TOP
Search the Dagstuhl Website
Looking for information on the websites of the individual seminars? - Then please:
Not found what you are looking for? - Some of our services have separate websites, each with its own search option. Please check the following list:
Schloss Dagstuhl - LZI - Logo
Schloss Dagstuhl Services
Seminars
Within this website:
External resources:
  • DOOR (for registering your stay at Dagstuhl)
  • DOSA (for proposing future Dagstuhl Seminars or Dagstuhl Perspectives Workshops)
Publishing
Within this website:
External resources:
dblp
Within this website:
External resources:
  • the dblp Computer Science Bibliography


Dagstuhl Seminar 17131

Mixed Criticality on Multicore / Manycore Platforms

( Mar 26 – Mar 31, 2017 )

(Click in the middle of the image to enlarge)

Permalink
Please use the following short url to reference this page: https://www.dagstuhl.de/17131

Organizers

Contact


Impacts

Schedule

Motivation

The Mixed Criticality Systems (MCS) have become an important topic in the real-time systems community. The first cluster of the European collaborative projects on Mixed Criticality will reach its end in September 2016 indicating quick maturing of the related concepts within the industry and academia. Nevertheless we consider that many of the challenges brought about by the integration of mixed criticality applications onto multicore and manycore architectures are still there. In reality the Mixed Criticality problems have inherited the difficulty of real-time systems: being at the frontier of several domains like real-time scheduling, real-time operating systems / runtime environments, timing analysis as well as hardware architectures. This seminar aims to promote lively interaction, cross fertilisation of ideas, synergies, and closer collaboration across different MCS communities. The seminar will attract industrialists from the aerospace and automotive industries with specific interest in MCS.

In common with the previous Dagstuhl Seminar on Mixed Criticality Systems (MCS), this seminar will focus on the two key conflicting requirements of MCS: separation between criticality levels for assurance and sharing for resource efficiency, along with the related requirement of time composability. The seminar will crucially span from the low level behaviour of the memory hierarchy, and network-on-chip or buses, through timing analysis (WCET) and delays relating to pre-emption and migration (CRPD), real-time operating system (RTOS) behaviour, and high level scheduling, and task allocation to the verification of end-to-end deadlines.

An important aspect of the seminar involves obtaining different industry perspectives on the key problems and considerations in building future mixed criticality real-time systems. These perspectives will help to ground the research questions that are addressed, ensuring that the solutions developed are of value to industry.

The sessions of the seminar will be structured around a set of themes. Particular attention will be paid to the interfaces between themes, as these are the areas that can most benefit from improved understanding and collaboration. We aim to promote a holistic approach to solving the problems of MCS.

  • Task and system models for MCS on multicore and manycore platforms, including use of the various resources (memory, interconnect) in addition to the processors.
  • Scheduling schemes and analyses for MCS, including the integration of appropriate models of overheads and delays.
  • Run-time environments and support for MCS, including data exchange and synchronisation across criticality levels, and issues relating to consistency of the criticality mode.
  • Analysis of worst-case execution times (WCET) relating to MCS on multicore and manycore platforms, including cache related pre-emption and migration delays.
  • Mixed criticality communications mechanisms and analysis, including Network-on-Chip support.
  • Probabilistic analysis techniques for MCS.

The seminar does not aim to cover security aspects that relate to some MCS. It aims to be cognisant of the needs for certification in some industries, but does not seek to address the certification process itself. As a result of feedback from the 1st seminar suggesting stronger industrial involvement, the seminar will include a series of talks by industrial speakers.

Copyright Sanjoy K. Baruah, Liliana Cucu-Grosjean, Robert Davis, Zoë Stephenson, and Benoît Triquet

Summary

Real-time applications are characterized by the need for both functional correctness and temporal correctness (appropriate timing behaviour). Real-time systems are present in many diverse areas such as avionics, automotive, space, robotics, and medical applications to cite only a few. Mixed Criticality Systems (MCS) have become an important topic for the real-time systems community. The first cluster of the European collaborative projects on MCS has been completed in September 2016, indicating a maturing of the related concepts within both industry and academia. Nevertheless many of the challenges brought about by the integration of mixed criticality applications onto multicore and manycore architectures remain to be solved. In reality mixed criticality problems have inherited the difficulty of real-time systems: being at the frontier of several domains including real-time scheduling, real-time operating systems / runtime environments, and timing analysis, as well as hardware architectures. This seminar promoted lively interaction, cross fertilization of ideas, synergies, and closer collaboration across different sub-communities of academics and industrialists from aerospace, automotive, and railway industries with specific interests in MCS, as well as with experts in certification.

In common with the first Dagstuhl Seminar on Mixed Criticality Systems, this seminar also focused on the two key conflicting requirements of MCS: separation between criticality levels for assurance and sharing for resource efficiency, along with the related requirement of time composability. An important aspect of this seminar was the presentation of different industry perspectives on the key problems. These perspectives formed the starting point of our seminar, with the first day mainly dedicated to industry statements on current practice and their perception of current work on MCS. The academic participants benefited from substantial and detailed arguments from the industry speakers. There were lively interactive discussions during the talks which led to much improved understanding of current industry practice, as well as helping to build a common vocabulary between academic and industry participants. The first day concluded with presentations by academic speakers presenting their thoughts on more practical mixed criticality models.

