Collaboration diagram for MapArray Kernels:

Functions
static std::string	skepu::MapArrayKernel_CL ("__kernel void MapArrayKernel_KERNELNAME(__global TYPE* input1, __global TYPE* input2, __global TYPE* output, unsigned int n)\n""{\n"" int i = get_global_id(0);\n"" unsigned int gridSize = get_local_size(0)*get_num_groups(0);\n"" while(i < n)\n"" {\n"" output[i] = FUNCTIONNAME(&input1[0], input2[i]);\n"" i += gridSize;\n"" }\n""}\n")
template<typename T , typename ArrayFunc >
__global__ void	skepu::MapArrayKernel_CU (ArrayFunc mapArrayFunc, T input1, T input2, T *output, unsigned int n)

Detailed Description

Definitions of CUDA and OpenCL kernels for the Map skeleton.

Function Documentation

static std::string skepu::MapArrayKernel_CL ( ) [static]

OpenCL MapArray kernel. Similar to Map computation but a pointer to first vector is supplied to the user function instead of just one element.

template<typename T , typename ArrayFunc >

__global__ void skepu::MapArrayKernel_CU	(	ArrayFunc	mapArrayFunc,
		T *	input1,
		T *	input2,
		T *	output,
		unsigned int	n
	)

CUDA MapArray kernel. Similar to Map computation but a pointer to first vector is supplied to the user function instead of just one element.