A generic fixed-capacity circular buffer using atomic indices for lock-free single-producer/single-consumer operation. Used for keyboard and mouse input buffering. Demonstrates careful memory ordering with Swift atomics and can be extracted as a standalone component for any producer-consumer scenario in a kernel or embedded context.

A heap-allocation-free printf implementation designed for use in interrupt handlers and kernel code. Uses Swift freestanding macros (#printf, #sprintf) to inline format parsing and argument packing into a Span of _PrintfArg values. The printing backend is decoupled via an UnicodeOutputStream protocol. This pattern is highly reusable in any constrained environment (embedded, RTOS, bootloaders) where dynamic allocation is prohibited.

A set of fixed-size bitmap allocators (8, 16, 32, 64, 128, and large variants) for managing pools of objects. Each allocator maintains a bitmask of free slots and provides allocate/free operations. Clean encapsulation and separation from kernel globals makes it suitable for extraction into any project needing lightweight object pooling.

Abstraction over memory-mapped I/O regions providing typed read/write methods (UInt8 through UInt64) with optional memory barriers. Supports sub-region slicing via MMIOSubRegion. The design decouples address translation from access semantics, making it reusable in drivers for different architectures or bus types. The file is self-contained and relies only on basic Swift primitives and a header for barrier intrinsics.