Bachelor and Master Theses

Title: A Data Access tracking approach for Memory analysis and Bad data access pattern detecting
Subject: Computer Science
Level: Advanced
Description: Data access tracking is the approach to reveal memory behavior in run-time, and is widely used in the memory related optimize analysis and bad data access pattern detecting. However most data access tracking approach is realized by code instrumentation, which will lead to extra performance cost of target system. In this paper, I present a new data access tracking approach: via interaction with instruction set simulator, data access trace is recorded in a non-intrusive way. By analyzing the data access trace, I also present the approaches for detecting some bad data access patterns. A proposal for detecting cache related bad patterns is also presented in the discussion part.
Company: Ericsson AB, kontaktperson: Marcus Jägemar
Prel. end date: 2012-06-14
Presentation date: 2012-06-13
Student: Li Jiang jg11001@student.mdh.se
IDT supervisor: Daniel Flemström
daniel.flemstrom@mdh.se, +46-21-151760
Examinator: Mikael Sjödin
Mikael Sjödin
mikael.sjodin@mdh.se, +46 70 288 2829

  • Mälardalen University |
  • Box 883 |
  • 721 23 Västerås/Eskilstuna |
  • 021-101300, 016-153600 |
  • webmaster |
  • Latest update: 2017.10.14