Improving the response time analysis of global fixed-priority multiprocessor scheduling

In this paper we address the problem of schedulability analysis for a set of sporadic tasks with arbitrary deadlines running on a multiprocessor system with global fixed-priority preemptive scheduling. We prove the existence of a class of critical instants for releasing a task, one of which results in the worst-case response time of that task. Then, we propose a new analysis that improves over current state-of-the-art Response Time Analysis (RTA) by reducing its pessimism. We also observe that, in the case of unconstrained deadlines, the current RTA may underestimate the carry-in workload, and we propose a new formulation that corrects the problem. Finally, we evaluate the performance improvement of our new response time analysis method by empirical experiments with randomly generated task sets. Experimental results show that our new analysis method can successfully accept a considerable amount of task sets that have to be treated as unschedulable by existing methods.