An analysis of random cache effects on real-time multi-core scheduling algorithms

The effect of sharing the last-level cache (LLC) among cores in a multi-core system has not been thoroughly investigated especially in the design of efficient scheduling algorithms. And with the growing interest in random caches, which allow for an easier estimation of the worst-case execution time of tasks in critical real-time embedded systems, tools that analyse the sensitivity of workloads to sharing the LLC become necessary. In this paper, we extend a realtime multiprocessor scheduling simulator, SimSo, with a framework that incorporates a random cache model for multi-level caches to evaluate emerging scheduling algorithms under the influence of shared caches. A set of experiments were performed to study the behavior of workloads with respect to worst-case response time, average slack time, and maximum utilization, with varying cache designs under different scheduling algorithms.