OSPERT 2013

9th annual workshop on

Operating Systems Platforms for Embedded Real-Time applications

July 9, 2013. Paris, France
held in conjunction with

Workshop Chairs:

Andrea Bastoni
SYSGO AG, Germany

Shinpei Kato
Nagoya University, Japan

Program Committee:

Björn B. Brandenburg
Max Planck Institute for Software Systems

Robert Kaiser
Hochschule RheinMain University of Applied Sciences

Gabriel Parmer
George Washington University

Rodolfo Pelizzoni
University of Waterloo

Thomas Nolte
Malardaren University

Neil Audsley
University of York

Richard West
Boston University

Paul McKenney
IBM

Wolfgang Mauerer
Siemens

Steven Rostedt
Red Hat

Proceedings (pdf)

Program

	Tuesday, July 9th 2013
8:30-9:30	Registration
9:30-11:00	Keynote Talk by Frederic Weisbecker (Red Hat): Title: Present and Future of Linux dynticks Abstract: Linux has been able to stop the tick for a few years now. This feature is known as dynticks. Although it was limited to idle CPUs, this was a great step towards enabling power-saving solutions. Now, the 3.10 Linux Kernel can extend this dynamic tick behavior not only to idle CPUs, but also to busy CPUs. This time, power-saving is not the only target, but rather latency and performance. For now the benefit is mostly to be expected in extreme real-time and High Performance Computing, but it may extend to more general purposes in the long run. This talk aims at diving into dynticks internals and speculates about its future improvements. Biography: Frederic Weisbecker is a Linux Kernel developer working for Red Hat. His involvement and role in the Linux community has evolved over time: he has been working on tracing with ftrace and perf events subsystems, on timers and dynticks-mode, and he helped to remove the big kernel lock. In 2010, he took up the challenge of disabling the tick interrupt on non-idle processors. Eventually, after many changes and helps from other Linux Kernel developers, his work has been merged into the 3.10 kernel.
11:00-11:30	Coffee Break
11:30-13:00	Session 1: Implementation and Performance Evaluation Investigation and Improvement on the Impact of TLB Misses in Real-Time Systems Takuya Ishikawa, Toshikazu Kato, Shinya Honda, Hiroaki Takada Implementation of the Multi-Level Adaptive Hierarchical Scheduling Framework Nima Moghaddami Khalilzad, Moris Behnam, Thomas Nolte A Comparison of Scheduling Latency in Linux, PREEMPT-RT, and LITMUSRT Felipe Cerqueria, Boern Brandenburg
13:30-14:30	Lunch
14:30-16:00	Session 2: Power-Management and Open-Source Projects Towards Power-Efficient Mixed-Critical Systems Florian Broekaert, Sergey Tverdyshev, Laurent San, Agnes Fritsch Reverse Engineering Power Management on NVIDIA GPUs - Anatomy of an Autonomic-Ready System Martin Peres The state of Composite Jiguo Song, Qi Wang, Gabriel Parmer
16:00-16:30	Coffee Break
16:30-18:00	Session 3: Resource Sharing and Locking Priority Inheritance on Condition Variables Tommaso Cucinotta Deterministic Fast User Space Synchronization Alexander Zupke
17:30-18:00	Discussion and Closing Thoughts

Call for papers (txt)

Research on innovative RTOS architectures and services is a hot topic. Developers of Real-Time Operating Systems (RTOS) are faced with many challenges arising from two opposite needs: extreme optimization of resource usage (processor, energy, network bandwidth, etc.) vs. dynamic configuration and upgrading, flexible scheduling, component-based development and deployment, etc. While real-time systems continue to be used in many small embedded applications, real-time services are being introduced and used in general- purpose operating systems. Notable examples are the various flavors Linux that provide support to time-sensitive applications.

This workshop is intended as a forum for researchers and practitioners of RTOS to discuss the recent advances in RTOS technology and the challenges that lie ahead. The workshop will consist of three categories of submitted papers:

• Open-source project reports. These papers presents key features, future directions and open issues of of existing university or industrial OS-related projects. These papers have a page limit of 1 page with the addition of useful links to the project webpages.
• Short papers and forward-looking papers. These papers focus on implementation discussions, innovative ideas, and vision about what an RTOS must consider to effectively support future real-time applications, and how they should be constructed over the upcoming years. The papers also focus on open problems and implementation issues faced when enhancing existing OS-technologies with new real-time capabilities. These papers consist of 2 to 5 pages.
• Technical papers. These papers are the traditional forms of OSPERTs consisting of technical contributions that advance the state of the art. They have a page limit of 10 pages.

Additionally, the workshop will feature invited presentations and publications about academic state-of-the-art and industrial state-of-practice within the area.

Upon acceptance, a complete version of the paper must be prepared and submitted. All papers will be made available to all participants a week before the workshop so that contributions can be examined prior to the workshop.

Scope

Topics include but are not limited to:

• Dynamic reconfiguration and upgrading
• Support for component based development
• Scalability, from very small scale embedded systems to full-fledged OSes
• Flexible scheduling
• Quality of Service guarantees
• Real-Time on Linux
• Interaction with reconfigurable hardware
• Support for embedded multiprocessor architectures
• Coordinated management of multiple resources
• Security and fault tolerance for embedded real-time systems
• Power-aware operating systems

Submission of Papers

Submitted papers should follow the IEEE conference format (2 columns, 10 pt, single-line spacing) and should not exceed the page limit listed above for the three submission categories respectively. On the title page, the submission should clearly state which category of papers the submission is for. Papers may be submitted in either PDF or Postscript format. The papers will be reviewed by the workshop Program Committee. All accepted papers will be made available to all participants one week before the workshop so that contributions can be examined prior to the event.

We want to encourage scientific reproducibility, and discourage redundant work within the community. Thus, code relevant to each submission must be publicly available, and a link to it included in the submission. If you require an exception, please contact the chairs. If your submission does not include an implementation, there are no such restrictions.

Papers should be submitted by using the START Conference Manager system. The system acknowledges receipt of each submission by sending an e-mail to the contact author, and it allows to revise a submission till the deadline.

If a paper is accepted, at least one author should register for the workshop following indications sent in the notification of acceptance, and present the paper at the workshop in person.

To submit a paper, visit the submission page.

© 2013 Report issues to Shinpei Kato
Template by Andreas Viklund