#include <device_array.hpp>

Inheritance diagram for kfusion::cuda::DeviceArray< T >:

Public Types
enum	{ elem_size = sizeof(T) }
	Element size. More...

typedef T	type
	Element type. More...

Public Member Functions
void	copyTo (DeviceArray &other) const
	Performs data copying. If destination size differs it will be reallocated. More...

void	create (size_t size)
	Allocates internal buffer in GPU memory. If internal buffer was created before the function recreates it with new size. If new and old sizes are equal it does nothing. More...

	DeviceArray ()
	Empty constructor. More...

	DeviceArray (size_t size)
	Allocates internal buffer in GPU memory. More...

	DeviceArray (T *ptr, size_t size)
	Initializes with user allocated buffer. Reference counting is disabled in this case. More...

	DeviceArray (const DeviceArray &other)
	Copy constructor. Just increments reference counter. More...

void	download (T *host_ptr) const
	Downloads data from internal buffer to CPU memory. More...

template<typename A >
void	download (std::vector< T, A > &data) const
	Downloads data from internal buffer to CPU memory. More...

	operator const T * () const
	Returns const pointer for internal buffer in GPU memory. More...

	operator T* ()
	Returns pointer for internal buffer in GPU memory. More...

DeviceArray &	operator= (const DeviceArray &other)
	Assigment operator. Just increments reference counter. More...

T *	ptr ()
	Returns pointer for internal buffer in GPU memory. More...

const T *	ptr () const
	Returns const pointer for internal buffer in GPU memory. More...

void	release ()
	Decrements reference counter and releases internal buffer if needed. More...

size_t	size () const
	Returns size in elements. More...

void	swap (DeviceArray &other_arg)
	Performs swap of data pointed with another device array. More...

void	upload (const T *host_ptr, size_t size)
	Uploads data to internal buffer in GPU memory. It calls create() inside to ensure that intenal buffer size is enough. More...

template<class A >
void	upload (const std::vector< T, A > &data)
	Uploads data to internal buffer in GPU memory. It calls create() inside to ensure that intenal buffer size is enough. More...

Public Member Functions inherited from kfusion::cuda::DeviceMemory
void	copyTo (DeviceMemory &other) const
	Performs data copying. If destination size differs it will be reallocated. More...

void	create (size_t sizeBytes_arg)
	Allocates internal buffer in GPU memory. If internal buffer was created before the function recreates it with new size. If new and old sizes are equal it does nothing. More...

	DeviceMemory ()
	Empty constructor. More...

	DeviceMemory (size_t sizeBytes_arg)
	Allocates internal buffer in GPU memory. More...

	DeviceMemory (void *ptr_arg, size_t sizeBytes_arg)
	Initializes with user allocated buffer. Reference counting is disabled in this case. More...

	DeviceMemory (const DeviceMemory &other_arg)
	Copy constructor. Just increments reference counter. More...

void	download (void *host_ptr_arg) const
	Downloads data from internal buffer to CPU memory. More...

bool	empty () const
	Returns true if unallocated otherwise false. More...

template<class U >
	operator PtrSz< U > () const
	Conversion to PtrSz for passing to kernel functions. More...

DeviceMemory &	operator= (const DeviceMemory &other_arg)
	Assigment operator. Just increments reference counter. More...

template<class T >
T *	ptr ()
	Returns pointer for internal buffer in GPU memory. More...

template<class T >
const T *	ptr () const
	Returns constant pointer for internal buffer in GPU memory. More...

void	release ()
	Decrements reference counter and releases internal buffer if needed. More...

size_t	sizeBytes () const

void	swap (DeviceMemory &other_arg)
	Performs swap of data pointed with another device memory. More...

void	upload (const void *host_ptr_arg, size_t sizeBytes_arg)
	Uploads data to internal buffer in GPU memory. It calls create() inside to ensure that intenal buffer size is enough. More...

	~DeviceMemory ()
	Destructor. More...

Detailed Description

template<class T>
class kfusion::cuda::DeviceArray< T >

DeviceArray class

Note: Typed container for GPU memory with reference counting.

Author: Anatoly Baksheev

Definition at line 20 of file device_array.hpp.

Member Typedef Documentation

◆ type

template<class T>

typedef T kfusion::cuda::DeviceArray< T >::type

Element type.

Definition at line 24 of file device_array.hpp.

Member Enumeration Documentation

◆ anonymous enum

template<class T>

anonymous enum

Element size.

Enumerator
elem_size

Definition at line 27 of file device_array.hpp.

Constructor & Destructor Documentation

◆ DeviceArray() [1/4]

template<class T >

kfusion::cuda::DeviceArray< T >::DeviceArray ( )

inline

Empty constructor.

Definition at line 234 of file device_array.hpp.

◆ DeviceArray() [2/4]

template<class T >

kfusion::cuda::DeviceArray< T >::DeviceArray ( size_t size )

inline

Allocates internal buffer in GPU memory.

Parameters

size_t number of elements to allocate

Definition at line 235 of file device_array.hpp.

