GPU Programming GPU编程基础.ppt

Synchronization Functions void __syncthreads() waits until all threads in the thread block have reached this point and all global and shared memory accesses made by these threads prior to __syncthreads() are visible to all threads in the block. __syncthreads() is used to coordinate communication between the threads of the same block. When some threads within a block access the same addresses in shared or global memory, there are potential read-after-write, write-after-read, or write-afterwrite hazards for some of these memory accesses. These data hazards can be avoided bysynchronizing threads in-between these accesses. Cuda Runtime API Memory Management Allocate memory on the device __host__ __device__ cudaError_t cudaMalloc (void **devPtr, size_t size) Parameters devPtr：Pointer to allocated device memory size：Requested allocation size in bytes Returns cudaSuccess, cudaErrorMemoryAllocation Initializes or sets device memory to a value __host__ cudaError_t cudaMemset (void *devPtr, int value, size_t count) Parameters devPtr: Pointer to device memory Cuda Runtime API value: Value to set for each byte of specified memory count: Size in bytes to set Returns cudaSuccess, cudaErrorInvalidValue, cudaErrorInvalidDevicePointer Copies data between host and device. __host__cudaError_t cudaMemcpy (void *dst, const void *src, size_t count, cudaMemcpyKind kind) Parameters dst: Destination memory address src: Source memory address count: Size in bytes to copy kind: Type of transfer,specifies the direction of the copy cudaMemcpyHostToHost, cudaMemcpyHostToDevice, cudaMemcpyDeviceToHost, cudaMemcpyDeviceToDevice, Cuda Runtime API Returns cudaSuccess, cudaErrorInvalidValue, cudaErrorInvalidDevicePointer, cudaErrorInvalidMemcpyDirection Frees memory on the device __host____device__cudaError_t cudaFree (void *devPtr) Parameters devPtr: Device pointer to memory to free Returns cudaSuccess, cudaErrorInvalidDevicePointer, cudaErrorInitializationError Outline Cuda Filename Suffix T