ExtINT

ExtINT is a software abstraction and runtime for managing extended and programmable interrupts in embedded and real-time systems. It provides an extensible interface to traditional hardware interrupt controllers, enabling software-defined interrupt sources, dynamic prioritization, and isolation between high- and low-criticality code.

Origin and goals: ExtINT emerged from the need to modularize interrupt handling as systems grew more complex. By decoupling device drivers from the core interrupt controller, it aims to improve portability across architectures and simplify verification in safety- and mission-critical contexts.

Architecture: The ExtINT stack consists of a core interrupt controller layer, an extensible descriptor set for interrupt sources, a handler registry, and a policy layer that determines priority, masking, and preemption. It supports both hardware interrupts and software-generated interrupts, and provides hooks for instrumentation such as latency measurement and fault containment.

Usage and features: Implementations typically integrate with an RTOS or bare-metal runtime. Features include dynamic priority adjustment, nesting control, selective masking, end-of-interrupt signaling, and optional isolation boundaries for handlers. It supports architecture-specific backends and a portable API that remains constant across platforms.

Limitations and reception: While ExtINT can improve flexibility, it adds complexity and potential portability challenges. Verification and real-time guarantees depend on the quality of the backend and the correctness of policy definitions. In safety-critical domains, formal methods and thorough testing are recommended.

Status: As a niche approach, ExtINT is maintained by community contributors in some projects and may be included as optional components in certain RTOS distributions. Availability and maturity vary by platform.