Module ring 

Zero-Copy SPSC Ring Buffer with AF_XDP Integration

This implementation uses pinned allocations and proper memory ordering for single-producer/single-consumer DMA-style hand-off.

Optimizations:

• 4x larger capacity (4096) for better batching
• Prefetching in hot paths for cache-friendly access
• Optimal cache-line padding (128 bytes) to avoid false sharing

Structs

RingConfig

Ring buffer configuration with backpressure thresholds

SpscRingBuffer

Zero-copy SPSC ring buffer with proper memory ordering