Compartilhar via


all_memory_fence Function

Blocks execution of all threads in a tile until all memory accesses have been completed. This ensures that all memory accesses are visible to other threads in the thread tile, and are executed in program order.

inline void all_memory_fence(
   const tile_barrier & _Barrier
) restrict(amp);

Parameters

  • _Barrier
    A tile_barrier object.

Requirements

Header: amp.h

Namespace: Concurrency::direct3d

See Also

Reference

Concurrency::direct3d Namespace