TNL::Backend::Stream Class Reference

Public Member Functions
	Stream ()=default
	Constructs a stream wrapping the CUDA 0 (NULL) stream.
	Stream (const Stream &other)=default
	Default copy-constructor.
	Stream (Stream &&other) noexcept=default
	Default move-constructor.
	operator const stream_t & () const
	Access the CUDA stream handle associated with this object.
Stream &	operator= (const Stream &other)=default
	Default copy-assignment operator.
Stream &	operator= (Stream &&other) noexcept=default
	Default move-assignment operator.

Static Public Member Functions
static Stream	create (unsigned int flags=StreamDefault, int priority=0)
	Creates a new stream.

Member Function Documentation

create()

Stream TNL::Backend::Stream::create ( unsigned int flags = StreamDefault, int priority = 0 )

inlinestatic

Creates a new stream.

The stream is created by calling cudaStreamCreateWithPriority with the following parameters:

Parameters

flags	Custom flags for stream creation. Possible values are: cudaStreamDefault: Default stream creation flag. cudaStreamNonBlocking: Specifies that work running in the created stream may run concurrently with work in stream 0 (the NULL stream), and that the created stream should perform no implicit synchronization with stream 0.
priority	Priority of the stream. Lower numbers represent higher priorities. See cudaDeviceGetStreamPriorityRange for more information about the meaningful stream priorities that can be passed.

operator const stream_t &()

TNL::Backend::Stream::operator const stream_t & ( ) const

inline

Access the CUDA stream handle associated with this object.

This routine permits the implicit conversion from Stream to stream_t.

Warning

The obtained stream_t handle becomes invalid when the originating Stream object is destroyed. For example, the following code is invalid, because the Stream object managing the lifetime of the stream_t handle is destroyed as soon as it is cast to stream_t:

const TNL::Backend::stream_t stream = TNL::Backend::Stream::create();
my_kernel<<< gridSize, blockSize, 0, stream >>>( args... );

---

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

• src/TNL/Backend/Stream.h