標題: 以內容觸發降低嵌入式行動裝置上虛擬機器垂直性能分析之負擔
Low Overhead Context-sensitive Vertical Profiling for Virtual Machine Based Mobile Embedded Devices
作者: 蘇資翔
Su, Tzu-hsiang
陳添福
Chen, Tien-Fu
資訊學院資訊學程
關鍵字: 垂直性能分析;Dalvik虛擬機器;Java;Android;記憶體通道;Dalvik插入測試模組;Vertical Profiling;Dalvik Virtual Machine;Java;Android;Direct Memory Tunnel;Dalvik Instrumentation Module
公開日期: 2011
摘要: Android系統採用了Dalvik應用層虛擬機器以便支援各種處理器硬體架構。但是使用Dalvik虛擬機器卻在應用虛擬機器層與Linux核心層間產生了Android上現有效能分析工具無法跨越的障礙。這道障礙使得現有效能分析工具無法有效判斷應用程式的執行瓶頸究竟來自於應用層的程式碼演算法、Linux使用者空間函式庫、或是更底層的Linux核心與驅動。本論文提出利用Dalvik虛擬機器的架構特色與虛擬記憶體替換通道技術達成不導入大量負擔即可克服訊息傳遞障礙的可調鎖定觸發機制垂直性能分析平台。這個性能分析平台可以簡化Android系統上的軟硬體協同設計過程。
Dalvik virtual machine in the Android system creates a profiling barrier between virtual machine applications and Linux user-space libraries. This barrier makes it difficult for existing profiling tools on the Android system to definitively identify whether a bottleneck occurred in application level, Linux user-space level, or Linux kernel level. The proposed the Reconfigurable Vertical Profiling Framework leverages the unique features of Dalvik runtime virtual machine to bridges the gap without introducing large overhead. This profiling framework streamlines the hardware-software co-design process for the Android runtime system.
URI: http://140.113.39.130/cdrfb3/record/nctu/#GT079679521
http://hdl.handle.net/11536/44071
Appears in Collections:Thesis


Files in This Item:

  1. 952101.pdf