tensor_sse-inl.hpp File Reference

support of sse2 optimization of some operations More...

#include <malloc.h>
#include "tensor_expr.h"
#include "tensor.h"
#include <emmintrin.h>

Go to the source code of this file.

Classes
struct	mshadow::sse2::FVec< FloatType >
	float vector real type, used for vectorization More...
struct	mshadow::sse2::FVec< float >
	vector real type for float More...
struct	mshadow::sse2::FVec< double >
	vector real type for float More...
struct	mshadow::sse2::SSEOp< OP >
	sse2 operator type of certain operator More...
struct	mshadow::sse2::SSEOp< op::plus >
struct	mshadow::sse2::SSEOp< op::minus >
struct	mshadow::sse2::SSEOp< op::mul >
struct	mshadow::sse2::SSEOp< op::div >
struct	mshadow::sse2::SSEOp< op::identity >
struct	mshadow::sse2::Saver< SV, TFloat >
struct	mshadow::sse2::Saver< sv::saveto, TFloat >
class	mshadow::expr::SSEPlan< ExpType >
class	mshadow::expr::SSEPlan< Tensor< Device, dim > >
class	mshadow::expr::SSEPlan< ScalarExp >
class	mshadow::expr::SSEPlan< BinaryMapExp< OP, TA, TB, etype > >
class	mshadow::expr::SSEPlan< UnaryMapExp< OP, TA, etype > >
struct	mshadow::expr::SSECheck< E >
	static check sse enable if a expression E can not be evaluated using sse, then kPass = false More...
struct	mshadow::expr::SSECheck< ScalarExp >
struct	mshadow::expr::SSECheck< Tensor< cpu, dim > >
struct	mshadow::expr::SSECheck< UnaryMapExp< OP, TA, etype > >
struct	mshadow::expr::SSECheck< BinaryMapExp< OP, TA, TB, etype > >
struct	mshadow::expr::SSEAlignCheck< dim, E >
struct	mshadow::expr::SSEAlignCheck< dim, ScalarExp >
struct	mshadow::expr::SSEAlignCheck< dim, Tensor< cpu, dim > >
struct	mshadow::expr::SSEAlignCheck< dim, UnaryMapExp< OP, TA, etype > >
struct	mshadow::expr::SSEAlignCheck< dim, BinaryMapExp< OP, TA, TB, etype > >

Namespaces
	mshadow
	namespace for mshadow
	mshadow::sse2
	namespace to support sse2 vectorization
	mshadow::expr
	namespace for abstract expressions and expressions template, have no dependecy on tensor.h, These data structure takes no charge in computations, they are only used to define operations and represent expression in a symbolic way

Functions
void *	mshadow::sse2::AlignedMallocPitch (size_t &pitch, size_t lspace, size_t num_line)
	analog to cudaMallocPitch, allocate a aligned space with num_line * lspace cells More...
void	mshadow::sse2::AlignedFree (void *ptr)
	free aligned space More...
bool	mshadow::sse2::CheckAlign (size_t pitch)
	check if a pointer is aligned
bool	mshadow::sse2::CheckAlign (void *ptr)
	check if a pointer is aligned
index_t	mshadow::sse2::UpperAlign (index_t size, size_t fsize)
	get upper bound of aligned index of size More...
index_t	mshadow::sse2::LowerAlign (index_t size, size_t fsize)
	get lower bound of aligned index of size More...
template<typename OP , typename TA , typename TB , int etype>
SSEPlan< BinaryMapExp< OP, TA, TB, etype > >	mshadow::expr::MakeSSEPlan (const BinaryMapExp< OP, TA, TB, etype > &e)
SSEPlan< ScalarExp >	mshadow::expr::MakeSSEPlan (const ScalarExp &e)
template<typename T >
SSEPlan< T >	mshadow::expr::MakeSSEPlan (const ContainerExp< T > &e)
template<typename T , int dim>
SSEPlan< T >	mshadow::expr::MakeSSEPlan (const MakeTensorExp< T, cpu, dim > &e)
template<typename OP , typename TA , int etype>
SSEPlan< UnaryMapExp< OP, TA, etype > >	mshadow::expr::MakeSSEPlan (const UnaryMapExp< OP, TA, etype > &e)
template<typename SV , typename E , int dim>
void	mshadow::MapSSEPlan (Tensor< cpu, dim > _dst, const expr::SSEPlan< E > &plan)
	use SSEPlan to compute result

Detailed Description

support of sse2 optimization of some operations

Author

Tianqi Chen