cv::cuda::Stream Class Reference

This class encapsulates a queue of asynchronous calls. More...

#include <opencv2/core/cuda.hpp>

Collaboration diagram for cv::cuda::Stream:

Public Types
typedef void(*	StreamCallback) (int status, void *userData)

Public Member Functions
	Stream ()
	creates a new asynchronous stream
	Stream (const Ptr< GpuMat::Allocator > &allocator)
	creates a new asynchronous stream with custom allocator
	Stream (const size_t cudaFlags)
	creates a new Stream using the cudaFlags argument to determine the behaviors of the stream
void *	cudaPtr () const
	return Pointer to CUDA stream
void	enqueueHostCallback (StreamCallback callback, void *userData)
	Adds a callback to be called on the host after all currently enqueued items in the stream have completed.
	operator bool_type () const
	returns true if stream object is not default (!= 0)
bool	queryIfComplete () const
	Returns true if the current stream queue is finished. Otherwise, it returns false.
void	waitEvent (const Event &event)
	Makes a compute stream wait on an event.
void	waitForCompletion ()
	Blocks the current CPU thread until all operations in the stream are complete.

Static Public Member Functions
static Stream &	Null ()
	return Stream object for default CUDA stream

Friends
class	BufferPool
class	DefaultDeviceInitializer
struct	StreamAccessor

Detailed Description

This class encapsulates a queue of asynchronous calls.

Note

Currently, you may face problems if an operation is enqueued twice with different data. Some functions use the constant GPU memory, and next call may update the memory before the previous one has been finished. But calling different operations asynchronously is safe because each operation has its own constant buffer. Memory copy/upload/download/set operations to the buffers you hold are also safe.

The Stream class is not thread-safe. Please use different Stream objects for different CPU threads.

void thread1()
{
    cv::cuda::Stream stream1;
    cv::cuda::func1(..., stream1);
}
 
void thread2()
{
    cv::cuda::Stream stream2;
    cv::cuda::func2(..., stream2);
}

Note

By default all CUDA routines are launched in Stream::Null() object, if the stream is not specified by user. In multi-threading environment the stream objects must be passed explicitly (see previous note).

Member Typedef Documentation

StreamCallback

typedef void(* cv::cuda::Stream::StreamCallback) (int status, void *userData)

Constructor & Destructor Documentation

Stream() [1/3]

cv::cuda::Stream::Stream

(

)

Python:
	cv.cuda.Stream(		) ->	<cuda_Stream object>
	cv.cuda.Stream(	allocator	) ->	<cuda_Stream object>
	cv.cuda.Stream(	cudaFlags	) ->	<cuda_Stream object>

creates a new asynchronous stream

Stream() [2/3]

cv::cuda::Stream::Stream

(

const Ptr< GpuMat::Allocator > &

allocator

)

Python:
	cv.cuda.Stream(		) ->	<cuda_Stream object>
	cv.cuda.Stream(	allocator	) ->	<cuda_Stream object>
	cv.cuda.Stream(	cudaFlags	) ->	<cuda_Stream object>

creates a new asynchronous stream with custom allocator

Stream() [3/3]

cv::cuda::Stream::Stream

(

const size_t

cudaFlags

)

Python:
	cv.cuda.Stream(		) ->	<cuda_Stream object>
	cv.cuda.Stream(	allocator	) ->	<cuda_Stream object>
	cv.cuda.Stream(	cudaFlags	) ->	<cuda_Stream object>

creates a new Stream using the cudaFlags argument to determine the behaviors of the stream

Note

The cudaFlags parameter is passed to the underlying api cudaStreamCreateWithFlags() and supports the same parameter values.

// creates an OpenCV cuda::Stream that manages an asynchronous, non-blocking,
// non-default CUDA stream
cv::cuda::Stream cvStream(cudaStreamNonBlocking);

Member Function Documentation

cudaPtr()

void * cv::cuda::Stream::cudaPtr

(

)

const

Python:
	cv.cuda.Stream.cudaPtr(		) ->	retval

return Pointer to CUDA stream

enqueueHostCallback()

void cv::cuda::Stream::enqueueHostCallback	(	StreamCallback	callback,
		void *	userData
	)

Adds a callback to be called on the host after all currently enqueued items in the stream have completed.

Note

Callbacks must not make any CUDA API calls. Callbacks must not perform any synchronization that may depend on outstanding device work or other callbacks that are not mandated to run earlier. Callbacks without a mandated order (in independent streams) execute in undefined order and may be serialized.

Null()

static Stream & cv::cuda::Stream::Null ( )

static

Python:
	cv.cuda.Stream.Null(		) ->	retval
	cv.cuda.Stream_Null(		) ->	retval

return Stream object for default CUDA stream

operator bool_type()

cv::cuda::Stream::operator bool_type

(

)

const

returns true if stream object is not default (!= 0)

queryIfComplete()

bool cv::cuda::Stream::queryIfComplete

(

)

const

Python:
	cv.cuda.Stream.queryIfComplete(		) ->	retval

Returns true if the current stream queue is finished. Otherwise, it returns false.

waitEvent()

void cv::cuda::Stream::waitEvent

(

const Event &

event

)

Python:
	cv.cuda.Stream.waitEvent(	event	) ->	None

Makes a compute stream wait on an event.

waitForCompletion()

void cv::cuda::Stream::waitForCompletion

(

)

Python:
	cv.cuda.Stream.waitForCompletion(		) ->	None

Blocks the current CPU thread until all operations in the stream are complete.

Friends And Related Symbol Documentation

BufferPool

friend class BufferPool

friend

DefaultDeviceInitializer

friend class DefaultDeviceInitializer

friend

StreamAccessor

friend struct StreamAccessor

friend

---

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

• opencv2/core/cuda.hpp