Preemption Point Selection in Limited Preemptive Scheduling using Probabilistic Preemption Costs

Transcription

1 Preemption Point Selection in Limited Preemptive Scheduling using Probabilistic Preemption Costs Filip Marković, Jan Carlson, Radu Dobrin Mälardalen Real-Time Research Centre, Dept. of Computer Science and Software Engineering, Mälardalen University, Sweden

2 Limited Preemptive Scheduling

3 Limited Preemptive Scheduling An attractive scheduling paradigm instead of fully-preemptive and non-preemptive scheduling. Enables control of preemption related overheads, thus reducing their impact on schedulability.

4 Limited Preemptive Scheduling An attractive scheduling paradigm instead of fully-preemptive and non-preemptive scheduling. Enables control of preemption related overheads, thus reducing their impact on schedulability. Fixed Preemption Points Preemption is allowed only at predefined selected locations inside the code, called preemption points.

5 Limited Preemptive Scheduling An attractive scheduling paradigm instead of fully-preemptive and non-preemptive scheduling. Enables control of preemption related overheads, thus reducing their impact on schedulability. Fixed Preemption Points Preemption is allowed only at predefined selected locations inside the code, called preemption points. higher P τ " lower P preemption point

6 Limited Preemptive Scheduling An attractive scheduling paradigm instead of fully-preemptive and non-preemptive scheduling. Enables control of preemption related overheads, thus reducing their impact on schedulability. Fixed Preemption Points Preemption is allowed only at predefined selected locations inside the code, called preemption points. higher P τ " lower P preemption point

7 Limited Preemptive Scheduling An attractive scheduling paradigm instead of fully-preemptive and non-preemptive scheduling. Enables control of preemption related overheads, thus reducing their impact on schedulability. Fixed Preemption Points Preemption is allowed only at predefined selected locations inside the code, called preemption points. higher P τ " lower P preemption point

8 Limited Preemptive Scheduling An attractive scheduling paradigm instead of fully-preemptive and non-preemptive scheduling. Enables control of preemption related overheads, thus reducing their impact on schedulability. Fixed Preemption Points Preemption is allowed only at predefined selected locations inside the code, called preemption points. higher P τ " lower P preemption point

9 Limited Preemptive Scheduling An attractive scheduling paradigm instead of fully-preemptive and non-preemptive scheduling. Enables control of preemption related overheads, thus reducing their impact on schedulability. Fixed Preemption Points Preemption is allowed only at predefined selected locations inside the code, called preemption points. higher P τ " lower P preemption point

10 Limited Preemptive Scheduling An attractive scheduling paradigm instead of fully-preemptive and non-preemptive scheduling. Enables control of preemption related overheads, thus reducing their impact on schedulability. Fixed Preemption Points Preemption is allowed only at predefined selected locations inside the code, called preemption points. higher P τ " lower P preemption point

11 Motivation The existing selection methods account for upper bounded preemption overheads, thus introducing a potentially high level of pessimism in the results.

12 Motivation The existing selection methods account for upper bounded preemption overheads, thus introducing a potentially high level of pessimism in the results. higher P τ " lower P Preemption overhead

13 Motivation The existing selection methods account for upper bounded preemption overheads, thus introducing a potentially high level of pessimism in the results. higher P τ " lower P Preemption overhead

14 Motivation The existing selection methods account for upper bounded preemption overheads, thus introducing a potentially high level of pessimism in the results. higher P τ " preemption overhead lower P Preemption overhead

15 Motivation The existing selection methods account for upper bounded preemption overheads, thus introducing a potentially high level of pessimism in the results. higher P τ " lower P preemption overhead deadline miss Preemption overhead

16 Motivation The existing selection methods account for upper bounded preemption overheads, thus introducing a potentially high level of pessimism in the results. higher P τ " lower P preemption overhead deadline miss Preemption overhead Can we reduce the pessimism by considering probabilistic information about overheads?

17 Contributions We propose a probabilistic distribution model of overheads and preemption point selection method which provides controllable probabilistic relaxations.

18 Contributions We propose a probabilistic distribution model of overheads and preemption point selection method which provides controllable probabilistic relaxations.

19 Contributions We propose a probabilistic distribution model of overheads and preemption point selection method which provides controllable probabilistic relaxations. Preemption overhead

20 Contributions We propose a probabilistic distribution model of overheads and preemption point selection method which provides controllable probabilistic relaxations. upper bound preemption overhead

21 Contributions We propose a probabilistic distribution model of overheads and preemption point selection method which provides controllable probabilistic relaxations. upper bound empirical samples of preemption overheads preemption overhead

22 Contributions We propose a probabilistic distribution model of overheads and preemption point selection method which provides controllable probabilistic relaxations. probability probability density function upper bound empirical samples of preemption overheads preemption overhead

23 Preemption Point Selection Algorithm

24 Preemption Point Selection Algorithm Input Task set with potential preemption points Associated probabilistic overhead distributions

25 Preemption Point Selection Algorithm Input Task set with potential preemption points Associated probabilistic overhead distributions Output Selected preemption points

26 Preemption Point Selection Algorithm Input Task set with potential preemption points Associated probabilistic overhead distributions Output Selected preemption points Algorithm Gradually decreases probabilistic factor for preemption overheads in order to find preemption point selection

27 Preemption Point Selection Algorithm iteration Input Task set with potential preemption points 1 Associated probabilistic overhead distributions. Output Selected preemption points Algorithm 2 focused overhead Gradually decreases probabilistic factor for preemption overheads in order to find preemption point 3 selection. implies selection of different points probability of a deadline miss (part of the future work)

28 Preemption Point Selection Algorithm iteration Input Task set with potential preemption points 1 Associated probabilistic overhead distributions. Output Selected preemption points Algorithm 2 focused overhead Gradually decreases probabilistic factor for preemption overheads in order to find preemption point 3 selection. implies selection of different points probability of a deadline miss (part of the future work)

29 Preemption Point Selection Algorithm iteration Input Task set with potential preemption points 1 Associated probabilistic overhead distributions. Output Selected preemption points Algorithm 2 focused overhead Gradually decreases probabilistic factor for preemption overheads in order to find preemption point 3 selection. implies selection of different points probability of a deadline miss (part of the future work)

30 Preliminary results Goal of the experiment: To investigate to what extent the relaxation of the considered overheads facilitates finding solutions to the preemption point selection problem. Task sets for which a selection is found (%) Upper bounds Quantile selection Utilisation

31 Summary and Future work Contributions Probabilistic overhead model Preemption point selection based on probabilistic overhead distributions Future work Probabilistic schedulability analysis techniques for tasks with fixed preemption points and associated probabilistic overheads Novel preemption point selection strategies to maximize schedulability