Measurement-Based Probabilistic Timing Analysis (MBPTA) has been shown to be an industrially viable method to\nestimate the Worst-Case Execution Time (WCET) of real-time programs running on processors including several\nhigh-performance features. MBPTA requires hardware/software support so that program�s execution time, and so its\nWCET, has a probabilistic behaviour and can be modelled with probabilistic and statistic methods. MBPTA also\nrequires that those events with high impact on execution time are properly captured in the (R) runs made at analysis\ntime. Thus, a representativeness argument is needed to provide evidence that those events have been captured.\nThis paper addresses the MBPTA representativeness problems caused by set-associative caches and presents a novel\nrepresentativeness validation method (ReVS) for cache placement. Building on cache simulation, ReVS explores the\nprobability and impact (miss count) of those cache placements that can occur during operation. ReVS determines the\nnumber of runs R, which can be higher than R, such that those cache placements with the highest impact are\neffectively observed in the analysis runs, and hence, MBPTA can be reliably applied to estimate the WCET.
Loading....