ViennaCL - The Vienna Computing Library  1.5.2
Data Structures | Namespaces | Defines | Typedefs | Functions
viennacl/backend/cuda.hpp File Reference

Implementations for the CUDA backend functionality. More...

#include <iostream>
#include <vector>
#include <cassert>
#include "viennacl/tools/shared_ptr.hpp"
#include <cuda_runtime.h>

Go to the source code of this file.

Data Structures

struct  cuda_deleter< U >
 Functor for deleting a CUDA handle. Used within the smart pointer class. More...

Namespaces

namespace  viennacl
 

Main namespace in ViennaCL. Holds all the basic types such as vector, matrix, etc. and defines operations upon them.


namespace  viennacl::backend
 

Namespace providing routines for handling the different memory domains.


namespace  viennacl::backend::cuda
 

Provides implementations for handling CUDA memory buffers.


namespace  viennacl::backend::cuda::detail
 

Holds implementation details for handling CUDA memory buffers. Not intended for direct use by library users.


Defines

#define VIENNACL_CUDA_ERROR_CHECK(err)   detail::cuda_error_check (err, __FILE__, __LINE__)

Typedefs

typedef
viennacl::tools::shared_ptr
< char > 
handle_type

Functions

void cuda_error_check (cudaError error_code, const char *file, const int line)
handle_type memory_create (vcl_size_t size_in_bytes, const void *host_ptr=NULL)
 Creates an array of the specified size on the CUDA device. If the second argument is provided, the buffer is initialized with data from that pointer.
void memory_copy (handle_type const &src_buffer, handle_type &dst_buffer, vcl_size_t src_offset, vcl_size_t dst_offset, vcl_size_t bytes_to_copy)
 Copies 'bytes_to_copy' bytes from address 'src_buffer + src_offset' on the CUDA device to memory starting at address 'dst_buffer + dst_offset' on the same CUDA device.
void memory_write (handle_type &dst_buffer, vcl_size_t dst_offset, vcl_size_t bytes_to_copy, const void *ptr, bool async=false)
 Writes data from main RAM identified by 'ptr' to the CUDA buffer identified by 'dst_buffer'.
void memory_read (handle_type const &src_buffer, vcl_size_t src_offset, vcl_size_t bytes_to_copy, void *ptr, bool async=false)
 Reads data from a CUDA buffer back to main RAM.

Detailed Description

Implementations for the CUDA backend functionality.


Define Documentation

#define VIENNACL_CUDA_ERROR_CHECK (   err)    detail::cuda_error_check (err, __FILE__, __LINE__)