Point Cloud Library (PCL)  1.14.1-dev
List of all members | Public Types | Public Member Functions
pcl::gpu::DeviceArray2D< T > Class Template Reference

DeviceArray2D class More...

#include </__w/1/s/gpu/containers/include/pcl/gpu/containers/device_array.h>

+ Inheritance diagram for pcl::gpu::DeviceArray2D< T >:
+ Collaboration diagram for pcl::gpu::DeviceArray2D< T >:

Public Types

enum  { elem_size = sizeof(T) }
 Element size. More...
 
using type = T
 Element type. More...
 

Public Member Functions

 DeviceArray2D ()
 Empty constructor. More...
 
 DeviceArray2D (int rows, int cols)
 Allocates internal buffer in GPU memory. More...
 
 DeviceArray2D (int rows, int cols, void *data, std::size_t stepBytes)
 Initializes with user allocated buffer. More...
 
 DeviceArray2D (const DeviceArray2D &other)
 Copy constructor. More...
 
DeviceArray2Doperator= (const DeviceArray2D &other)
 Assignment operator. More...
 
void create (int rows, int cols)
 Allocates internal buffer in GPU memory. More...
 
void release ()
 Decrements reference counter and releases internal buffer if needed. More...
 
void copyTo (DeviceArray2D &other) const
 Performs data copying. More...
 
void upload (const void *host_ptr, std::size_t host_step, int rows, int cols)
 Uploads data to internal buffer in GPU memory. More...
 
void download (void *host_ptr, std::size_t host_step) const
 Downloads data from internal buffer to CPU memory. More...
 
void swap (DeviceArray2D &other_arg)
 Performs swap of data pointed with another device array. More...
 
template<class A >
void upload (const std::vector< T, A > &data, int cols)
 Uploads data to internal buffer in GPU memory. More...
 
template<class A >
void download (std::vector< T, A > &data, int &cols) const
 Downloads data from internal buffer to CPU memory. More...
 
T * ptr (int y=0)
 Returns pointer to given row in internal buffer. More...
 
const T * ptr (int y=0) const
 Returns const pointer to given row in internal buffer. More...
 
 operator T* ()
 Returns pointer for internal buffer in GPU memory. More...
 
 operator const T * () const
 Returns const pointer for internal buffer in GPU memory. More...
 
int cols () const
 Returns number of elements in each row. More...
 
int rows () const
 Returns number of rows. More...
 
std::size_t elem_step () const
 Returns step in elements. More...
 
- Public Member Functions inherited from pcl::gpu::DeviceMemory2D
 DeviceMemory2D ()
 Empty constructor. More...
 
 ~DeviceMemory2D ()
 Destructor. More...
 
 DeviceMemory2D (int rows_arg, int colsBytes_arg)
 Allocates internal buffer in GPU memory. More...
 
 DeviceMemory2D (int rows_arg, int colsBytes_arg, void *data_arg, std::size_t step_arg)
 Initializes with user allocated buffer. More...
 
 DeviceMemory2D (const DeviceMemory2D &other_arg)
 Copy constructor. More...
 
DeviceMemory2Doperator= (const DeviceMemory2D &other_arg)
 Assignment operator. More...
 
void create (int rows_arg, int colsBytes_arg)
 Allocates internal buffer in GPU memory. More...
 
void release ()
 Decrements reference counter and releases internal buffer if needed. More...
 
void copyTo (DeviceMemory2D &other) const
 Performs data copying. More...
 
void upload (const void *host_ptr_arg, std::size_t host_step_arg, int rows_arg, int colsBytes_arg)
 Uploads data to internal buffer in GPU memory. More...
 
void download (void *host_ptr_arg, std::size_t host_step_arg) const
 Downloads data from internal buffer to CPU memory. More...
 
void swap (DeviceMemory2D &other_arg)
 Performs swap of data pointed with another device memory. More...
 
