摘要: Reuse signature, or reuse distance pattern, is an accurate model for program memory accessing behaviors. It has been studied and shown to be effective in analysis optimizations by many recent works. However, the high overhead associated with measurement restricts scope of its application. This paper explores applying sampling signature collection reduce time overhead. We compare different strategies show that enhanced systematic a uniform coverage all ranges can used extrapolate distribution. Based on analysis, we present novel method accuracy more than 99%. Our average speedup 7.5 while best improvement observed 34. first attempt utilize measuring signatures. Experiments varied programs instrumentation tools great potential promoting practical uses signatures enabling optimization opportunities.
acm.org
sci-hub.se 参考文章(33)
Martin Hirzel, Bursty Tracing: A Framework for Low-Overhead Temporal Profiling ACM. ,(2001)
Rabin Andrew Sugumar, Multi-configuration simulation algorithms for the evaluation of computer architecture designs University of Michigan. ,(1993)
Chen Ding, Ken Kennedy, Improving effective bandwidth through compiler enhancement of global and dynamic cache reuse Rice University. ,(2000)
Frank Olken, Efficient methods for calculating the success function of fixed space replacement policies Performance Evaluation. ,vol. 3, pp. 153- 154 ,(1983) , 10.1016/0166-5316(83)90024-X
M. Arnold, D. Grove, Collecting and Exploiting High-Accuracy Call Graph Profiles in Virtual Machines symposium on code generation and optimization. pp. 51- 62 ,(2005) , 10.1109/CGO.2005.9
Xipeng Shen, Jonathan Shaw, Brian Meeker, Chen Ding, Locality approximation using time symposium on principles of programming languages. ,vol. 42, pp. 55- 61 ,(2007) , 10.1145/1190215.1190227
Changpeng Fang, S. Can, S. Onder, Zhenlin Wang, Instruction based memory distance analysis and its application to optimization international conference on parallel architectures and compilation techniques. pp. 27- 37 ,(2005) , 10.1109/PACT.2005.26
Chen Ding, Yutao Zhong, Predicting whole-program locality through reuse distance analysis programming language design and implementation. ,vol. 38, pp. 245- 257 ,(2003) , 10.1145/780822.781159
Yul H. Kim, Mark D. Hill, David A. Wood, Implementing stack simulation for highly-associative memories measurement and modeling of computer systems. ,vol. 19, pp. 212- 213 ,(1991) , 10.1145/107971.107995
Marco Zagha, Brond Larson, Steve Turner, Marty Itzkowitz, Performance Analysis Using the MIPS R10000 Performance Counters conference on high performance computing (supercomputing). pp. 16- 16 ,(1996) , 10.1145/369028.369059