◆ DeviceArray() [3/4]

template<class T>

kfusion::cuda::DeviceArray< T >::DeviceArray	(	T *	ptr,
		size_t	size
	)

inline

Initializes with user allocated buffer. Reference counting is disabled in this case.

Parameters

ptr	pointer to buffer
size	elemens number

Definition at line 236 of file device_array.hpp.

◆ DeviceArray() [4/4]

template<class T>

kfusion::cuda::DeviceArray< T >::DeviceArray ( const DeviceArray< T > & other )

inline

Copy constructor. Just increments reference counter.

Definition at line 237 of file device_array.hpp.

Member Function Documentation

◆ copyTo()

template<class T >

void kfusion::cuda::DeviceArray< T >::copyTo ( DeviceArray< T > & other ) const

inline

Performs data copying. If destination size differs it will be reallocated.

Parameters

other_arg destination container

Definition at line 246 of file device_array.hpp.

◆ create()

template<class T >

void kfusion::cuda::DeviceArray< T >::create ( size_t size )

inline

Allocates internal buffer in GPU memory. If internal buffer was created before the function recreates it with new size. If new and old sizes are equal it does nothing.

Parameters

size	elemens number

Definition at line 241 of file device_array.hpp.

◆ download() [1/2]

template<class T>

void kfusion::cuda::DeviceArray< T >::download ( T * host_ptr ) const

inline

Downloads data from internal buffer to CPU memory.

Parameters

host_ptr_arg pointer to buffer to download

Definition at line 250 of file device_array.hpp.

◆ download() [2/2]

template<class T>

template<class A >

void kfusion::cuda::DeviceArray< T >::download ( std::vector< T, A > & data ) const

inline

Downloads data from internal buffer to CPU memory.

Parameters

data	host vector to download to

Definition at line 263 of file device_array.hpp.

◆ operator const T *()

template<class T >

kfusion::cuda::DeviceArray< T >::operator const T * ( ) const

inline

Returns const pointer for internal buffer in GPU memory.

Definition at line 256 of file device_array.hpp.

◆ operator T*()

template<class T >

kfusion::cuda::DeviceArray< T >::operator T* ( )

inline

Returns pointer for internal buffer in GPU memory.

Definition at line 255 of file device_array.hpp.

◆ operator=()

template<class T >

kfusion::cuda::DeviceArray< T > & kfusion::cuda::DeviceArray< T >::operator= ( const DeviceArray< T > & other )

inline

Assigment operator. Just increments reference counter.

Definition at line 238 of file device_array.hpp.

◆ ptr() [1/2]

template<class T >

T * kfusion::cuda::DeviceArray< T >::ptr ( )

inline

Returns pointer for internal buffer in GPU memory.

Definition at line 259 of file device_array.hpp.

◆ ptr() [2/2]

template<class T >

const T * kfusion::cuda::DeviceArray< T >::ptr ( ) const

inline

Returns const pointer for internal buffer in GPU memory.

Definition at line 260 of file device_array.hpp.

◆ release()

template<class T >

void kfusion::cuda::DeviceArray< T >::release ( )

inline

Decrements reference counter and releases internal buffer if needed.

Definition at line 243 of file device_array.hpp.

◆ size()

template<class T >

size_t kfusion::cuda::DeviceArray< T >::size ( ) const

inline

Returns size in elements.

Definition at line 257 of file device_array.hpp.

◆ swap()

template<class T >

void kfusion::cuda::DeviceArray< T >::swap ( DeviceArray< T > & other_arg )

Performs swap of data pointed with another device array.

Parameters

other device array to swap with

Definition at line 253 of file device_array.hpp.

◆ upload() [1/2]

template<class T>

void kfusion::cuda::DeviceArray< T >::upload	(	const T *	host_ptr,
		size_t	size
	)

inline

Uploads data to internal buffer in GPU memory. It calls create() inside to ensure that intenal buffer size is enough.

Parameters

host_ptr_arg	pointer to buffer to upload
size	elemens number

Definition at line 248 of file device_array.hpp.

◆ upload() [2/2]

template<class T>

template<class A >

void kfusion::cuda::DeviceArray< T >::upload ( const std::vector< T, A > & data )

inline

Uploads data to internal buffer in GPU memory. It calls create() inside to ensure that intenal buffer size is enough.

Parameters

data	host vector to upload from

Definition at line 262 of file device_array.hpp.

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

device_array.hpp

Public Types

Public Member Functions

Detailed Description

template<class T> class kfusion::cuda::DeviceArray< T >

Member Typedef Documentation

◆ type

Member Enumeration Documentation

◆ anonymous enum

Constructor & Destructor Documentation

◆ DeviceArray() [1/4]

◆ DeviceArray() [2/4]

◆ DeviceArray() [3/4]

◆ DeviceArray() [4/4]

Member Function Documentation

◆ copyTo()

◆ create()

◆ download() [1/2]

◆ download() [2/2]

◆ operator const T *()

◆ operator T*()

◆ operator=()

◆ ptr() [1/2]

◆ ptr() [2/2]

◆ release()

◆ size()

◆ swap()

◆ upload() [1/2]

◆ upload() [2/2]

template<class T>
class kfusion::cuda::DeviceArray< T >