ABSTRACT
The Worst-Case Execution Time (WCET), energy consumption, and code size are among the most important criteria of hard real-time systems. To estimate the WCET and energy consumption at compile time, static analyzers are often used: they estimate the objectives by invoking time-consuming microarchitecture, data flow, and control flow analyses. The expensive analyses make it almost infeasible to use evolutionary algorithms for solving multiobjective problems with these two objectives at compile time, since any evolutionary algorithm extensively evaluates objectives to find solutions. We propose a method that speeds up an evolutionary algorithm supplying it with a reduced search space and prediction model fitted on the reduced search space, so the algorithm needs to explore a smaller search space and can use fast predictions instead of time-consuming estimations to evaluate the WCET and energy consumption. The proposed approach is general enough to be used for any compiler-based optimization. We demonstrate the advantages of it solving a multiobjective function inlining problem at compile time.
- AbsInt Angewandte Informatik, GmbH. 2021. aiT Worst-Case Execution Time Analyzers.Google Scholar
- P. Altenbernd, J. Gustafsson, B. Lisper, and F. Stappert. 2016. Early execution time-estimation through automatically generated timing models. 52, 6 (2016), 731-760. https://doi.org/10.1007/s11241-016-9250-7 Google ScholarDigital Library
- A. Bonenfant, D. Claraz, M. de Michiel, and P. Sotin. 2017. Early WCET Prediction Using Machine Learning. In International Workshop on Worst-Case Execution Time Analysis, Vol. 57. 5:1-5:9. https://doi.org/10.4230/OASIcs.WCET.2017.5Google Scholar
- C. C. Coello, G.B. Lamont, and D. A. van Veldhuizen. 2007. Evolutionary Algorithms for Solving Multi-Objective Problems. Springer-Verlag GmbH. https://www.ebook.de/de/product/19293629/carlos_coello_coello_gary_b_lamont_david_a_van_veldhuizen_evolutionary_algorithms_for_solving_multi_objective_problems.html Google ScholarDigital Library
- A. M. C. Demetrios, D. De Sensi, A. F. Lorenzon, K. Georgiou, J. Nunez-Yanez, K. Eder, and S. Xavier de Souza. [n.d.]. Performance and Energy Trade-Offs for Parallel Applications on Heterogeneous Multi-Processing Systems. 13, 9 ([n. d.]), 2409. https://doi.org/10.3390/en13092409Google Scholar
- H. Falk, S. Altmeyer, P. Hellinckx, B. Lisper, W. Puffitsch, C. Rochange, M. Schoeberl, R. B. Sørensen, P. Wägemann, and S. Wegener. 2016. TACLeBench: A Benchmark Collection to Support Worst-Case Execution Time Research. In International Workshop on Worst-Case Execution Time Analysis (2016), Vol. 55. 2:1-2:10. https://doi.org/10.4230/OASICS.WCET.2016.2Google Scholar
- H. Falk and P. Lokuciejewski. 2010. A compiler framework for the reduction of worst-case execution times. Real-Time Systems 46, 2 (2010), 251-300. https://doi.org/10.1007/s11241-010-9101-x Google ScholarDigital Library
- T. Huybrechts, S. Mercelis, and P. Hellinckx. 2018. A New Hybrid Approach on WCET Analysis for Real-Time Systems Using Machine Learning. In International Workshop on Worst-Case Execution Time Analysis (2018), Vol. 63. 5:1-5:12. https://doi.org/10.4230/OASICS.WCET.2018.5Google Scholar
- P. Lokuciejewski, F. Gedikli, P. Marwedel, and K. Morik. 2009. Automatic WCET Reduction by Machine Learning Based Heuristics for Function Inlining. In SMART.Google Scholar
- P. Lokuciejewski, S. Plazar, H. Falk, P. Marwedel, and L. Thiele. 2011. Approximating Pareto optimal compiler optimization sequences - a trade-off between WCET, ACET and code size. Software: Practice and Experience 41, 12 (2011), 1437-1458. https://doi.org/10.1002/spe.1079 Google ScholarDigital Library
- N. Mishra, H. Zhang, J. D. Lafferty, and H. Hoffmann. [n.d.]. A Probabilistic Graphical Model-based Approach for Minimizing Energy Under Performance Constraints. 50, 4 ([n. d.]), 267-281. https://doi.org/10.1145/2775054.2694373 Google ScholarDigital Library
- K. Muts and H. Falk. 2020. Multi-Criteria Function Inlining for Hard Real-Time Systems. In International Conference on Real-Time Networks and Systems. https://doi.org/10.1145/3394810.3394819 Google ScholarDigital Library
- K. Muts and H.Falk. 2021, Accepted for publication. Predicting Worst-Case Execution Times During Multi-Criterial Function Inlining. In International Conference on machine Learning, Optimization and Data science.Google Scholar
- F. Pedregosa, G. Varoquaux, A. Gramfort, V. Michel, B. Thirion, O. Grisel, M. Blondel, A. Müller, J. Nothman, G. Louppe, P. Prettenhofer, R. Weiss, V. Dubourg, J. Vanderplas, A. Passos, D. Cournapeau, M. Brucher, M. Perrot, and É. Duchesnay. 2012. Scikit-learn: Machine Learning in Python. (2012). Google ScholarDigital Library
- S.Purini and L. Jain.2013. Finding good optimization sequences covering program space. 9, 4 (2013), 1-23. https://doi.org/10.1145/2400682.2400715 Google ScholarDigital Library
- J. Thomson, M. O'Boyle, G. Fursin, and B. Franke. [n.d.]. Reducing Training Time in a One-Shot Machine Learning-Based Compiler. In Languages and Compilers for Parallel Computing. Springer Berlin Heidelberg, 399-407. https://doi.org/10.1007/978-3-642-13374-9_28 Google ScholarDigital Library
- S. Triantafyllis, M. Vachharajani, N. Vachharajani, and D. I. August. 2003. Compiler Optimization-Space Exploration. IEEE Computer Society, 204-215. Google ScholarDigital Library
- L. Wang, H. Ni, W. Zhou, P. M. Pardalos, J. Fang, and M. Fei. 2014. MBPOA-based LQR controller and its application to the double-parallel inverted pendulum system. Engineering Applications of Artificial Intelligence 36 (2014), 262-268. https://doi.org/10.1016/j.engappai.2014.07.023 Google ScholarDigital Library
Index Terms
- Predicting Objectives on a Reduced Search Space of Multiobjective Function Inlining
Recommendations
Multi-Criteria Function Inlining for Hard Real-Time Systems
RTNS '20: Proceedings of the 28th International Conference on Real-Time Networks and SystemsModern hard real-time systems shall satisfy some special requirements. Besides timing constraints, the additional design criteria such as code size and energy consumption are also not negligible. To optimize a system towards the mentioned specifications ...
Predicting Worst-Case Execution Times During Multi-criterial Function Inlining
Machine Learning, Optimization, and Data ScienceAbstractIn the domain of hard real-time systems, the Worst-Case Execution Time (WCET) is one of the most important design criteria. Safely and accurately estimating the WCET during a static WCET analysis is computationally demanding because of the ...
Integrating decision space diversity into hypervolume-based multiobjective search
GECCO '10: Proceedings of the 12th annual conference on Genetic and evolutionary computationMultiobjective optimization in general aims at learning about the problem at hand. Usually the focus lies on objective space properties such as the front shape and the distribution of optimal solutions. However, structural characteristics in the ...
Comments