基于simics全系统模拟器的研究.docVIP

本科学生毕业论文 基于Simics全系统模拟器的研究 Research on A Full System Simulation Platform 论文作者 院 系 信息工程学院 专 业 信息工程专业 学 号 指导老师 完成日期 中文摘要 本文主要对全系统模拟器Simics的搭建与使用进行研究。通过在服务器与普通计算机上构建Simics 4.0模拟平台，实现了在Simics环境中，快速生成一个虚拟的目标系统，建立集群的网络连接。并且通过对简单嵌入式平台Ebony的构建与使用，了解了Simics模拟环境的搭建与调试方法。 以前，研究人员对于多处理器的研究是建立在真实设备上的，既浪费时间又浪费资金，而利用Simics可以搭建各种多处理器环境。本文设计了两种简单多处理器模型，包括对称(SMP)与非对称(AMP)模型。利用Simics提供的硬件设备模型语言DML与Python语言搭建了多处理器硬件模拟环境，并且在模拟环境中安装了Linux操作系统。最后通过Simics提供的环境测试命令、读写CPUID以及使用CPU亲和力函数验证了使用Simics模拟的多处理器系统搭建的正确性。 关键词：Simics 模拟器 DML 多处理器 CPUID CPU亲和力 Abstract This paper studies on the putting up and application of full system simulator Simics. The speedy generation of virtual objective system and establishment of cluster network in Simics environment are achieved through the setting up of Simics 4.0 simulation platform on server and general computers. And the putting up and debug methods of Simics simulation environment is comprehended via the putting up and application of simple embedded platform Ebony. In the past, the research of multi-processor is conducted by researchers on real device. It’s time-costing and a waste of capital. With the application of Simics, various multi-processor environments can be established. This paper designs two simple multi-processor models, including symmetric (SMP) model and asymmetric (AMP) model. With the hardware modeling language DML applied by Simics, multi-processor hardware simulation environment with Python language can be put up. Linux operating system is also installed in the simulation environment. At last, the correctness of putting up simulation multi-processor system by Simics is verified by the environment testing order, read and write CPUID, and application of CPU affinity function provided by Simics. Keyword: Simcis Simulator DML Multi-processor CPUID CPU Affinity目录 第一章 引言 1 1.1 研究背景 1 1.2 Simics的技术特点 2 1.2.1 容易扩展 2 1.2.2 支持回退机制 2 1.3 研究目标 3 第二章 Simics全系统模拟器的安装与调试