Subsequent to the publication of the seminal work by Liu and Layland in 1973, researchers and practitioners alike started discussing which online scheduling algorithm was to be preferred between FPS and EDF. Results published in 2005 sustained the superiority of EDF, already proven in theory, also from an implementation perspective. With this work, we aim at digging deeper into the roots of those results. To this end, we took the first-ever instance of an Ada Ravenscar runtime, with its FPS scheduler, combined with its IPCP locking policy companion, and developed a variant of it that implements EDF scheduling coupled with DFP locking. In this manner, we were able to transparently attach those two runtime variants to a suite of synthetic benchmarks, which we used to perform an extensive quantitative comparison between those two runtimes, getting to the bottom of where one prevails on the other.

Ravenscar-EDF: Comparative benchmarking of an EDF variant of a ravenscar runtime

CARLETTO, PAOLO
Membro del Collaboration Group
;
Vardanega, Tullio
Membro del Collaboration Group
2017

Abstract

Subsequent to the publication of the seminal work by Liu and Layland in 1973, researchers and practitioners alike started discussing which online scheduling algorithm was to be preferred between FPS and EDF. Results published in 2005 sustained the superiority of EDF, already proven in theory, also from an implementation perspective. With this work, we aim at digging deeper into the roots of those results. To this end, we took the first-ever instance of an Ada Ravenscar runtime, with its FPS scheduler, combined with its IPCP locking policy companion, and developed a variant of it that implements EDF scheduling coupled with DFP locking. In this manner, we were able to transparently attach those two runtime variants to a suite of synthetic benchmarks, which we used to perform an extensive quantitative comparison between those two runtimes, getting to the bottom of where one prevails on the other.
2017
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
9783319605876
File in questo prodotto:
Non ci sono file associati a questo prodotto.
Pubblicazioni consigliate

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11577/3257899
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 2
  • ???jsp.display-item.citation.isi??? 2
social impact