Boosting Job-Level Migration by Static Analysis

From an operating system’s perspective, task migration is a potent instrument to exploit multi-core processors. Like full preemption, full migration is particularly advantageous as it allows the scheduler to relocate tasks at arbitrary times between cores. However, in real-time systems, migration is accompanied by a tremendous drawback: poor predictability and thus inevitable overapproximations in the worst-case executiontime analysis. This is due to the non-constant size of the tasks’ resident set and the costs associated with its transfer between cores. As a result, migration is banned in many real-time systems, regressing the developer to a static allocation of tasks to cores with disadvantageous effects on the overall utilization and schedulability. In this paper, we tackle the shortcomings of full migration in real-time systems by reducing the associated costs and increasing its predictability. Our approach is to analyze a task’s source code to identify beneficial migration points considering the size of scheduling units and the associated migration costs. Consequently, we can do both: generate schedules that benefit from static migration as well as provide information about advantageous migration points to dynamic scheduling, making full migration more predictable. Our experiments show improved schedulability and a reduction in transfer size of up to 76 percent.