Semi-partitioning under a Blocking-Aware Task Allocation

—Semi-partitioned scheduling is a resource efficient scheduling approach compared to the conventional multiprocessor scheduling approaches in terms of system utilization and migration overhead. Semi-partitioned scheduling can better utilize processor bandwidth compared to the partitioned scheduling while introducing less overhead compared to the global scheduling. Various techniques have been proposed to schedule tasks in a semi-partitioned environment, however, they have used blockingagnostic allocation mechanisms in presence of resource sharing protocols. Since, the allocation mechanism can highly affect the system schedulability, in this paper we provide a blocking-aware allocation mechanism for semi-partitioned scheduling framework under a suspension-based resource sharing protocol. We have applied new heuristics for sorting the tasks in the algorithm that shows improvements upon system schedulability. Finally, we present our preliminary results.

• Type: Journal
• Year: 2015
• Where: RTSS
• Authors: Sara Afshar, Moris Behnam, Thomas Nolte
• Keywords: Control Systems, RTSS 2015