The next three days each included sessions devoted to an invited tutorial from a academic speaker. These covered the one-out-of-m multicore problem, Networks-on-Chip and mixed criticality, resource management, and statistical approaches to worst-case execution time estimation. The remaining sessions covered a range of fascinating open problems. In addition, a number of ad-hoc small working groups formed to collaborate on specific topics. We are pleased to report that a significant number of these initial collaborations have gained traction resulting in further work after the seminar, and in some cases the development and submission of papers.

Organization of the seminar report. Section 3 is an overview of the industry talks and Section 4 provides an overview of the academic talks. Section 5 presents working group discussions. Section 6 summarizes open problems discussed during the seminar. Finally outcomes from the seminar are listed in Section 7. As organizers, we would like to thank Prof. Reinhard Wilhelm for joining us, Dagstuhl’s Scientific Directorate for allowing us to run a second seminar on mixed criticality systems, and to the staff at Schloss Dagstuhl for their superb support during the seminar itself. Finally, we would like to thank all of the participants for the very lively and open discussions. As organizers, we appreciated the feedback and enthusiasm which made running the seminar a great pleasure.

Copyright Liliana Cucu-Grosjean, Robert Davis, Sanjoy K. Baruah, and Zoë Stephenson

Participants
  • Kunal Agrawal (Washington University - St. Louis, US) [dblp]
  • Sebastian Altmeyer (University of Amsterdam, NL) [dblp]
  • James H. Anderson (University of North Carolina at Chapel Hill, US) [dblp]
  • Sanjoy Baruah (University of North Carolina at Chapel Hill, US) [dblp]
  • Iain Bate (University of York, GB) [dblp]
  • Enrico Bini (University of Turin, IT) [dblp]
  • Björn B. Brandenburg (MPI-SWS - Kaiserslautern, DE) [dblp]
  • Alan Burns (University of York, GB) [dblp]
  • Thidapat (Tam) Chantem (Virginia Polytechnic Institute - Arlington, US) [dblp]
  • Jian-Jia Chen (TU Dortmund, DE) [dblp]
  • Liliana Cucu-Grosjean (INRIA - Paris, FR) [dblp]
  • Robert Davis (University of York, GB) [dblp]
  • Arvind Easwaran (Nanyang TU - Singapore, SG) [dblp]
  • Pontus Ekberg (Uppsala University, SE) [dblp]
  • Sébastien Faucou (University of Nantes, FR) [dblp]
  • Madeleine Faugère (Thales Research and Technology - Palaiseau, FR) [dblp]
  • Christian Ferdinand (AbsInt - Saarbrücken, DE) [dblp]
  • Laurent George (ESIEE - Champs sur Marne, FR) [dblp]
  • Adriana Gogonel (INRIA - Paris, FR) [dblp]
  • Sathish Gopalakrishnan (University of British Columbia - Vancouver, CA) [dblp]
  • Emmanuel Grolleau (ENSMA - Chasseneuil, FR) [dblp]
  • Zhishan Guo (University of Missouri - Rolla, US) [dblp]
  • Leandro Soares Indrusiak (University of York, GB) [dblp]
  • Jaewoo Lee (University of Pennsylvania - Philadelphia, US) [dblp]
  • Jing Li (Washington University - St. Louis, US) [dblp]
  • Martina Maggio (Lund University, SE) [dblp]
  • Alberto Marchetti-Spaccamela (Sapienza University of Rome, IT) [dblp]
  • Cristian Maxim (Airbus S.A.S. - Toulouse, FR) [dblp]
  • Dorin Maxim (LORIA & INRIA - Nancy, FR) [dblp]
  • Mitra Nasri (MPI-SWS - Kaiserslautern, DE) [dblp]
  • Claire Pagetti (ONERA - Toulouse, FR) [dblp]
  • Kirk Pruhs (University of Pittsburgh, US) [dblp]
  • Gurulingesh Raravi (ISEP Porto, PT) [dblp]
  • Jan Reineke (Universität des Saarlandes, DE) [dblp]
  • Stefan Resch (Thales - Wien, AT) [dblp]
  • Philippa Ryan (Adelard - London, GB) [dblp]
  • Luca Santinelli (ONERA - Toulouse, FR) [dblp]
  • Zoë Stephenson (Rapita Systems Ltd. - York, GB) [dblp]
  • Sascha Uhrig (Airbus - München, DE) [dblp]
  • Wang Yi (Uppsala University, SE) [dblp]
  • Dirk Ziegenbein (Robert Bosch GmbH - Stuttgart, DE) [dblp]

Related Seminars
  • Dagstuhl Seminar 15121: Mixed Criticality on Multicore/Manycore Platforms (2015-03-15 - 2015-03-20) (Details)

Classification
  • optimization / scheduling
  • semantics / formal methods
  • verification / logic

Keywords
  • real-time systems
  • mixed-criticality
  • multi- and many-cores