Locking in pfair-scheduled multiprocessor systems

Abstract

The goal of this dissertation is to extend the Pfair scheduling approach in order to enable its efficient implementation on a real-time multiprocessor. At present, Pfair scheduling is the only known means for optimally scheduling recurrent real-time tasks on multiprocessors. In addition, there has been growing practical interest in such approaches due to their fairness guarantees. Unfortunately, prior work in this area has considered only the scheduling of independent tasks, which do not communicate with each other or share resources. The work presented herein focuses both on adding support for these actions and also on developing techniques for reducing various forms of implementation overhead, including that produced by task migrations and context switches. The thesis of this dissertation is: tasks can be efficiently synchronized in Pfair-scheduled systems and overhead due to common system events, such as context switches and migrations, can be reduced.This thesis is established through research in three areas. First, the pre-existing Pfair scheduling theory is extended to support the scheduling of groups of tasks as a single entity. Second, mechanisms for supporting both lock-based and lock-free synchronization are presented. Lockbased synchronization coordinates access to shared resources through the use of semaphores.On the other hand, lock-free operations are optimistically attempted and then retried if the operation fails. Last, the deployment of Pfair scheduling on a symmetric multiprocessor is considered.

Synchronization in Embedded Real-Time Operating Systems

“…They allow FIFO-queue based access to locked resources and present different techniques for handling short and long critical sections [37].…”

Efficient Synchronization under Global EDF Scheduling on Multiprocessors

“…By (8), in PS τ , the total allocation to Ψ in [t, t ) cannot exceed (t − t) · U sum (τ ). By (14) and Part (ii) of Claim 2, at most m − k jobs of Ψ execute at any instant in [t, t ).…”

“…In [14], an extension of the priority-ceiling protocol, which is for use with semaphores, has been presented for partitioned rate-monotonic scheduling. Semaphore-based approaches and the lock-free approach have been considered for Pfair-scheduled systems in [9] and [8], respectively. For use in multiprocessor-based real-time kernels, an approach based on spin locks has been proposed in [15].…”

Desynchronized Pfair Scheduling on Multiprocessors

“…One such limitation is the requirement that tasks be allocated processor time in fixed-sized quanta that align across all processors. It is known that if this requirement is not satisfied, then deadlines can be missed under an otherwiseoptimal Pfair scheduling algorithm [7]. Fig.…”

“…This idling can be reduced if the quanta are desynchronized and each processor scheduled independently. Fourth, synchronization protocols that have been proposed may cause a task to block on a lock request in the middle of a quantum [7]. Under the SFQ model, the remainder of the quantum is wasted.…”