cub::GridQueue< OffsetT > Class Template Reference

Detailed description

template< typename OffsetT>
class cub::GridQueue< OffsetT >

GridQueue is a descriptor utility for dynamic queue management.

Overview

GridQueue descriptors provides abstractions for "filling" or "draining" globally-shared vectors.

A "filling" GridQueue works by atomically-adding to a zero-initialized counter, returning a unique offset for the calling thread to write its items. The GridQueue maintains the total "fill-size". The fill counter must be reset using GridQueue::ResetFill by the host or kernel instance prior to the kernel instance that will be filling.

Similarly, a "draining" GridQueue works by works by atomically-incrementing a zero-initialized counter, returning a unique offset for the calling thread to read its items. Threads can safely drain until the array's logical fill-size is exceeded. The drain counter must be reset using GridQueue::ResetDrain or GridQueue::FillAndResetDrain by the host or kernel instance prior to the kernel instance that will be filling. (For dynamic work distribution of existing data, the corresponding fill-size is simply the number of elements in the array.)

Iterative work management can be implemented simply with a pair of flip-flopping work buffers, each with an associated set of fill and drain GridQueue descriptors.

Template Parameters

OffsetT

Signed integer type for global offsets

Public Methods
__host__ __device__ __forceinline__	GridQueue ()
	Constructs an invalid GridQueue descriptor. More...
__host__ __device__ __forceinline__	GridQueue (void *d_storage)
	Constructs a GridQueue descriptor around the device storage allocation. More...
__host__ __device__ __forceinline__ cudaError_t	FillAndResetDrain (OffsetT fill_size, cudaStream_t stream=0)
	This operation sets the fill-size and resets the drain counter, preparing the GridQueue for draining in the next kernel instance. To be called by the host or by a kernel prior to that which will be draining. More...
__host__ __device__ __forceinline__ cudaError_t	ResetDrain (cudaStream_t stream=0)
	This operation resets the drain so that it may advance to meet the existing fill-size. To be called by the host or by a kernel prior to that which will be draining. More...
__host__ __device__ __forceinline__ cudaError_t	ResetFill (cudaStream_t stream=0)
	This operation resets the fill counter. To be called by the host or by a kernel prior to that which will be filling. More...
__host__ __device__ __forceinline__ cudaError_t	FillSize (OffsetT &fill_size, cudaStream_t stream=0)
	Returns the fill-size established by the parent or by the previous kernel. More...
__device__ __forceinline__ OffsetT	Drain (OffsetT num_items)
	Drain num_items from the queue. Returns offset from which to read items. To be called from CUDA kernel. More...
__device__ __forceinline__ OffsetT	Fill (OffsetT num_items)
	Fill num_items into the queue. Returns offset from which to write items. To be called from CUDA kernel. More...

Static Public Methods
__host__ __device__ static __forceinline__ size_t	AllocationSize ()
	Returns the device allocation size in bytes needed to construct a GridQueue instance. More...

Constructor & Destructor Documentation

template<typename OffsetT >

__host__ __device__ __forceinline__ cub::GridQueue< OffsetT >::GridQueue ( )

inline

Constructs an invalid GridQueue descriptor.

template<typename OffsetT >

__host__ __device__ __forceinline__ cub::GridQueue< OffsetT >::GridQueue ( void *  d_storage)

inline

Constructs a GridQueue descriptor around the device storage allocation.

Parameters

d_storage

Device allocation to back the GridQueue. Must be at least as big as AllocationSize().

Member Function Documentation

template<typename OffsetT >

__host__ __device__ static __forceinline__ size_t cub::GridQueue< OffsetT >::AllocationSize ( )

inlinestatic

Returns the device allocation size in bytes needed to construct a GridQueue instance.

template<typename OffsetT >

__host__ __device__ __forceinline__ cudaError_t cub::GridQueue< OffsetT >::FillAndResetDrain ( OffsetT  fill_size, cudaStream_t  stream = 0  )

inline

This operation sets the fill-size and resets the drain counter, preparing the GridQueue for draining in the next kernel instance. To be called by the host or by a kernel prior to that which will be draining.

template<typename OffsetT >

__host__ __device__ __forceinline__ cudaError_t cub::GridQueue< OffsetT >::ResetDrain ( cudaStream_t  stream = 0)

inline

This operation resets the drain so that it may advance to meet the existing fill-size. To be called by the host or by a kernel prior to that which will be draining.

template<typename OffsetT >

__host__ __device__ __forceinline__ cudaError_t cub::GridQueue< OffsetT >::ResetFill ( cudaStream_t  stream = 0)

inline

This operation resets the fill counter. To be called by the host or by a kernel prior to that which will be filling.

template<typename OffsetT >

__host__ __device__ __forceinline__ cudaError_t cub::GridQueue< OffsetT >::FillSize ( OffsetT &  fill_size, cudaStream_t  stream = 0  )

inline

Returns the fill-size established by the parent or by the previous kernel.

template<typename OffsetT >

__device__ __forceinline__ OffsetT cub::GridQueue< OffsetT >::Drain ( OffsetT  num_items)

inline

Drain num_items from the queue. Returns offset from which to read items. To be called from CUDA kernel.

template<typename OffsetT >

__device__ __forceinline__ OffsetT cub::GridQueue< OffsetT >::Fill ( OffsetT  num_items)

inline

Fill num_items into the queue. Returns offset from which to write items. To be called from CUDA kernel.

---

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

• grid_queue.cuh