MayaFlux::Memory::SingleThreadedPolicy Struct Reference

Non-atomic single-threaded operation. More...

#include <RingBuffer.hpp>

Collaboration diagram for MayaFlux::Memory::SingleThreadedPolicy:

Classes
struct	State

Static Public Attributes
static constexpr bool	is_thread_safe = false
static constexpr bool	requires_trivial_copyable = false
static constexpr bool	requires_fixed_storage = false

Detailed Description

Non-atomic single-threaded operation.

Provides minimal overhead when all access is from a single thread or when synchronization is handled externally (e.g., mutex-protected).

Thread Safety:

• NO thread synchronization
• Use when externally synchronized OR single-threaded

Benefits:

• No atomic overhead (~2-5 CPU cycles faster per operation)
• Works with any element type (no TriviallyCopyable requirement)
• Compatible with both FixedStorage and DynamicStorage

When to Use:

• DSP delay lines in audio callback (single thread)
• Polynomial node history buffers (single thread)
• Non-realtime data structures
• Already protected by external mutex

// Single-threaded delay line for audio DSP
using AudioDelay = RingBuffer<double,
                              DynamicStorage<double>,
                              SingleThreadedPolicy,
                              HomogeneousAccess>;
 
void audio_process_callback() {
    AudioDelay delay(1000);  // No thread sync overhead
    delay.push(input_sample);
    auto delayed = delay.peek_oldest();
}

Definition at line 216 of file RingBuffer.hpp.

---

The documentation for this struct was generated from the following file:

• RingBuffer.hpp