ePrints@IIScePrints@IISc Home | About | Browse | Latest Additions | Advanced Search | Contact | Help

Fast and Precise Worst-Case Interference Placement for Shared Cache Analysis

Nagar, Kartik and Srikant, YN (2016) Fast and Precise Worst-Case Interference Placement for Shared Cache Analysis. In: ACM TRANSACTIONS ON EMBEDDED COMPUTING SYSTEMS, 15 (3).

[img] PDF
ACM_Tra_Emb_Com_Sys_15-3_45_2016.pdf - Published Version
Restricted to Registered users only

Download (794kB) | Request a copy
Official URL: http://dx.doi.org/10.1145/2854151


Real-time systems require a safe and precise estimate of the worst-case execution time (WCET) of programs. In multicore architectures, the precision of a program's WCET estimate highly depends on the precision of its predicted shared cache behavior. Prediction of shared cache behavior is difficult due to the uncertain timing of interfering shared cache accesses made by programs running on other cores. Given the assignment of programs to cores, the worst-case interference placement (WCIP) technique tries to find the worst-case timing of interfering accesses, which would cause the maximum number of cache misses on the worst case path of the program, to determine its WCET. Although WCIP generates highly precise WCET estimates, the current ILP-based approach is also known to have very high analysis time. In this work, we investigate the WCIP problem in detail and determine its source of hardness. We show that performing WCIP is an NP-hard problem by reducing the 0-1 knapsack problem. We use this observation to make simplifying assumptions, which make the WCIP problem tractable, and we propose an approximate greedy technique for WCIP, whose time complexity is linear in the size of the program. We perform extensive experiments to show that the assumptions do not affect the precision of WCIP but result in significant reduction of analysis time.

Item Type: Journal Article
Additional Information: Copy right for this article belongs to the ASSOC COMPUTING MACHINERY, 2 PENN PLAZA, STE 701, NEW YORK, NY 10121-0701 USA
Department/Centre: Division of Electrical Sciences > Computer Science & Automation
Depositing User: Id for Latest eprints
Date Deposited: 22 Oct 2016 09:27
Last Modified: 22 Oct 2016 09:27
URI: http://eprints.iisc.ac.in/id/eprint/55031

Actions (login required)

View Item View Item