An Efficient Parallel Replay Method Based on Shared Memory Dependency Pair

YING Huan; WANG Dong-hui; WANG Zhe

Chinese

您当前的位置：

首页 >

文章列表页 >

An Efficient Parallel Replay Method Based on Shared Memory Dependency Pair

更新时间：2026-09-04

- An Efficient Parallel Replay Method Based on Shared Memory Dependency Pair
- Vol. 34, Issue 1, Pages: 6-11(2017)
- 作者机构：
  
  1. 中国科学院机构声学研究所,北京,100190
  2. 中国科学院大学机构,北京,101408
  3. 中国科学院计算技术研究所,北京,100080
- 作者简介：
- 基金信息：
- DOI：
  CLC：
- Published：2017
- 稿件说明：
移动端阅览
应欢, 王东辉, 王喆. 一种基于访存依赖对的并行重放方法[J]. 微电子学与计算机, 2017, 34(1): 6-11.

YING Huan, WANG Dong-hui, WANG Zhe. An Efficient Parallel Replay Method Based on Shared Memory Dependency Pair[J]. 2017, 34(1): 6-11.
应欢, 王东辉, 王喆. 一种基于访存依赖对的并行重放方法[J]. 微电子学与计算机, 2017, 34(1): 6-11. DOI：

YING Huan, WANG Dong-hui, WANG Zhe. An Efficient Parallel Replay Method Based on Shared Memory Dependency Pair[J]. 2017, 34(1): 6-11. DOI：

摘要

确定性重放技术能够为并行程序提供复现执行的功能.为了加速重放执行

本文提出一种基于访存依赖对的并行重放方法

该方法通过离线分析记录阶段生成的日志文件

识别出日志文件中具有访存依赖关系的日志项

在重放执行时仅按序执行上述日志项.采用PARSEC测试集进行性能评估

实验结果表明该方法在重放阶段引入的开销较小.

Abstract

Deterministic replay provides the possibility of reproducing parallel program execution. In order to get excellent performance in replay-runs

this paper proposes a parallel replay method based on shared memory dependency parir

which conducted by offline analysis of original log generated in record-runs to recognise the record items with memory dependency relationship

and each thread only executes these record items in order in replay-runs. Performance evaluation using PARSEC benchmark shows that this method introduces little overhead in replay-runs.

关键词

Keywords

references

Honarmand N, Torrellas J. Replay debugging: Leveraging record and replay for program debugging[C]//ISCA 2014. USA, Minnesota, 2014: 445-456.

唐士斌. 多线程程序的确定性调试方法[D]. 北京: 中国科学院计算技术研究所, 2014.

Pokam G, Pereira C, Danne K, et al. Hardware and software approaches for deterministic multi-processor replay of concurrent programs[J]. Intel Technol, 2009, 13(4)20-41. DOI: 20

Leblanc T J, Mellor-Crummey J M. Debugging parallel programs with instant replay[J]. IEEE Trans. Comput, 1987, 34(4): 471-482. DOI: 471

Liu P, Zhang X, Tripp O, et al. Light: Replay via tightly bounded recording[C]//PLDI 2015. USA, Portland, 2015: 55-64.

Patil H, Pereira C, Stallcup M, et al. PinPlay: a framework for deterministic replay and reproducible analysis of parallel programs[C]//CGO 2010. Canada, Toronto, 2010: 2-11.

Courtois P J, Heymans F, Parnas D L. Concurrent Control with 'eaders' and 'writers'[J]. Commun. ACM, 1971, 14(10): 667-668. DOI: 667

Bhansali S, Chen W K, de Jong S, et al. Framework for instruction-level tracing and analysis of program executions[C]//VEE 2006. Canada, Ottawa, 2006: 154-163.

Bienia C, Kumar S, Singh J P, et al. The PARSEC benchmark suite: characterization and architectural implications[C]//PACT 2008. Canada, Toronto, 2008: 72-81.

应欢, 王东辉, 武成岗, 等. 适用于商用系统环境的低开销确定性重放技术[J]. 吉林大学学报(工学版), 2017, 49(1): 1782-1790. DOI: 1782

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Design and Implementation on the Data Race Free System Based on Combining Parallel and Serial Mechanism

Lightweight countermeasures against Spectre attacks on the branch prediction unit

Design of ultra-low power real-time clock circuits

Simulation on the DC characteristics of double heterojunction p-GaN HEMT with AlN/GaN hybrid buffer layer

Research on assembly action recognition based on MSS-YOLO-MPHands fusion architecture

Related Author

SHA Han

YING Huan

TANG Bo-wen

QIAO Yang

Related Institution

Capital Normal University

AI问答

Postal code：100079
Tel：（010）53879206 Email：tmw@bjxintong.com.cn
Technical support is provided by Beijing Founder electronics co., LTD 京ICP备09082226号-64 京公网安备11010602201714号
It is recommended to read the content of this site in Chrome&IE9+. Please switch to extreme mode in browser 360.
Cookies We use cookies to help provide and enhance our service and tailor content. By continuing, you agree to the use of cookies.

⁰