Cache-Persistence-Aware Response-Time Analysis for Fixed-Priority Preemptive Systems

Abstract

A task can be preempted by several jobs of higherpriority tasks during its response time. Assuming the worstcasememory demand for each of these jobs leads to pessimistic worstcase response time (WCRT) estimations. Indeed, there is a bigchance that a large portion of the instructions and data associatedwith the preempting task τj are still available in the cache when τj releases its next jobs. Accounting for this observation allowsthe pessimism of WCRT analysis to be significantly reduced,which is not considered by existing work.The four main contributions of this paper are: 1) The conceptof persistent cache blocks is introduced in the context of WCRT analysis, which allows re-use of cache blocks to be captured, 2) A cache-persistence-aware WCRT analysis for fixed-prioritypreemptive systems exploiting the PCBs to reduce the WCRTbound, 3) An multi-set extension of the analysis that furtherimproves the WCRT bound, and 4) An evaluation showing thatour cachepersistence-aware WCRT analysis results in up to 10% higher schedulability than state-of-the-art approaches. Abstract-A task can be preempted by several jobs of higher priority tasks during its response time. Assuming the worst-case memory demand for each of these jobs leads to pessimistic worstcase response time (WCRT) estimations. Indeed, there is a big chance that a large portion of the instructions and data associated with the preempting task τj are still available in the cache when τj releases its next jobs. Accounting for this observation allows the pessimism of WCRT analysis to be significantly reduced, which is not considered by existing work.The four main contributions of this paper are: 1) The concept of persistent cache blocks is introduced in the context of WCRT analysis, which allows re-use of cache blocks to be captured, 2) A cache-persistence-aware WCRT analysis for fixed-priority preemptive systems exploiting the PCBs to reduce the WCRT bound, 3) An multi-set extension of the analysis that further improves the WCRT bound, and 4) An evaluation showing that our cache-persistence-aware WCRT analysis results in up to 10% higher schedulability than state-of-the-art approaches.

Integrated Analysis of Cache Related Preemption Delays and Cache Persistence Reload Overheads

“…The notion of cache persistence and CPRO was recently introduced in [21]. Methods to compute the CPRO cost and to integrate it in the WCRT analysis for FPPS were proposed, showing significant improvement in the accuracy of the response time analysis.…”

“…Their analysis considers both the CRPD and CPRO and dominates the stateof-the-art approaches that only consider CRPD. However, the analysis in [21] may sometimes result in over-estimation of the task response times. This is due to the fact that CRPD and CPRO are calculated separately, providing independent upper bounds on the two classes of overheads.…”

“…These cache blocks are called Persistent Cache Blocks (PCBs) and this concept is referred to as cache persistence 1 [21]. Analysis of cache persistence can be used to reduce pessimism in the computation of interference from multiple jobs of a higher priority task in state-of-the-art worst-case response time (WCRT) analysis for systems using Fixed Priority Preemptive Scheduling (FPPS).…”

“…In [21], the authors derived two analyses for CPRO that were integrated into an improved response time analysis for FPPS that takes account of reductions in memory demand due to cache persistence, along with the CRPD. Their analysis considers both the CRPD and CPRO and dominates the stateof-the-art approaches that only consider CRPD.…”

Timing Predictability in Future Multi-Core Avionics Systems

“…This is an extensively studied issue (e.g., [7,9,20,55,76]). On a multi-core system on the other hand, requests can be issued at the same time and may have to be arbitrated by the hardware somehow.…”