Expand description
Performance optimization infrastructure and implementations.
This module provides optimized implementations of critical paths in the threading system, including CPU cache-aware algorithms, NUMA optimization, and architecture-specific acceleration.
Modules§
- cache_
aware - Cache-aware optimizations for scheduler hot paths.
- context_
switch_ opt - Context switch optimizations with micro-benchmarking support.
- cpu_
dispatch - CPU-specific optimizations and SIMD acceleration.
- fast_
paths - Lock-free fast paths for common threading operations.
- memory_
pools - Per-CPU memory pools for optimized allocation performance.
- numa
- NUMA-aware thread placement and memory allocation optimizations.
Structs§
- Perf
Config - Performance configuration structure.
- Perf
Counters - Performance counters for monitoring optimization effectiveness.
Statics§
- PERF_
COUNTERS - Global performance counters instance.
Functions§
- get_
perf_ config - Get the current performance configuration.
- init_
perf_ optimization - Initialize performance subsystem with optimal configuration.