template<class T >
T * ptr (int y_arg=0)
 Returns pointer to given row in internal buffer. More...
 
template<class T >
const T * ptr (int y_arg=0) const
 Returns constant pointer to given row in internal buffer. More...
 
template<class U >
 operator PtrStep< U > () const
 Conversion to PtrStep for passing to kernel functions. More...
 
template<class U >
 operator PtrStepSz< U > () const
 Conversion to PtrStepSz for passing to kernel functions. More...
 
bool empty () const
 Returns true if unallocated otherwise false. More...
 
int colsBytes () const
 Returns number of bytes in each row. More...
 
int rows () const
 Returns number of rows. More...
 
std::size_t step () const
 Returns stride between two consecutive rows in bytes for internal buffer. More...
 

Detailed Description

template<class T>
class pcl::gpu::DeviceArray2D< T >

DeviceArray2D class

Note
Typed container for pitched GPU memory with reference counting.
Author
Anatoly Baksheev

Definition at line 188 of file device_array.h.

Member Typedef Documentation

◆ type

template<class T >
using pcl::gpu::DeviceArray2D< T >::type = T

Element type.

Definition at line 191 of file device_array.h.

Member Enumeration Documentation

◆ anonymous enum

template<class T >
anonymous enum

Element size.

Enumerator
elem_size 

Definition at line 194 of file device_array.h.

Constructor & Destructor Documentation

◆ DeviceArray2D() [1/4]

template<class T >
pcl::gpu::DeviceArray2D< T >::DeviceArray2D
inline

Empty constructor.

Definition at line 189 of file device_array.hpp.

◆ DeviceArray2D() [2/4]

template<class T >
pcl::gpu::DeviceArray2D< T >::DeviceArray2D ( int  rows,
int  cols 
)
inline

Allocates internal buffer in GPU memory.

Parameters
rowsnumber of rows to allocate
colsnumber of elements in each row

Definition at line 193 of file device_array.hpp.

◆ DeviceArray2D() [3/4]

template<class T >
pcl::gpu::DeviceArray2D< T >::DeviceArray2D ( int  rows,
int  cols,
void *  data,
std::size_t  stepBytes 
)
inline

Initializes with user allocated buffer.

Reference counting is disabled in this case.

Parameters
rowsnumber of rows
colsnumber of elements in each row
datapointer to buffer
stepBytesstride between two consecutive rows in bytes

Definition at line 198 of file device_array.hpp.

◆ DeviceArray2D() [4/4]

template<class T >
pcl::gpu::DeviceArray2D< T >::DeviceArray2D ( const DeviceArray2D< T > &  other)
inline

Copy constructor.

Just increments reference counter.

Definition at line 206 of file device_array.hpp.

Member Function Documentation

◆ cols()

template<class T >
int pcl::gpu::DeviceArray2D< T >::cols
inline

Returns number of elements in each row.

Definition at line 310 of file device_array.hpp.

Referenced by pcl::gpu::AsyncCopy< T >::download(), pcl::gpu::TextureBinder::TextureBinder(), and pcl::gpu::AsyncCopy< T >::upload().

◆ copyTo()

template<class T >
void pcl::gpu::DeviceArray2D< T >::copyTo ( DeviceArray2D< T > &  other) const
inline

Performs data copying.

If destination size differs it will be reallocated.

Parameters
otherdestination container

Definition at line 234 of file device_array.hpp.

References pcl::gpu::DeviceMemory2D::copyTo().

◆ create()

template<class T >
void pcl::gpu::DeviceArray2D< T >::create ( int  rows,
int  cols 
)
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
rowsnumber of rows to allocate
colsnumber of elements in each row

Definition at line 220 of file device_array.hpp.

References pcl::gpu::DeviceMemory2D::create().

◆ download() [1/2]

template<class T >
template<class A >
void pcl::gpu::DeviceArray2D< T >::download ( std::vector< T, A > &  data,
int &  cols 
) const
inline

Downloads data from internal buffer to CPU memory.

Parameters
datahost vector to download to
colsOutput stride in elements between two consecutive rows for host vector.

Definition at line 267 of file device_array.hpp.

◆ download() [2/2]

template<class T >
void pcl::gpu::DeviceArray2D< T >::download ( void *  host_ptr,
std::size_t  host_step 
) const
inline

Downloads data from internal buffer to CPU memory.

User is responsible for correct host buffer size.

Parameters
host_ptrpointer to host buffer to download
host_stepstride between two consecutive rows in bytes for host buffer

Definition at line 251 of file device_array.hpp.

References pcl::gpu::DeviceMemory2D::download().

◆ elem_step()

template<class T >
std::size_t pcl::gpu::DeviceArray2D< T >::elem_step
inline

Returns step in elements.

Definition at line 324 of file device_array.hpp.

References pcl::gpu::DeviceMemory2D::step().

◆ operator const T *()

template<class T >
pcl::gpu::DeviceArray2D< T >::operator const T *
inline

Returns const pointer for internal buffer in GPU memory.

Definition at line 303 of file device_array.hpp.

◆ operator T*()

template<class T >
pcl::gpu::DeviceArray2D< T >::operator T*
inline

Returns pointer for internal buffer in GPU memory.

Definition at line 297 of file device_array.hpp.

◆ operator=()

template<class T >
DeviceArray2D< T > & pcl::gpu::DeviceArray2D< T >::operator= ( const DeviceArray2D< T > &  other)
inline

Assignment operator.

Just increments reference counter.

Definition at line 212 of file device_array.hpp.

◆ ptr() [1/2]

template<class T >
T * pcl::gpu::DeviceArray2D< T >::ptr ( int  y = 0)
inline

Returns pointer to given row in internal buffer.

Parameters
yrow index

Definition at line 284 of file device_array.hpp.

Referenced by pcl::gpu::AsyncCopy< T >::download(), pcl::gpu::TextureBinder::TextureBinder(), and pcl::gpu::AsyncCopy< T >::upload().

◆ ptr() [2/2]

template<class T >
const T * pcl::gpu::DeviceArray2D< T >::ptr ( int  y = 0) const
inline

Returns const pointer to given row in internal buffer.

Parameters
yrow index

Definition at line 291 of file device_array.hpp.

◆ release()

template<class T >
void pcl::gpu::DeviceArray2D< T >::release
inline

Decrements reference counter and releases internal buffer if needed.

Definition at line 227 of file device_array.hpp.

◆ rows()

template<class T >
int pcl::gpu::DeviceArray2D< T >::rows
inline

◆ swap()

template<class T >
void pcl::gpu::DeviceArray2D< T >::swap ( DeviceArray2D< T > &  other_arg)

Performs swap of data pointed with another device array.

Parameters
other_argdevice array to swap with

Definition at line 277 of file device_array.hpp.

◆ upload() [1/2]

template<class T >
template<class A >
void pcl::gpu::DeviceArray2D< T >::upload ( const std::vector< T, A > &  data,
int  cols 
)
inline

Uploads data to internal buffer in GPU memory.

It calls create() inside to ensure that internal buffer size is enough.

Parameters
datahost vector to upload from
colsstride in elements between two consecutive rows for host buffer

Definition at line 259 of file device_array.hpp.

◆ upload() [2/2]

template<class T >
void pcl::gpu::DeviceArray2D< T >::upload ( const void *  host_ptr,
std::size_t  host_step,
int  rows,
int  cols 
)
inline

Uploads data to internal buffer in GPU memory.

It calls create() inside to ensure that internal buffer size is enough.

Parameters
host_ptrpointer to host buffer to upload
host_stepstride between two consecutive rows in bytes for host buffer
rowsnumber of rows to upload
colsnumber of elements in each row

Definition at line 241 of file device_array.hpp.

References pcl::gpu::DeviceMemory2D::upload().


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