Apache Arrow (C++)
A columnar in-memory analytics layer designed to accelerate big data.
Public Member Functions | List of all members
arrow::gpu::CudaContext Class Reference

Friendlier interface to the CUDA driver API. More...

#include <arrow/gpu/cuda_context.h>

Inheritance diagram for arrow::gpu::CudaContext:

Public Member Functions

 ~CudaContext ()
 
Status Close ()
 
Status Allocate (int64_t nbytes, std::shared_ptr< CudaBuffer > *out)
 Allocate CUDA memory on GPU device for this context. More...
 
Status OpenIpcBuffer (const CudaIpcMemHandle &ipc_handle, std::shared_ptr< CudaBuffer > *buffer)
 Open existing CUDA IPC memory handle. More...
 
int64_t bytes_allocated () const
 

Detailed Description

Friendlier interface to the CUDA driver API.

Constructor & Destructor Documentation

◆ ~CudaContext()

arrow::gpu::CudaContext::~CudaContext ( )

Member Function Documentation

◆ Allocate()

Status arrow::gpu::CudaContext::Allocate ( int64_t  nbytes,
std::shared_ptr< CudaBuffer > *  out 
)

Allocate CUDA memory on GPU device for this context.

Parameters
[in]nbytesnumber of bytes
[out]outthe allocated buffer
Returns
Status

◆ bytes_allocated()

int64_t arrow::gpu::CudaContext::bytes_allocated ( ) const

◆ Close()

Status arrow::gpu::CudaContext::Close ( )

◆ OpenIpcBuffer()

Status arrow::gpu::CudaContext::OpenIpcBuffer ( const CudaIpcMemHandle ipc_handle,
std::shared_ptr< CudaBuffer > *  buffer 
)

Open existing CUDA IPC memory handle.

Parameters
[in]ipc_handleopaque pointer to CUipcMemHandle (driver API)
[out]buffera CudaBuffer referencing
Returns
Status

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