A tensor plus its reference to its parent tensor. More...

#include <TTL_tensors.h>

Collaboration diagram for TTL_sub_tensor< TENSORTYPE >:

Data Structures
struct	Origin

Public Member Functions
	TTL_sub_tensor (TENSORTYPE *base, const TTL_shape &shape, const TTL_layout &layout, const TTL_dim elem_size, const TTL_offset offset, const TTL_shape origin_shape, TTL_offset origin_offset)
	TTL_sub_tensor(TTL_create_, const_1, location, type, sub, _7_params)

	TTL_sub_tensor (TENSORTYPE *const base, const TTL_shape &shape, const TTL_layout &layout, const TTL_tensor< TENSORTYPE > &origin_tensor, const TTL_offset &sub_offset)
	TTL_sub_tensor(TTL_create_, const_1, location, type, sub, _5_params)

	TTL_sub_tensor (TENSORTYPE const base, const TTL_tensor< TENSORTYPE > &origin_tensor)
	TTL_sub_tensor(TTL_create_, const_1, location, type, sub, _2_params)

	TTL_sub_tensor (const TTL_tensor< TENSORTYPE > &origin_tensor)
	TTL_sub_tensor(TTL_create_, const_1, location, type, sub, _1_param)

	TTL_sub_tensor ()
	Create an empty tiled internal tensor. Empty means it has all dimensions set to zero.

const TENSORTYPE &	read (const unsigned int x, const unsigned int y=0, const unsigned int z=0) const
	Read a value from a sub_tensor.

TENSORTYPE	write (const TENSORTYPE value, const unsigned int x, const unsigned int y=0, const unsigned int z=0)
	Write a value to a sub_tensor.

bool	empty ()
	A Tensor is empty if its shape is empty.

Data Fields
TTL_tensor< TENSORTYPE >	tensor

Origin	origin

Detailed Description

template<typename TENSORTYPE>
struct TTL_sub_tensor< TENSORTYPE >

A tensor plus its reference to its parent tensor.

TTL_sub_tensor contains both the logical dimensions of a tile as well as its physical mapping to memory.

Parameters

TTL_scope	The scope of the creation can be TTL_global or TTL_local
const_1	The const name to place after the prefix - should be empty or const_
location	The location of the tensor - should be ext or int
type	The type of the tensor - should be any valid c type
const_2	The const type to create - should be empty or const

Definition at line 275 of file TTL_cpp/TTL_tensors.h.

Constructor & Destructor Documentation

◆ TTL_sub_tensor() [1/5]

template<typename TENSORTYPE>

TTL_sub_tensor< TENSORTYPE >::TTL_sub_tensor	(	TENSORTYPE *	base,
		const TTL_shape &	shape,
		const TTL_layout &	layout,
		const TTL_dim	elem_size,
		const TTL_offset	offset,
		const TTL_shape	origin_shape,
		TTL_offset	origin_offset )

inline

TTL_sub_tensor(TTL_create_, const_1, location, type, sub, _7_params)

Parameters

base	The ## TTL_scope ## base address of the tensor in ## location ## memory
shape	Description of the shape of the tensor that base points
layout	The layout of the ## location ## tensor
elem_size	The size of a single element in the
offset	The offset of the tensor from the base.

See also: TTL_offset

Parameters

origin_shape	The shape of the tensor that originated the sub tensor
origin_offset	The offset of the tensor from the souce tensor.

See also: TTL_offset

Returns: return a TTL_sub_tensor(TTL_, const_1, location, type, sub, _t)

Definition at line 289 of file TTL_cpp/TTL_tensors.h.

◆ TTL_sub_tensor() [2/5]

template<typename TENSORTYPE>

TTL_sub_tensor< TENSORTYPE >::TTL_sub_tensor	(	TENSORTYPE *const	base,
		const TTL_shape &	shape,
		const TTL_layout &	layout,
		const TTL_tensor< TENSORTYPE > &	origin_tensor,
		const TTL_offset &	sub_offset )

inline

TTL_sub_tensor(TTL_create_, const_1, location, type, sub, _5_params)

Parameters

base	The ## TTL_scope ## base address of the tensor in ## location ## memory
shape	Description of the shape of the tensor that base points
layout	The layout of the ## location ## tensor
origin_tensor	The tensor that originated the sub tensor
sub_offset	The offset of the tensor from the souce tensor.

See also: TTL_offset

The element size of the tensor is taken from the origin tensor The offset of the sub tensor is taken to be (0, 0, 0)

Returns: return a TTL_sub_tensor(TTL_, const_1, location, type, sub, _t)

Definition at line 308 of file TTL_cpp/TTL_tensors.h.

