Software queue-based algorithms for pipelined synchronization on multiprocessors

作者: M. Takesue

DOI: 10.1109/ICPPW.2003.1240361

关键词:

摘要: Synchronization either ensures mutual exclusion on shared data or forces a processor to wait until set of variables becomes specific state; the latter is called conditional synchronization. We have improved performance multiprocessors by allowing processors concurrently access different parts in pipelined manner (Takesue, 2002). A special software tree queue-tail pointers key scheme for pipelining, but it requires other hardware schemes such as queue distributed caches. This paper proposes queue-based algorithms synchronization only with Fetch&Inc. support. pipeline exploiting tree. Conditional declaring semaphore structure, and simulating P V operations so that can be eagerly performed before accessing data. Evaluation results an RTL (register transfer level) simulator show compared non-pipelined synchronization, speedup our pipelining reaches up over 2.0 large heavily contentious cases.

参考文章(20)
Masaru Takesue, A model of pipelined mutual exclusion on cache-coherent multiprocessors european conference on parallel processing. pp. 917- 922 ,(2003) , 10.1007/978-3-540-45209-6_126
John L. Hennessy, David A. Patterson, Computer Architecture: A Quantitative Approach ,(1989)
Alain Kägi, Doug Burger, James R. Goodman, Efficient synchronization Proceedings of the 24th annual international symposium on Computer architecture - ISCA '97. ,vol. 25, pp. 170- 180 ,(1997) , 10.1145/264107.264166
Andrew S. Tanenbaum, Modern Operating Systems ,(1992)
T.E. Anderson, The performance of spin lock alternatives for shared-money multiprocessors IEEE Transactions on Parallel and Distributed Systems. ,vol. 1, pp. 6- 16 ,(1990) , 10.1109/71.80120
James Laudon, Daniel Lenoski, The SGI Origin Proceedings of the 24th annual international symposium on Computer architecture - ISCA '97. ,vol. 25, pp. 241- 251 ,(1997) , 10.1145/264107.264206
Edsger W. Dijkstra, The structure of the “THE”-multiprogramming system Communications of The ACM. ,vol. 11, pp. 341- 346 ,(1968) , 10.1145/363095.363143
James R. Goodman, Mary K. Vernon, Philip J. Woest, Efficient synchronization primitives for large-scale cache-coherent multiprocessors architectural support for programming languages and operating systems. ,vol. 17, pp. 64- 75 ,(1989) , 10.1145/68182.68188
A. Dinning, A survey of synchronization methods for parallel computers IEEE Computer. ,vol. 22, pp. 66- 77 ,(1989) , 10.1109/2.30733