Exokernelbased

Exokernel-based systems are operating systems built on the exokernel architecture. In these systems, the kernel is deliberately minimal and focuses on secure, low-level resource multiplexing, while resource management and policy decisions are implemented in user-space libraries or library operating systems. The goal is to expose raw hardware resources to applications under strict protection, enabling specialized, high-performance configurations.

Core principles include the separation of policy from mechanism, and the use of secure interfaces and capabilities to control access to resources. The kernel provides only primitive abstractions for resources such as memory, CPUs, and I/O devices, and enforces protection without imposing high-level scheduling or caching decisions. Applications deploy libOS components that implement their own resource management and OS services, potentially tailored to specific workloads.

Implementation typically features a minimal kernel that exports resource primitives and enforces protection, while user-space components handle resource allocation, multiplexing, and policy. Performance-critical paths can bypass kernel overhead by performing management in user space, with the kernel ensuring safe sharing and isolation.

Advantages include potential for higher performance, greater flexibility, and finer-grained control over resources. By moving policy and management into user space, systems can be customized for real-time, multimedia, or datacenter workloads, and experiments with novel resource sharing strategies are enabled. Critics note increased development complexity, larger trusted computing base in user space, portability concerns, and a lack of broad commercial adoption. Security relies on correct implementation of capabilities and robust boundaries between kernel and library OS code.

Historically, the concept originated with MIT researchers in the 1990s, leading to the Exokernel and ExOS projects as a prominent example of exokernel-based design. The term remains a reference point in computer systems research for policy-mechanism separation and lean kernels.