Experimental Evaluation of Code Properties for WCET Analysis

  • Authors:
  • Antoine Colin;Stefan M. Petters

  • Affiliations:
  • -;-

  • Venue:
  • RTSS '03 Proceedings of the 24th IEEE International Real-Time Systems Symposium
  • Year:
  • 2003

Quantified Score

Hi-index 0.00

Visualization

Abstract

This paper presents a quantification of the timing effectsthat advanced processor features like data and instructioncache, pipelines, branch prediction units and out-of-orderexecution units have on the worst-case execution time(WCET) of programs. These features are present in processors(e.g. PowerPC) that are being widely used in embeddedand real-time systems. We present an experimental evaluationof the execution time of a series of synthetic benchmarksand real-life case studies. The execution time is evaluatedusing extensive testing and a simple WCET technique.We show that the most important factor in reduction of executiontime is cache size (both instruction and data cache).Other factors like branch prediction and out-or-order executionhave minimal improvements that is cancelled out bythe pessimism of the analysis. We also argue that some of theperformance gain of advanced processor features also appliesto the worst case and although WCET estimates maybe more pessimistic the overall impact is that they result inlower WCET estimates.