entry each store operation on load different iteration dependent. replacing standard with special configured determine if operand the matches in one entries. Another remove following corresponding operation. If currently executing thread table, setting checkpoint, completing pre-fetch mode, re-executing from checkpoint.
lens.org 参考文章(27)
David A Helder, Kalyan Muthukumar, Compare speculation in software-pipelined loops ,(2000)
Anthony Roide, Antonio Valles, Dattatraya Kulkarni, Gautam Doshi, Dependence compensation for sparse computations ,(2002)
Ramsey W. Haddad, Stephan G. Meier, James B. Keller, Scheduler which retries load/store hit situations ,(2000)
Rajendra K. Bera, Detection of reduction variables in an assignment statement ,(2002)
Ramsey W. Haddad, Stephan G. Meier, James B. Keller, Scheduler which discovers non-speculative nature of an instruction after issuing and reissues the instruction ,(2000)
Daniel M. Lavery, Christopher J. Hughes, Jamison Collins, Ralph Kling, Hong Wang, Yong-Fong Lee, John Shen, Register rotation prediction and precomputation ,(2001)
Atsufumi Shibayama, Naoki Nishi, Sunao Torii, Satoshi Matsushita, Data dependency detection using history table of entry number hashed from memory address ,(2005)
David C. Sehr, Chris Newburn, Milind Girkar, Perry Wang, Xinmin Tian, Richard Grove, Hong Wang, Wei Li, John Shen, Thread-data affinity optimization using compiler ,(2005)
Hiroaki Shikano, Hironori c, Keiji c, Global compiler for controlling heterogeneous multiprocessor ,(2007)
Zhiyuan Li, Jenn-Yuan Tsai, Xin Wang, Pen-Chung Yew, Bess Zheng, Compiler techniques for concurrent multithreading with hardware speculation support Languages and Compilers for Parallel Computing. pp. 175- 191 ,(1997) , 10.1007/BFB0017252