liu.seSearch for publications in DiVA
Change search
ReferencesLink to record
Permanent link

Direct link
Cache-Related Preemption Delay Analysis for Multilevel Noninclusive Caches
Linköping University, Department of Computer and Information Science, Software and Systems. Linköping University, The Institute of Technology.
National University of Singapore, Singapore.
2014 (English)In: ACM Transactions on Embedded Computing Systems, ISSN 1539-9087, E-ISSN 1558-3465, Vol. 13, no 147Article in journal (Refereed) Published
Abstract [en]

With the rapid growth of complex hardware features, timing analysis has become an increasingly difficult problem. The key to solving this problem lies in the precise and scalable modeling of performance-enhancing processor features (e.g., cache). Moreover, real-time systems are often multitasking and use preemptive scheduling, with fixed or dynamic priority assignment. For such systems, cache related preemption delay (CRPD) may increase the execution time of a task. Therefore, CRPD may affect the overall schedulability analysis. Existing works propose to bound the value of CRPD in a single-level cache. In this article, we propose a CRPD analysis framework that can be used for a two-level, noninclusive cache hierarchy. In addition, our proposed framework is also applicable in the presence of shared caches. We first show that CRPD analysis faces several new challenges in the presence of a multilevel, noninclusive cache hierarchy. Our proposed framework overcomes all such challenges and we can formally prove the correctness of our framework. We have performed experiments with several subject programs, including an unmanned aerial vehicle (UAV) controller and an in-situ space debris monitoring instrument. Our experimental results suggest that we can provide sound and precise CRPD estimates using our framework.

Place, publisher, year, edition, pages
Association for Computing Machinery (ACM) , 2014. Vol. 13, no 147
Keyword [en]
Design; Performance; Verification; Cache-related preemption delay; WCET; multilevel caches; multicore; shared caches
National Category
Computer and Information Science
URN: urn:nbn:se:liu:diva-113373DOI: 10.1145/2632156ISI: 000346416300003OAI: diva2:781557

Funding Agencies|A*STAR Public Sector Funding "Scalable Timing Analysis Methods for Embedded Software" [1121202007]

Available from: 2015-01-16 Created: 2015-01-16 Last updated: 2015-01-16

Open Access in DiVA

No full text

Other links

Publisher's full text

Search in DiVA

By author/editor
Chattopadhyay, Sudipta
By organisation
Software and SystemsThe Institute of Technology
In the same journal
ACM Transactions on Embedded Computing Systems
Computer and Information Science

Search outside of DiVA

GoogleGoogle Scholar
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

Altmetric score

Total: 28 hits
ReferencesLink to record
Permanent link

Direct link