首页 | 本学科首页   官方微博 | 高级检索  
     

跨平台系统级虚拟机的访存优化
引用本文:蔡嵩松,刘奇,沈海华,章隆兵.跨平台系统级虚拟机的访存优化[J].计算机研究与发展,2012(Z1):131-136.
作者姓名:蔡嵩松  刘奇  沈海华  章隆兵
作者单位:中国科学院计算机系统结构重点实验室;中国科学院计算技术研究所;中国科学院研究生院;龙芯中科技术有限公司
基金项目:国家自然科学基金项目(60736012,60921002,61070025);国家“八六三”高技术研究发展计划基金项目(2008AA010901);国家“九七三”重点基础研究发展计划基金项目(2005CB321600);国家“核高基”科技重大专项基金项目(2009ZX01028-002-003,2009ZX01029-001-003)
摘    要:跨平台系统级虚拟机软件模拟访存操作效率低,严重影响了虚拟机的性能.为提高跨平台虚拟机访存效率,提出了一种使用宿主系统TLB硬件、加速跨平台系统级虚拟机访存地址转换的软硬件协同优化方法.该方法相对于软件访存模拟方法,有效利用了宿主系统的硬件资源,提高了跨平台系统级虚拟机执行访存操作效率.实验结果表明该方法将虚拟机系统的整体性能提高了近15%.提出的方法已实际应用在龙芯系统级跨平台虚拟机中.

关 键 词:跨平台系统级虚拟机  访存模拟  地址转换  页表  TLB

Memory Accessing Optimization of Cross-Platform System-Level Virtual Machine
Cai Songsong, Liu Qi, Shen Haihua, and Zhang Longbing.Memory Accessing Optimization of Cross-Platform System-Level Virtual Machine[J].Journal of Computer Research and Development,2012(Z1):131-136.
Authors:Cai Songsong  Liu Qi  Shen Haihua  and Zhang Longbing
Affiliation:(Key Laboratory of Computer System and Architecture, Chinese Academy of Sciences, Beijing 100190) (Institute of Computing Technology, Chinese Academy of Sciences, Beijing 100190) (Graduate University of Chinese Academy of Sciences, Beijing 100049) (Loongson Technology Corporation Limited, Beijing 100190)
Abstract:The memory accessing efficiency of cross-platform system-level virtual machine is low, which affects the performance of the virtual machine greatly. In order to improve the efficiency, we present a hardware-software co-design method which can speed up the address translation of cross-platform system-level virtual machine by using TLB in host platform. It uses the hardware of the host platform efficiently, and improves the memory accessing efficiency of cross-platform system-level virtual machine. The result of evaluation shows that it improves the performance of virtual machine by 15%. The method is implemented in the system-level virtual machine of Loongson.
Keywords:cross-platform system-level virtual machine  simulation of memory accessing  address translation  page table  TLB
本文献已被 CNKI 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号