vector_kernel.cuh File Reference

CUDA kernel methods for basic operations on vectors. More...

#include "cudpp_util.h"
#include "sharedmem.h"
#include "cudpp.h"

Functions
Vector Functions
CUDA kernel methods for basic operations on vectors.
template<class T >
__global__ void	vectorAddConstant (T *d_vector, T constant, int n, int baseIndex)
	Adds a constant value to all values in the input d_vector. More...
template<class T >
__global__ void	vectorAddUniform (T *d_vector, const T *d_uniforms, int numElements, int blockOffset, int baseIndex)
	Add a uniform value to each data element of an array. More...
template<typename T >
__global__ void	vectorAddUniform2 (T *g_data, T *uniforms, int n, int eltsPerBlock)
template<class T , class Oper , int elementsPerThread, bool fullBlocks>
__global__ void	vectorAddUniform4 (T *d_vector, const T *d_uniforms, int numElements, int vectorRowPitch, int uniformRowPitch, int blockOffset, int baseIndex)
	Add a uniform value to each data element of an array (vec4 version) More...
template<class T >
__global__ void	vectorAddVector (T *d_vectorA, const T *d_vectorB, int numElements, int baseIndex)
	Adds together two vectors. More...
template<class T , class Oper , bool isLastBlockFull>
__global__ void	vectorSegmentedAddUniform4 (T *d_vector, const T *d_uniforms, const unsigned int *d_maxIndices, unsigned int numElements, int blockOffset, int baseIndex)
	Add a uniform value to data elements of an array (vec4 version) More...
template<class T , class Oper , bool isLastBlockFull>
__global__ void	vectorSegmentedAddUniformToRight4 (T *d_vector, const T *d_uniforms, const unsigned int *d_minIndices, unsigned int numElements, int blockOffset, int baseIndex)
	Add a uniform value to data elements of an array (vec4 version) More...

Detailed Description

CUDA kernel methods for basic operations on vectors.

vector_kernel.cu

CUDA kernel methods for basic operations on vectors.

Examples:

• vectorAddConstant(): d_vector + constant
• vectorAddUniform(): d_vector + uniform (per-block constants)
• vectorAddVectorVector(): d_vector + d_vector