tf::cudaStreamBase< Creator, Deleter > Class Template Reference

class to create a CUDA stream with unique ownership More...

#include <taskflow/cuda/cuda_stream.hpp>

Inheritance diagram for tf::cudaStreamBase< Creator, Deleter >:

Collaboration diagram for tf::cudaStreamBase< Creator, Deleter >:

Public Types
using	base_type = std::unique_ptr<std::remove_pointer_t<cudaStream_t>, Deleter>
	base type for the underlying unique pointer

Public Member Functions
template<typename... ArgsT>
	cudaStreamBase (ArgsT &&... args)
	constructs a cudaStream object by passing the given arguments to the stream creator
	cudaStreamBase (cudaStreamBase &&)=default
	constructs a cudaStream from the given rhs using move semantics
cudaStreamBase &	operator= (cudaStreamBase &&)=default
	assign the rhs to *this using move semantics
cudaStreamBase &	synchronize ()
	synchronizes the associated stream
void	begin_capture (cudaStreamCaptureMode m=cudaStreamCaptureModeGlobal) const
	begins graph capturing on the stream
cudaGraph_t	end_capture () const
	ends graph capturing on the stream
void	record (cudaEvent_t event) const
	records an event on the stream
void	wait (cudaEvent_t event) const
	waits on an event
template<typename C, typename D>
cudaStreamBase &	run (const cudaGraphExecBase< C, D > &exec)
	runs the given executable CUDA graph
cudaStreamBase &	run (cudaGraphExec_t exec)
	runs the given executable CUDA graph

Detailed Description

template<typename Creator, typename Deleter>
class tf::cudaStreamBase< Creator, Deleter >

class to create a CUDA stream with unique ownership

Template Parameters

Creator	functor to create the stream (used in constructor)
Deleter	functor to delete the stream (used in destructor)

The cudaStream class encapsulates a cudaStream_t using std::unique_ptr, ensuring that CUDA events are properly created and destroyed with a unique ownership.

Member Typedef Documentation

base_type

template<typename Creator, typename Deleter>

using tf::cudaStreamBase< Creator, Deleter >::base_type = std::unique_ptr<std::remove_pointer_t<cudaStream_t>, Deleter>

base type for the underlying unique pointer

This alias provides a shorthand for the underlying std::unique_ptr type that manages CUDA stream resources with an associated deleter.

Constructor & Destructor Documentation

cudaStreamBase()

template<typename Creator, typename Deleter>

template<typename... ArgsT>

tf::cudaStreamBase< Creator, Deleter >::cudaStreamBase ( ArgsT &&... args )

inlineexplicit

constructs a cudaStream object by passing the given arguments to the stream creator

Constructs a cudaStream object by passing the given arguments to the stream creator

Parameters

args	arguments to pass to the stream creator

Member Function Documentation

begin_capture()

template<typename Creator, typename Deleter>

void tf::cudaStreamBase< Creator, Deleter >::begin_capture ( cudaStreamCaptureMode m = cudaStreamCaptureModeGlobal ) const

inline

begins graph capturing on the stream

When a stream is in capture mode, all operations pushed into the stream will not be executed, but will instead be captured into a graph, which will be returned via cudaStream::end_capture.

A thread's mode can be one of the following:

• cudaStreamCaptureModeGlobal: This is the default mode. If the local thread has an ongoing capture sequence that was not initiated with cudaStreamCaptureModeRelaxed at cuStreamBeginCapture, or if any other thread has a concurrent capture sequence initiated with cudaStreamCaptureModeGlobal, this thread is prohibited from potentially unsafe API calls.
• cudaStreamCaptureModeThreadLocal: If the local thread has an ongoing capture sequence not initiated with cudaStreamCaptureModeRelaxed, it is prohibited from potentially unsafe API calls. Concurrent capture sequences in other threads are ignored.
• cudaStreamCaptureModeRelaxed: The local thread is not prohibited from potentially unsafe API calls. Note that the thread is still prohibited from API calls which necessarily conflict with stream capture, for example, attempting cudaEventQuery on an event that was last recorded inside a capture sequence.

end_capture()

template<typename Creator, typename Deleter>

cudaGraph_t tf::cudaStreamBase< Creator, Deleter >::end_capture ( ) const

inline

ends graph capturing on the stream

Equivalently calling cudaStreamEndCapture to end capture on stream and returning the captured graph. Capture must have been initiated on stream via a call to cudaStream::begin_capture. If capture was invalidated, due to a violation of the rules of stream capture, then a NULL graph will be returned.

record()

template<typename Creator, typename Deleter>

void tf::cudaStreamBase< Creator, Deleter >::record ( cudaEvent_t event ) const

inline

records an event on the stream

Equivalently calling cudaEventRecord to record an event on this stream, both of which must be on the same CUDA context.

run() [1/2]

template<typename Creator, typename Deleter>

template<typename C, typename D>

cudaStreamBase & tf::cudaStreamBase< Creator, Deleter >::run

(

const cudaGraphExecBase< C, D > &

exec

)

runs the given executable CUDA graph

Parameters

exec	the given cudaGraphExec

run() [2/2]

template<typename SC, typename SD>

cudaStreamBase< SC, SD > & tf::cudaStreamBase< SC, SD >::run

(

cudaGraphExec_t

exec

)

runs the given executable CUDA graph

Parameters

exec	the given cudaGraphExec_t

synchronize()

template<typename Creator, typename Deleter>

cudaStreamBase & tf::cudaStreamBase< Creator, Deleter >::synchronize ( )

inline

synchronizes the associated stream

Equivalently calling cudaStreamSynchronize to block until this stream has completed all operations.

wait()

template<typename Creator, typename Deleter>

void tf::cudaStreamBase< Creator, Deleter >::wait ( cudaEvent_t event ) const

inline

waits on an event

Equivalently calling cudaStreamWaitEvent to make all future work submitted to stream wait for all work captured in event.

---

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

• taskflow/cuda/cuda_stream.hpp
• taskflow/cuda/cuda_graph_exec.hpp