PVCuda::CudaPoolingDeliverKernel Class Reference

Inheritance diagram for PVCuda::CudaPoolingDeliverKernel:

Data Structures
struct	Params

Public Member Functions
	CudaPoolingDeliverKernel (CudaDevice *inDevice)
void	setArgs (PVLayerLoc const *preLoc, PVLayerLoc const *postLoc, int nxpPost, int nypPost, cudnnPoolingMode_t poolingMode, int multiplier, CudaBuffer *inputBuffer, CudaBuffer *outputBuffer, int channel)

Static Public Member Functions
static int	calcStride (int preRestricted, int postRestricted)

Protected Member Functions
int	calcBorderExcess (int preRestricted, int postRestricted, int border, int patchSizePostPerspective)
int	calcManyScale (int preRestricted, int postRestricted)
virtual int	do_run () override

Protected Attributes
int	mBorderExcessX = 0
int	mBorderExcessY = 0
CudaBuffer *	mCudnnDataStore = nullptr
CudaBuffer *	mCudnnGSyn = nullptr
float *	mDataStore = nullptr
cudnnTensorDescriptor_t	mDataStoreDescriptor = nullptr
float *	mGSyn = nullptr
cudnnTensorDescriptor_t	mGSynDescriptor = nullptr
float	mMultiplier = 1.0f
cudnnPoolingDescriptor_t	mPoolingDescriptor = nullptr
cudnnPoolingMode_t	mPoolingMode = CUDNN_POOLING_MAX
PVLayerLoc const *	mPostLoc = nullptr
PVLayerLoc const *	mPreLoc = nullptr

Detailed Description

Definition at line 16 of file CudaPoolingDeliverKernel.hpp.

---

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

• OpenPV/src/cudakernels/CudaPoolingDeliverKernel.hpp
• OpenPV/src/cudakernels/CudaPoolingDeliverKernel.cpp