◆ TTL_sub_tensor() [3/5]

template<typename TENSORTYPE>

TTL_sub_tensor< TENSORTYPE >::TTL_sub_tensor	(	TENSORTYPE const	base,
		const TTL_tensor< TENSORTYPE > &	origin_tensor )

inline

TTL_sub_tensor(TTL_create_, const_1, location, type, sub, _2_params)

Simply create a sub_tensor from an origin tensor

Parameters

base	The ## TTL_scope ## base address of the tensor in ## location ## memory
origin_tensor	The tensor that originated the sub tensor

The element size of the tensor is taken from the origin tensor The offset of the sub tensor is taken to be (0, 0, 0) The offset of the sub tensor relative to the source tensor is taken to be (0, 0, 0)

Returns: return a TTL_sub_tensor(TTL_, const_1, location, type, sub, _t)

Definition at line 327 of file TTL_cpp/TTL_tensors.h.

◆ TTL_sub_tensor() [4/5]

template<typename TENSORTYPE>

TTL_sub_tensor< TENSORTYPE >::TTL_sub_tensor ( const TTL_tensor< TENSORTYPE > & origin_tensor )

inline

TTL_sub_tensor(TTL_create_, const_1, location, type, sub, _1_param)

Simply create a sub_tensor from an origin tensor

Parameters

origin_tensor The tensor that originated the sub tensor

Effective creates a sub-tensor that is a tensor covering 100% of the source tensor

Returns: return a TTL_sub_tensor(TTL_, const_1, location, type, sub, _t)

Definition at line 343 of file TTL_cpp/TTL_tensors.h.

◆ TTL_sub_tensor() [5/5]

template<typename TENSORTYPE>

TTL_sub_tensor< TENSORTYPE >::TTL_sub_tensor ( )

inline

Create an empty tiled internal tensor. Empty means it has all dimensions set to zero.

Most operations on an empty tensor should turn into no-ops and so an empty tensor is the safest default state.

Definition at line 354 of file TTL_cpp/TTL_tensors.h.

Member Function Documentation

◆ empty()

template<typename TENSORTYPE>

bool TTL_sub_tensor< TENSORTYPE >::empty ( )

inline

A Tensor is empty if its shape is empty.

See also: TTL_shape_empty

Parameters

tensor The tensor to test for emptiness

Returns: true is the tensor is empty; false is the tensor is not empty

Definition at line 391 of file TTL_cpp/TTL_tensors.h.

◆ read()

template<typename TENSORTYPE>

const TENSORTYPE & TTL_sub_tensor< TENSORTYPE >::read	(	const unsigned int	x,
		const unsigned int	y = 0,
		const unsigned int	z = 0 ) const

inline

Read a value from a sub_tensor.

Parameters

x	The offset in the x dimension
y	The offset in the y dimension
z	The offset in the z dimension

No bounds checking is performed.

Returns: The value read

Definition at line 367 of file TTL_cpp/TTL_tensors.h.

◆ write()

template<typename TENSORTYPE>

TENSORTYPE TTL_sub_tensor< TENSORTYPE >::write	(	const TENSORTYPE	value,
		const unsigned int	x,
		const unsigned int	y = 0,
		const unsigned int	z = 0 )

inline

Write a value to a sub_tensor.

Parameters

value	The value to write
x	The offset in the x dimension
y	The offset in the y dimension
z	The offset in the z dimension

Definition at line 379 of file TTL_cpp/TTL_tensors.h.

Field Documentation

◆ origin

template<typename TENSORTYPE>

Origin TTL_sub_tensor< TENSORTYPE >::origin

Definition at line 408 of file TTL_cpp/TTL_tensors.h.

◆ tensor

template<typename TENSORTYPE>

TTL_tensor<TENSORTYPE> TTL_sub_tensor< TENSORTYPE >::tensor

Definition at line 407 of file TTL_cpp/TTL_tensors.h.

The documentation for this struct was generated from the following file:

TTL_cpp/TTL_tensors.h

Data Structures

Public Member Functions

Data Fields

Detailed Description

Constructor & Destructor Documentation

◆ TTL_sub_tensor() [1/5]

◆ TTL_sub_tensor() [2/5]

◆ TTL_sub_tensor() [3/5]

◆ TTL_sub_tensor() [4/5]

◆ TTL_sub_tensor() [5/5]

Member Function Documentation

◆ empty()

◆ read()

◆ write()

Field Documentation

◆ origin

◆ tensor