gs_nvshmem_kernels.h File Reference

#include <nvshmemx.h>

Include dependency graph for gs_nvshmem_kernels.h:

Go to the source code of this file.

Functions
template<typename T >
__global__ void	pack_pushShmemKernel (const T *__restrict__ u, T *dest, T *__restrict__ src, const int *__restrict__ dof, const int destRank, const int n, uint64_t iter, uint64_t *doneSig, uint64_t *readySlot)
template<>
__global__ void	pack_pushShmemKernel (const float *__restrict__ u, float *dest, float *__restrict__ src, const int *__restrict__ dof, const int destRank, const int n, uint64_t iter, uint64_t *doneSig, uint64_t *readySlot)
template<>
__global__ void	pack_pushShmemKernel (const double *__restrict__ u, double *dest, double *__restrict__ src, const int *__restrict__ dof, const int destRank, const int n, uint64_t iter, uint64_t *doneSig, uint64_t *readySlot)
__global__ void	pushShmemKernelWait (uint64_t iter, uint64_t *doneSlot)
__global__ void	postReadyShmemKernel (uint64_t *readySlot, uint64_t iter, const int srcRank)
template<typename T >
__global__ void	pack_pushShmemKernel_vec (const T *__restrict__ u, T *dest, T *__restrict__ src, const int *__restrict__ dof, const int destRank, const int n, const int nc, const int ns, uint64_t iter, uint64_t *doneSig, uint64_t *readySlot)
template<>
__global__ void	pack_pushShmemKernel_vec (const float *__restrict__ u, float *dest, float *__restrict__ src, const int *__restrict__ dof, const int destRank, const int n, const int nc, const int ns, uint64_t iter, uint64_t *doneSig, uint64_t *readySlot)
template<>
__global__ void	pack_pushShmemKernel_vec (const double *__restrict__ u, double *dest, double *__restrict__ src, const int *__restrict__ dof, const int destRank, const int n, const int nc, const int ns, uint64_t iter, uint64_t *doneSig, uint64_t *readySlot)

Function Documentation

pack_pushShmemKernel() [1/3]

template<>

__global__ void pack_pushShmemKernel	(	const double *__restrict__	u,
		double *	dest,
		double *__restrict__	src,
		const int *__restrict__	dof,
		const int	destRank,
		const int	n,
		uint64_t	iter,
		uint64_t *	doneSig,
		uint64_t *	readySlot
	)

Definition at line 108 of file gs_nvshmem_kernels.h.

Here is the call graph for this function:

pack_pushShmemKernel() [2/3]

template<>

__global__ void pack_pushShmemKernel	(	const float *__restrict__	u,
		float *	dest,
		float *__restrict__	src,
		const int *__restrict__	dof,
		const int	destRank,
		const int	n,
		uint64_t	iter,
		uint64_t *	doneSig,
		uint64_t *	readySlot
	)

Definition at line 78 of file gs_nvshmem_kernels.h.

Here is the call graph for this function:

pack_pushShmemKernel() [3/3]

template<typename T >

__global__ void pack_pushShmemKernel	(	const T *__restrict__	u,
		T *	dest,
		T *__restrict__	src,
		const int *__restrict__	dof,
		const int	destRank,
		const int	n,
		uint64_t	iter,
		uint64_t *	doneSig,
		uint64_t *	readySlot
	)

pack_pushShmemKernel_vec() [1/3]

template<>

__global__ void pack_pushShmemKernel_vec	(	const double *__restrict__	u,
		double *	dest,
		double *__restrict__	src,
		const int *__restrict__	dof,
		const int	destRank,
		const int	n,
		const int	nc,
		const int	ns,
		uint64_t	iter,
		uint64_t *	doneSig,
		uint64_t *	readySlot
	)

Definition at line 213 of file gs_nvshmem_kernels.h.

Here is the call graph for this function:

pack_pushShmemKernel_vec() [2/3]

template<>

__global__ void pack_pushShmemKernel_vec	(	const float *__restrict__	u,
		float *	dest,
		float *__restrict__	src,
		const int *__restrict__	dof,
		const int	destRank,
		const int	n,
		const int	nc,
		const int	ns,
		uint64_t	iter,
		uint64_t *	doneSig,
		uint64_t *	readySlot
	)

Definition at line 179 of file gs_nvshmem_kernels.h.

Here is the call graph for this function:

pack_pushShmemKernel_vec() [3/3]

template<typename T >

__global__ void pack_pushShmemKernel_vec	(	const T *__restrict__	u,
		T *	dest,
		T *__restrict__	src,
		const int *__restrict__	dof,
		const int	destRank,
		const int	n,
		const int	nc,
		const int	ns,
		uint64_t	iter,
		uint64_t *	doneSig,
		uint64_t *	readySlot
	)

postReadyShmemKernel()

__global__ void postReadyShmemKernel	(	uint64_t *	readySlot,
		uint64_t	iter,
		const int	srcRank
	)

Definition at line 150 of file gs_nvshmem_kernels.h.

Here is the call graph for this function:

pushShmemKernelWait()

__global__ void pushShmemKernelWait	(	uint64_t	iter,
		uint64_t *	doneSlot
	)

Definition at line 139 of file gs_nvshmem_kernels.h.

Here is the call graph for this function: