tf::cudaDeviceAllocator< T > Class Template Reference

class to create a CUDA device allocator More...

#include <cuda_memory.hpp>

Classes
struct	rebind
	its member type U is the equivalent allocator type to allocate elements of type U More...

Public Types
using	value_type = T
	element type
using	pointer = T *
	element pointer type
using	reference = T &
	element reference type
using	const_pointer = const T *
	const element pointer type
using	const_reference = const T &
	constant element reference type
using	size_type = std::size_t
	size type
using	difference_type = std::ptrdiff_t
	pointer difference type

Public Member Functions
	cudaDeviceAllocator () noexcept
	Constructs a device allocator object.
	cudaDeviceAllocator (const cudaDeviceAllocator &) noexcept
	Constructs a device allocator object from another device allocator object.
template<typename U >
	cudaDeviceAllocator (const cudaDeviceAllocator< U > &) noexcept
	Constructs a device allocator object from another device allocator object with a different element type.
	~cudaDeviceAllocator () noexcept
	Destructs the device allocator object.
pointer	address (reference x)
	Returns the address of x.
const_pointer	address (const_reference x) const
	Returns the address of x.
pointer	allocate (size_type n, std::allocator< void >::const_pointer=0)
	allocates block of storage.
void	deallocate (pointer ptr, size_type)
	Releases a block of storage previously allocated with member allocate and not yet released.
size_type	max_size () const noexcept
	returns the maximum number of elements that could potentially be allocated by this allocator
void	construct (pointer, const_reference)
	ignored to avoid de-referencing device pointer from the host
void	destroy (pointer)
	ignored to avoid de-referencing device pointer from the host
template<typename U >
bool	operator== (const cudaDeviceAllocator< U > &) const noexcept
	compares two allocator of different types using ==
template<typename U >
bool	operator!= (const cudaDeviceAllocator< U > &) const noexcept
	compares two allocator of different types using !=

Detailed Description

template<typename T>
class tf::cudaDeviceAllocator< T >

class to create a CUDA device allocator

Template Parameters

T	element type

A cudaDeviceAllocator enables device-specific allocation for standard library containers. It is typically passed as template parameter when declaring standard library containers (e.g. std::vector).

Member Function Documentation

address() [1/2]

template<typename T >

const_pointer tf::cudaDeviceAllocator< T >::address ( const_reference  x ) const

inline

Returns the address of x.

This effectively means returning &x.

Parameters

x	reference to an object

Returns

a pointer to the object

address() [2/2]

template<typename T >

pointer tf::cudaDeviceAllocator< T >::address ( reference  x )

inline

Returns the address of x.

This effectively means returning &x.

Parameters

x	reference to an object

Returns

a pointer to the object

allocate()

template<typename T >

pointer tf::cudaDeviceAllocator< T >::allocate ( size_type  n, std::allocator< void >::const_pointer  = 0  )

inline

allocates block of storage.

Attempts to allocate a block of storage with a size large enough to contain n elements of member type, value_type, and returns a pointer to the first element.

The storage is aligned appropriately for object of type value_type, but they are not constructed.

The block of storage is allocated using cudaMalloc and throws std::bad_alloc if it cannot allocate the total amount of storage requested.

Parameters

n	number of elements (each of size sizeof(value_type)) to be allocated

Returns

a pointer to the initial element in the block of storage.

deallocate()

template<typename T >

void tf::cudaDeviceAllocator< T >::deallocate ( pointer  ptr, size_type    )

inline

Releases a block of storage previously allocated with member allocate and not yet released.

The elements in the array are not destroyed by a call to this member function.

Parameters

ptr	pointer to a block of storage previously allocated with allocate

max_size()

template<typename T >

size_type tf::cudaDeviceAllocator< T >::max_size ( ) const

inlinenoexcept

returns the maximum number of elements that could potentially be allocated by this allocator

A call to member allocate with the value returned by this function can still fail to allocate the requested storage.

Returns

the nubmer of elements that might be allcoated as maximum by a call to member allocate

operator!=()

template<typename T >

template<typename U >

bool tf::cudaDeviceAllocator< T >::operator!= ( const cudaDeviceAllocator< U > &  ) const

inlinenoexcept

compares two allocator of different types using !=

Device allocators of different types are always equal to each other because the storage allocated by the allocator a1 can be deallocated through a2.

operator==()

template<typename T >

template<typename U >

bool tf::cudaDeviceAllocator< T >::operator== ( const cudaDeviceAllocator< U > &  ) const

inlinenoexcept

compares two allocator of different types using ==

Device allocators of different types are always equal to each other because the storage allocated by the allocator a1 can be deallocated through a2.

---

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

• cuda_memory.hpp