Where to call __syncthreads() function in CUDA

Question

I have some bool variable , and my each threads must work with that , So I need to synchronise them it means I need to call __syncthreads() function in my program , but I don't know where to call it.

Answer 1

__syncthreads() function is __device__ function , so you must call it from your __global__ kernel or some other __device__ function , additional __syncthreads() is a synchronization barrier primitive that only synchronizes threads in the same block.