Discover flexible, scalable synchronization for parallel systems with path expressions.
This book presents a high-level language for coordinating many processes that share data, aiming to make complex synchronization easier to implement and reason about. It argues that path expressions offer more flexible control than traditional approaches and can adapt to a variety of architectures.
Path expressions define when events may execute and which may occur concurrently, using a structured syntax that supports sequencing, selection, and repetition. The discussion covers semantics, models, and practical considerations for implementing these ideas in real machines. The work also examines how arbitration works in parallel environments and introduces fairness concepts to ensure progress across competing events.
- Learn how path expressions constrain event execution to achieve correct synchronization.
- Explore the arbitration approach and how fairness can be realized in parallel systems.
- See how the concepts are formalized and analyzed, including semantic models and complexity notes.
- Read about portability and applicability to different shared-memory, MIMD architectures.
Ideal for readers of parallel computing, synchronization theory, and systems architecture who want a high-level framework for coordinating concurrent processes.