Distributed priority inheritance for real-time and embedded systems

  • Authors:

  • César Sánchez;Henny B. Sipma;Christopher D. Gill;Zohar Manna

  • Affiliations:

  • Computer Science Department, Stanford University, Stanford, CA;Computer Science Department, Stanford University, Stanford, CA;Department of Computer Science and Engineering, Washington University, St. Louis, MO;Computer Science Department, Stanford University, Stanford, CA

  • Venue:
  • OPODIS'06 Proceedings of the 10th international conference on Principles of Distributed Systems
  • Year:
  • 2006

Quantified Score

Hi-index 0.00

Visualization

Abstract

We study the problem of priority inversion in distributed real-time and embedded systems and propose a solution based on a distributed version of the priority inheritance protocol (PIP). Previous approaches to priority inversions in distributed systems use variations of the priority ceiling protocol (PCP), originally designed for centralized systems as a modification of PIP that also prevents deadlock. PCP, however, requires maintaining a global view of the acquired resources, which in distributed systems leads to high communication overhead. This paper presents a distributed PIP built on top of a deadlock avoidance schema that requires much less communication than PCP. Since the system is already deadlock free and priority inversions can be detected locally, we obtain an efficient dynamic resource allocation system that prevents deadlocks and handles priority inversions.