2026-08-15T05:50:51Z https://drops.dagstuhl.de/oai

oai:drops-oai.dagstuhl.de:1186 2024-03-06T10:28:26Z ddc:004 open_access

Automatic Amortised Worst-Case Execution Time Analysis Herrmann, Christoph A. Bonenfant, Armelle Hammond, Kevin Jost, Steffen Loidl, Hans-Wolfgang Pointon, Robert Amortisation functional programming performance measurement static analysis type and effect systems worst-case execution time Our research focuses on formally bounded WCET analysis, where we aim to provide absolute guarantees on execution time bounds. In this paper, we describe how amortisation can be used to improve the quality of the results that are obtained from a fully-automatic and formally guaranteed WCET analysis, by delivering analysis results that are parameterised on specific input patterns and which take account of relations between these patterns. We have implemented our approach to give a tool that is capable of predicting execution costs for a typical embedded system development platform, a Renesas board with a Renesas M32C/85U processor. We show that not only is the amortised approach applicable in theory, but that it can be applied automatically to yield good WCET results. Schloss Dagstuhl – Leibniz-Zentrum für Informatik Christoph A. Herrmann and Armelle Bonenfant and Kevin Hammond and Steffen Jost and Hans-Wolfgang Loidl and Robert Pointon 2007 Is Part Of OASIcs, Volume 6, 7th International Workshop on Worst-Case Execution Time Analysis (WCET'07) (2007) InProceedings Text doc-type:ResearchArticle publishedVersion application/pdf doi:10.4230/OASIcs.WCET.2007.1186 urn:nbn:de:0030-drops-11868 https://drops.dagstuhl.de/entities/document/10.4230/OASIcs.WCET.2007.1186 eng https://creativecommons.org/licenses/by-nc-nd/3.0/legalcode