cu_sell_kernel.h File Reference

#include "ghost/types.h"
#include "ghost/cu_complex.h"
#include <cuda.h>

Include dependency graph for cu_sell_kernel.h:

This graph shows which files directly or indirectly include this file:

Go to the source code of this file.

Classes
struct	CustomSum

Functions
template<typename v_t >
__device__ v_t	ghost_shfl_down32 (v_t var, unsigned int srcLane)
template<typename v_t >
__device__ v_t	ghost_shfl_down (v_t var, unsigned int srcLane, int width)
template<>
__device__ cuFloatComplex	ghost_shfl_down< cuFloatComplex > (cuFloatComplex var, unsigned int srcLane, int width)
template<>
__device__ cuDoubleComplex	ghost_shfl_down< cuDoubleComplex > (cuDoubleComplex var, unsigned int srcLane, int width)
template<typename v_t >
__inline__ __device__ v_t	ghost_warpReduceSum (v_t val)
template<typename v_t >
__inline__ __device__ v_t	ghost_partialWarpReduceSum (v_t val, int size, int width)
template<typename v_t , int size>
__inline__ __device__ v_t	ghost_partialWarpReduceSumFast (v_t val)
template<>
__inline__ __device__ double3	ghost_warpReduceSum< double3 > (double3 val)
template<typename v_t >
__inline__ __device__ v_t	ghost_partialBlockReduceSum (v_t val, int size)
template<typename v_t >
__inline__ __device__ v_t	ghost_1dPartialBlockReduceSum (v_t val, int nwarps)
template<typename v_t >
__inline__ __device__ v_t	ghost_blockReduceSum (v_t val)
template<>
__inline__ __device__ double3	ghost_blockReduceSum< double3 > (double3 val)
template<typename v_t >
__global__ void	ghost_deviceReduceSum (v_t *in, v_t *out, ghost_lidx N)
template<typename v_t >
__global__ void	ghost_deviceReduceSumMultiple (v_t *in, v_t *out, ghost_lidx N, ghost_lidx ncols)
template<typename T >
__device__ __inline__ T	streaming_load (const T *addr)
template<>
__device__ __inline__ double	streaming_load (const double *addr)
template<>
__device__ __inline__ float	streaming_load (const float *addr)
template<>
__device__ __inline__ cuDoubleComplex	streaming_load (const cuDoubleComplex *addr)
template<>
__device__ __inline__ cuFloatComplex	streaming_load (const cuFloatComplex *addr)
template<typename T >
__device__ __inline__ void	streaming_store (T *addr, const T val)
template<>
__device__ __inline__ void	streaming_store (double *addr, const double val)
template<>
__device__ __inline__ void	streaming_store (float *addr, const float val)
template<>
__device__ __inline__ void	streaming_store (cuDoubleComplex *addr, const cuDoubleComplex val)

Variables
__shared__ char	shared []

Function Documentation

template<typename v_t >

__inline__ __device__ v_t ghost_1dPartialBlockReduceSum	(	v_t	val,
		int	nwarps
	)

Here is the call graph for this function:

template<typename v_t >

__inline__ __device__ v_t ghost_blockReduceSum

(

v_t

val

)

Here is the call graph for this function:

template<>

__inline__ __device__ double3 ghost_blockReduceSum< double3 >

(

double3

val

)

Here is the call graph for this function:

template<typename v_t >

__global__ void ghost_deviceReduceSum	(	v_t *	in,
		v_t *	out,
		ghost_lidx	N
	)

Here is the call graph for this function:

template<typename v_t >

__global__ void ghost_deviceReduceSumMultiple	(	v_t *	in,
		v_t *	out,
		ghost_lidx	N,
		ghost_lidx	ncols
	)

Here is the call graph for this function:

template<typename v_t >

__inline__ __device__ v_t ghost_partialBlockReduceSum	(	v_t	val,
		int	size
	)

Here is the call graph for this function:

template<typename v_t >

__inline__ __device__ v_t ghost_partialWarpReduceSum	(	v_t	val,
		int	size,
		int	width
	)

Here is the call graph for this function:

template<typename v_t , int size>

__inline__ __device__ v_t ghost_partialWarpReduceSumFast

(

v_t

val

)

Here is the call graph for this function:

template<typename v_t >

__device__ v_t ghost_shfl_down ( v_t  var, unsigned int  srcLane, int  width  )

inline

template<typename v_t >

__device__ v_t ghost_shfl_down32 ( v_t  var, unsigned int  srcLane  )

inline

Here is the call graph for this function:

template<>

__device__ cuDoubleComplex ghost_shfl_down< cuDoubleComplex > ( cuDoubleComplex  var, unsigned int  srcLane, int  width  )

inline

Here is the call graph for this function:

template<>

__device__ cuFloatComplex ghost_shfl_down< cuFloatComplex > ( cuFloatComplex  var, unsigned int  srcLane, int  width  )

inline

Here is the call graph for this function:

template<typename v_t >

__inline__ __device__ v_t ghost_warpReduceSum

(

v_t

val

)

Here is the call graph for this function:

template<>

__inline__ __device__ double3 ghost_warpReduceSum< double3 >

(

double3

val

)

Here is the call graph for this function:

template<typename T >

__device__ __inline__ T streaming_load

(

const T *

addr

)

template<>

__device__ __inline__ double streaming_load

(

const double *

addr

)

template<>

__device__ __inline__ float streaming_load

(

const float *

addr

)

template<>

__device__ __inline__ cuDoubleComplex streaming_load

(

const cuDoubleComplex *

addr

)

template<>

__device__ __inline__ cuFloatComplex streaming_load

(

const cuFloatComplex *

addr

)

template<typename T >

__device__ __inline__ void streaming_store	(	T *	addr,
		const T	val
	)

template<>

__device__ __inline__ void streaming_store	(	double *	addr,
		const double	val
	)

template<>

__device__ __inline__ void streaming_store	(	float *	addr,
		const float	val
	)

template<>

__device__ __inline__ void streaming_store	(	cuDoubleComplex *	addr,
		const cuDoubleComplex	val
	)

Variable Documentation

__shared__ char shared[]