期刊文献+

基于代码剖析的定制指令识别

Custom Instructions Identification Based on Code Profiling
下载PDF
导出
摘要 代码剖析是分析程序行为和发现性能瓶颈的重要手段。根据可重配置处理器的定制指令设计需要,提出了一种基于虚拟机技术的代码剖析方法AID-prof。该方法的优势在于结构无关以及静态\动态分析的无缝契合。基于AID-prof,设计了一种自动化的定制指令识别过程CID。实验显示,AID-prof可有效地发掘程序热点,并且示例程序通过CID产生的定制指令获得明显加速。 Code profiling is a key technique for application behavior analysis and bottlenecks discovery. According to the requirement of designing custom instructions for reconfigurable processors, AID-prof, a novel profiling method based on virtual machine was presented. The benefits of the presented method are architecture-independent and close combination between static and dynamic analysis. Based on AID-prof, an automatic custom instructions identification procedure named CID was proposed. Experiments results show that AID-prof can discover application hot spots effectively, and custom instructions identified by CID can markedly speed up application execution.
出处 《计算机科学》 CSCD 北大核心 2010年第3期144-148,共5页 Computer Science
基金 国家863高技术研究发展计划项目(2007AA01Z2b3) 国家973重点基础研究发展计划前期研究专项项目(2007CB310608)资助
关键词 可重配置处理器 定制指令 识别 代码剖析 虚拟机 Reconfigurable processors, Custom instructions, Identification, Code profiling, Virtual machine
  • 相关文献

参考文献10

  • 1GNU. Profiling a Program: Where Does It Spend Its Time ? [EB/OL]. http://sourceware. org/binutils/docs-2. 16/gprof/.
  • 2Seward J, Nethercote N, Weidendorfer J. Valgrind 3. 3-Advanced Debugging and Profiling for GNU/Linux applications [M]. Network Theory Ltd.
  • 3Srivastava A, Eustace A. ATOM: A System for Building Customized Program Analysis Tools [C]//Proc. of ACM SIGPLAN Conference on Programming Language Design and Implementation. Orlando, Florida, 1994 : 196-205.
  • 4Karuri K, Faruque M A A, et al. Fine-grained application source code profiling for asip design [C]//Proc. of the 42nd Annual Conference on Design Automation. Anaheim, California, USA, 2005 : 329-334.
  • 5Wilson R P, French R S, Wilson C S, et al. SUIF: An Infrastructure for Research on Parallelizing and Optimizing Compilers[J]. SIGPLAN Notices (S0362-1340), 1994,29(12) : 31-37.
  • 6Smith M D. Machine-SUIF: A Research Infrastructure for Profile-driven and Machine-specific Optimizations [EB/OL]. http://www. eecs. harvard.edu/hube/software/software. html.
  • 7Atasu K, Pozzi L, Ienne P. Exact and Approximate Algorithms for the Extension of Embedded Processor Instruction Sets [J]. IEEE Transactions on Computer-aided Design of Integrated Circuits and Systems (S0278-0070), 2006,25 (7) : 1209-1229.
  • 8Lattner C, Adve V. LLVM: A Compilation Framework for Lifelong Program Analysis & Transformation [C]//Proc. of the 2004 International Symposium on Code Generation and Optimization. Palo Alto, California, USA, 2004 : 75-86.
  • 9Lee C,Potkonjak M,Mangione-Smith W H. Mediabench.-A Tool for Evaluating and Synthesizing Multimedia and Communications Systems [C]//Proc. Int. Syrup. Microarch. , 1997:330-337.
  • 10Guthaus M R,Ringenberg J S,Ernst D,et al. MiBench:A Free, Commercially Representative Embedded Benchmark Suite [C]// Proc. IEEE 4th Ann. Workshop Workload Characterization. 2001 : 3-14.

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部