Ertss2010 multicore scheduling
0 likes355 views
1) The document discusses multicore scheduling challenges for automotive electronic control units (ECUs) as the automotive industry moves toward multicore architectures. 2) It presents a model for scheduling numerous tasks or "runnables" across multiple cores of an ECU. The model considers attributes of each runnable like period, worst case execution time, initial offset, and core allocation constraints. 3) The document proposes using load balancing algorithms like Least Loaded and variants to build schedule tables that assign runnables to slots on each core, aiming to balance load over time while avoiding peaks. It has implemented these approaches in a scheduling tool.
![9
Solution
Partitioning is dealt with as a bin packing problem
Worst fit decreasing algorithm with fixed number of bins
Load Balancing is done with the Least Loaded algorithm (LL)
inspired from CAN domain [Grenier and Navet ERTSS2008]
Extended to handle non harmonic runnable sets (G-LL)
Improved so as to reduce further load peaks (G-LLσ)
Implemented in a tool
Freely available soon at http://www.realtimeatwork.com](https://image.slidesharecdn.com/ertss2010multicorescheduling-111114030115-phpapp01/85/Ertss2010-multicore-scheduling-9-320.jpg)
Ertss2010 multicore scheduling
- 1. Multicore scheduling in automotive ECUs Aurélien Monot - PSA Peugeot Citroën, LORIA Nicolas Navet - INRIA, RealTime-at-Work Françoise Simonot - LORIA Bernard Bavoux - PSA Peugeot Citroën Talk at ERTS2 2010 Toulouse, May 21st 2010
- 2. 2 Outlook Context: New tools and methodologies are needed as multicore ECUs are being introduced in the automotive EE architecture. Problem: How to address the scheduling of numerous runnables on a multicore ECUs in the context of the automotive domain? Method: Deployment of load balancing algorithms in ECU configuration tools.
- 3. 3 The case of a generic car manufacturer Typical number of ECUs in a car in 2000 : 20 Typical number of ECUs in a car in 2010 : over 40 The number of ECUs has more than doubled in 10 years Other examples Between 60 and 80 ECUs in the Audi A8 Over 100 ECUs in some Lexus !
- 4. 4 Moving towards multicore architecture Decreasing the complexity of in-vehicle architecture: reduces EE design and verification efforts decreases number of network interfaces decreases traffic on CAN network reduces costs ECU 1 ECU 2 Tas Tas Tas k k k Tas Tas Tas k k k
- 5. 5 Moving towards multicore architecture Other use cases for the automotive domain Dealing with resource demanding applications ‣ engine control, image processing... Improving the safety ‣ segragation of multi-source software, ISO26262... Dedicated use of core ‣ monitoring, event-triggered tasks General benefits of multi-core reduced power consumption reduced heat reduced EMC
- 6. 6 AutoSAR requirements • Static partitioning • Static cyclic scheduling using schedule tables • BSW are all allocated on the same core
- 7. 7 Problem Goal: schedule numerous runnables on a multicore ECU Two sub-problems Partitioning ‣ 600 runnables on 2 cores : 2600 possible allocations Build schedule table ‣ 300 runnables in 200 slots (expiry points) : 200300 schedules Sub-objectives and criteria Avoid load peaks ‣ Max Balance load over time ‣ Standard Deviation
- 8. 8 Model Runnables P1=10 R P2=10 R1 C1=2 C2=1 2 •Period O1=0 O2=5 •WCET •Initial Offset P3=20 P4=20 R3 C3=3 R4 C4=2 •Core allocation constraint O3=5 O4=15 •Colocation constraint Sequencer task R R R R R1 R3 R1 R4 R1 R3 R1 R4 2 2 2 2 0 5 10 15 20 25 30 35 40 Ttic Slots Tcycle
- 9. 9 Solution Partitioning is dealt with as a bin packing problem Worst fit decreasing algorithm with fixed number of bins Load Balancing is done with the Least Loaded algorithm (LL) inspired from CAN domain [Grenier and Navet ERTSS2008] Extended to handle non harmonic runnable sets (G-LL) Improved so as to reduce further load peaks (G-LLσ) Implemented in a tool Freely available soon at http://www.realtimeatwork.com
- 11. 11 Harmonic task sets LL Max: 4.79 Min: 4.52 StdDvt: 0.038 G-LLσ Max: 4.75 Min: 4.65 StdDvt: 0.018 Generated load: 94%, Ttic=5ms, Tcycle = 1s
- 12. 12 Non harmonic task sets Cmax Schedulability bound in the harmonic case 1− Ttic Generated 95% 97% 95% 97% Max WCET (μs) 150 300 900 CPU load Schedulability 94% 82% Schedulability 97% 94% 82% bound in the Max WCET = Max WCET = 900μs bound in the harmonic case 300μs harmonic case Success % of 64% 18% 12% 1% Success % of LL 96% 96% 92% LL Success % of 94% 94% 30% 5% Success % of G-LL 100% 100% 100% G-LL Success % of 100% 100% 97% 76% G-LL1σ Statistics collected over 1000 generated runnable sets
- 13. 13 Multiple synchronized sequencer tasks per core Incremental scheduling of three synchronized sequencer tasks with respective load of 45%, 35% and 15% resulting in 95% of the core capacity. Tcycle=1000ms and Ttic=5ms
- 14. 14 Multiple non synchronized sequencer tasks per core Case arises for sequencer tasks using different tic counters Engine control applications (standard time vs RPM) Any offset between the sequencer tasks and all clock rates are possible during runtime each sequencer task needs to be balanced independantly Verification is possible considering maximum clock rates Multi-frame scheduling results can be used
- 15. 15 Conclusion Adoption of multicore ECU raises new challenges Evolution of software architecture design Scheduling of software components We propose runnable scheduling heuristics for ECUs Fast and performant Easily adaptable for more advanced applications Compatible with AutoSAR R4.0 and its multicore extensions Future work Precedence constraints Lockstep synchronization Distributed timing chains
- 16. 16 Thank you for your attention