Multiprocessor Real-Time Systems with Shared Resources: Utilization Bound and Mapping

Abstract

Abstract-In real-time systems, both scheduling theory and resource access protocols have been studied extensively. However, there is very limited research focusing on scheduling algorithms specifically for real-time systems with shared resources, where the problem becomes more prominent with the emergence of multicore processors. In this paper, focusing on the partitioned-EDF scheduling with the MSRP resource access protocol, we study the utilization bound and efficient task mapping schemes for a set of periodic real-time tasks running on a multicore/multiprocessor system with shared resources. Specifically, we first illustrate the scheduling anomaly where the tasks are schedulable when being mapped on to fewer but not more processors due to synchronization overhead. Then, with such synchronization overhead being considered, we develop a synchronization-cognizant utilization bound. Moreover, we show that finding the optimal mapping of tasks with shared resources is NP-hard and propose two efficient synchronization-cognizant task mapping algorithms (SC-TMA) that rely on the new tightened synchronization overhead and have the goal of achieving better schedulability and balanced workload on deployed processors. Finally, the proposed SC-TMA schemes are evaluated through extensive simulations with synthetic tasks. The results show that, the schedulability ratio and (average) system load under SC-TMA are close to that of an INLP (Integer Non-Linear Programming) based solution for small task systems. When compared to the existing task mapping algorithms, SC-TMA obtain much better schedulability ratio and lower/balanced workload on all processors.

Resource-Oriented Partitioning for Multiprocessor Systems with Shared Resources

“…More recently, Wieder and Brandenburg [43] proposed a greedy slacker partitioning heuristic with the MSRP. Han et al [27], [28] addressed the problem of utilization bounds for P-EDF with the MSRP and proposed the synchronization-cognizant task mapping algorithms [28]. However, no algorithmic evidence was provided to necessitate these approaches from the perspective of resource augmentation.…”

“…Suppose that task set t is schedulable under R-PCPrm-rm when each processor is at least x times as fast. Then, summing over the corresponding terms in inequalities ( 25), ( 26), (27), and (28) and to contradict to inequality (24)…”

Triple Speed: Energy-Aware Real-Time Task Synchronization in Homogeneous Multi-Core Systems

“…Hsiu et al [25] proposed a dedicated-core framework to adequately manage blocking. Han et al [26] proposed a synchronizationcognizant task mapping algorithm to realize optimal schedulability and workload balancing. To reduce synchronization overheads, previous studies have assigned tasks to the same core according to accessed resources [10], [28].…”

“…Previous studies have extended the stack resource policy (SRP) [21] and priority ceiling protocol (PCP) [22] for multiprocessor systems to reduce the remote blocking resulting from resource sharing among different cores [18], [19], namely multiprocessor SRP (MSRP) and multiprocessor PCP (MPCP), respectively. Numerous studies have been conducted for investigating synchronization-aware task allocation to enhance the schedulability of tasks with resource sharing in multicore systems [10], [23], [24], [25], [26], [27], [28]. Although the effect of remote blocking on schedulability has been adequately addressed in the mentioned studies, the requirement of global resource sharing cannot be avoided.…”

A Correlation Framework for different Resource Access Protocol in Real Time System

“…In the lock-free approach [1,9], operations on simple software objects, e.g., stacks, linked lists, are performed by retry loops, i.e., operations are retried until the object is accessed successfully. The advantages of lock-free algorithm are that they do not require kernel support and as there is no need to lock, priority inversion does not occur.…”