Compartilhar via


global_memory_fence Function

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

inline void global_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