Schedulability analysis and stack size minimization for adaptive mixed criticality scheduling with semi-Clairvoyance and preemption thresholds
Section snippets
Introduction and related work
Certain application domains, e.g., Automotive Electrical/Electronic (E/E) systems, are very resource-constrained due to size, weight and power (SWaP) constraints, as well as monetary cost constraints thanks to intense cost-cutting pressures in today’s competitive consumer marketplace. Auto-makers are very cost-sensitive due to the mass production of vehicles. As a modern high-end vehicle may contain tens to 100+ Electronic Control Units (ECUs) in it, and each vehicle model may sell millions of
PT-AMC-sem and its schedulability test
We consider a mixed-criticality task set with independent sporadic real-time tasks on a uniprocessor. Based on Vestal’s task model [13], each task is given multiple Worst-Case Execution Time (WCET) estimates, one for each criticality level. We consider a dual-criticality MCS with two criticality levels: Low (LO) and High (HI). Task has a tight, optimistic WCET estimate in Low-Criticality (abbreviated as LO-crit) mode that may be occasionally exceeded, and a loose,
Illustrating example
We use an example to illustrate the schedulability tests. Consider a task set with three tasks, with the parameters shown in Table 1. To apply the analysis of PT-AMC-sem, we have assigned priorities and PTs to tasks in the running example. Let us consider HI-crit task as the task under analysis. According to Eq. (1), it is schedulable if , where .
First, let us compute ’s WCRT in LO-crit mode. We have , , ,
Performance evaluation
We use the same task set generation methodology as [24] to generate synthetic task sets for our experiments, with the following parameters:
- •
Each task ’s utilization in LO-crit mode() is generated using UUnifast [28], providing an unbiased distribution, with default task set utilization in LO-crit mode set to 0.85.
- •
Each task ’s period is generated according to a log-uniform distribution [29] in the interval [10,200] ms.
- •
Each task ’s deadline is set to be equal to its period
Conclusions and future work
In this paper, we present a schedulability test to enable integration of Preemption Threshold Scheduling (PTS) with Mixed-Criticality Scheduling with Semi-Clairvoyance. Performance evaluation indicates that our approach results in significant benefits in terms of improving schedulability and reducing system stack usage compared to Fully Preemptive Scheduling, which helps to implement MCS more efficiently in resource-constrained embedded systems. As part of future work, we plan to consider
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgment
This work was supported by the National Natural Science Foundation of China (Grant No. 61902185 and 61877015), the Jiangsu Provincial Natural Science Foundation (Grant No. BK20190448 and BK20190447), and “111” Program (No. B13022).
Qingling Zhao received the Ph.D. degree from Zhejiang University(2015). She is currently an associate professor in the school of computer science and engineering at the Nanjing University of Science and Technology. Her research interests include real-time systems, mixed-criticality systems, and cyber-physical systems.
References (31)
Efficient exact Boolean schedulability tests for fixed priority preemption threshold scheduling
J. Syst. Softw.
(2017)- et al.
Schedulability analysis and stack size minimization with preemption thresholds and mixed-criticality scheduling
J. Syst. Archit.
(2018) - et al.
Design optimization for AUTOSAR models with preemption thresholds and mixed-criticality scheduling
J. Syst. Archit.
(2017) - et al.
Scheduling fixed-priority tasks with preemption threshold
- et al.
Preemption threshold scheduling: Stack optimality, enhancements and analysis
- et al.
Optimal priority and threshold assignment for fixed-priority preemption threshold scheduling
ACM Sigbed Rev.
(2018) - et al.
Minimizing stack and communication memory usage in real-time embedded applications
ACM Trans. Embed. Comput. Syst.
(2014) Assigning priorities for fixed priority preemption threshold scheduling
Sci. World J.
(2015)- U. Keskin, R.J. Bril, J.J. Lukkien, Exact response-time analysis for fixed-priority preemption-threshold scheduling,...
- H.J. Rivera-Verduzco, R.J. Bril, Best-case response times of real-time tasks under fixed-priority scheduling with...
Global fixed priority scheduling with preemption threshold: Schedulability analysis and stack size minimization
IEEE Trans. Parallel Distrib. Syst.
Minimizing stack memory for hard real-time applications on multicore platforms with partitioned fixed-priority or EDF scheduling
ACM Trans. Design Automa. Electron. Syst. (TODAES)
Cache-partitioned preemption threshold scheduling
ACM Trans. Embed. Comput. Syst.
Fixed priority scheduling with pre-emption thresholds and cache-related pre-emption delays: integrated analysis and evaluation
Real-Time Syst.
Preemptive scheduling of multi-criticality systems with varying degrees of execution time assurance
Cited by (5)
Towards an energy-efficient quarter-clairvoyant mixed-criticality system
2022, Journal of Systems ArchitectureCitation Excerpt :These frameworks, which trigger a mode switch when they detect over-execution of a Hi-task, are also called “non-clairvoyant MCSs” [12,13]. Different from non-clairvoyant MCSs, Baruah et al. [12] and Agrawal et al. [13] introduce clairvoyant and semi-clairvoyant MCS theoretical models, which assume that whether a Hi-task will overrun its Lo-WCET is known before or at release [14,15]. As shown in [12,13], both clairvoyant and semi-clairvoyant MCSs outperform non-clairvoyant MCSs.
Criticality-Aware EDF Scheduling for Constrained-Deadline Imprecise Mixed-Criticality Systems
2024, IEEE Transactions on Computer-Aided Design of Integrated Circuits and SystemsEfficient Schedulability Analysis of Semi-Clairvoyant Sporadic Task Systems With Graceful Degradation
2022, ACM International Conference Proceeding SeriesCompensating Adaptive Mixed Criticality Scheduling
2022, ACM International Conference Proceeding SeriesMixed Criticality on Multi-cores Accounting for Resource Stress and Resource Sensitivity
2022, ACM International Conference Proceeding Series
Qingling Zhao received the Ph.D. degree from Zhejiang University(2015). She is currently an associate professor in the school of computer science and engineering at the Nanjing University of Science and Technology. Her research interests include real-time systems, mixed-criticality systems, and cyber-physical systems.
Mengfei Qu He is currently a master graduate student in the school of computer science and engineering at the Nanjing University of Science and Technology. His research interests include mixed criticality system, embedded system, etc.
Bo Huang received the Ph.D. degree from Nanjing University of Science and Technology (2006). He is currently a professor in the school of computer science and engineering at Nanjing University of Science and Technology. His research interests include discrete event systems, Petri nets, system modeling and intelligent scheduling.
Zhe Jiang received his Ph.D. from University of York (2019). He is currently working as the system design engineer in ARM Ltd and visit research associate in University of York. He is research int rests include safety-critical system, system architecture, and system micro-architecture. He can be reached at: [email protected].
Haibo Zeng (Member, IEEE) is with Department of Electrical and Computer Engineering at Virginia Tech. He received his Ph.D. in Electrical Engineering and Computer Sciences from University of California at Berkeley. He was a Senior Researcher at General Motors R&D until October 2011, and an Assistant Professor at McGill University until August 2014. His research interests are embedded systems, cyber-physical systems, and real-time systems. He received five best / outstanding paper awards in the above fields.
- 1
Co-first authors of the article.