CpuWinogradConv2dTransformOutputKernel Class Reference

#include <CpuWinogradConv2dKernel.h>

Collaboration diagram for CpuWinogradConv2dTransformOutputKernel:

Public Member Functions
	CpuWinogradConv2dTransformOutputKernel (const CpuWinogradConv2dTransformOutputKernel &)=delete
	Prevent instances of this class from being copied (As this class contains pointers) More...
CpuWinogradConv2dTransformOutputKernel &	operator= (const CpuWinogradConv2dTransformOutputKernel &)=delete
	Prevent instances of this class from being copied (As this class contains pointers) More...
	CpuWinogradConv2dTransformOutputKernel (CpuWinogradConv2dTransformOutputKernel &&)=delete
	Prevent instances of this class from being moved it contains references. More...
CpuWinogradConv2dTransformOutputKernel &	operator= (CpuWinogradConv2dTransformOutputKernel &&)=delete
	Prevent instances of this class from being moved it contains references. More...
	CpuWinogradConv2dTransformOutputKernel (arm_conv::winograd::WinogradImpl &w_impl, arm_conv::ConvolutionArgs &_c_args, uint32_t nthreads)
void	run_op (ITensorPack &tensors, const Window &window, const ThreadInfo &info) override
	Execute the kernel on the passed window. More...
const char *	name () const override
	Name of the kernel. More...
Public Member Functions inherited from ICPPKernel
virtual	~ICPPKernel ()=default
	Default destructor. More...
virtual void	run (const Window &window, const ThreadInfo &info)
	Execute the kernel on the passed window. More...
virtual void	run_nd (const Window &window, const ThreadInfo &info, const Window &thread_locator)
	legacy compatibility layer for implemantions which do not support thread_locator In these cases we simply narrow the interface down the legacy version More...
virtual size_t	get_mws (const CPUInfo &platform, size_t thread_count) const
	Return minimum workload size of the relevant kernel. More...
Public Member Functions inherited from IKernel
	IKernel ()
	Constructor. More...
virtual	~IKernel ()=default
	Destructor. More...
virtual bool	is_parallelisable () const
	Indicates whether or not the kernel is parallelisable. More...
virtual BorderSize	border_size () const
	The size of the border for that kernel. More...
const Window &	window () const
	The maximum window the kernel can be executed on. More...
bool	is_window_configured () const
	Function to check if the embedded window of this kernel has been configured. More...

Additional Inherited Members
Static Public Member Functions inherited from ICpuKernel< CpuWinogradConv2dTransformOutputKernel >
static const auto *	get_implementation (const SelectorType &selector, KernelSelectionType selection_type=KernelSelectionType::Supported)
	Micro-kernel selector. More...
Static Public Attributes inherited from ICPPKernel
static constexpr size_t	default_mws = 1

Detailed Description

Definition at line 74 of file CpuWinogradConv2dKernel.h.

Constructor & Destructor Documentation

CpuWinogradConv2dTransformOutputKernel() [1/3]

CpuWinogradConv2dTransformOutputKernel ( const CpuWinogradConv2dTransformOutputKernel &  )

delete

Prevent instances of this class from being copied (As this class contains pointers)

CpuWinogradConv2dTransformOutputKernel() [2/3]

CpuWinogradConv2dTransformOutputKernel ( CpuWinogradConv2dTransformOutputKernel &&  )

delete

Prevent instances of this class from being moved it contains references.

CpuWinogradConv2dTransformOutputKernel() [3/3]

CpuWinogradConv2dTransformOutputKernel	(	arm_conv::winograd::WinogradImpl &	w_impl,
		arm_conv::ConvolutionArgs &	_c_args,
		uint32_t	nthreads
	)

Definition at line 64 of file CpuWinogradConv2dKernel.cpp.

    : _winograd_impl{w_impl}, _conv_args{_c_args}, _nthreads{nthreads}
{
}

Member Function Documentation

name()

const char* name ( ) const

inlineoverridevirtual

Name of the kernel.

Returns

Kernel name

Implements ICPPKernel.

Definition at line 96 of file CpuWinogradConv2dKernel.h.

    {
        return "CpuWinogradConv2dTransformOutputKernel";
    }

operator=() [1/2]

CpuWinogradConv2dTransformOutputKernel& operator= ( const CpuWinogradConv2dTransformOutputKernel &  )

delete

Prevent instances of this class from being copied (As this class contains pointers)

operator=() [2/2]

CpuWinogradConv2dTransformOutputKernel& operator= ( CpuWinogradConv2dTransformOutputKernel &&  )

delete

Prevent instances of this class from being moved it contains references.

run_op()

void run_op ( ITensorPack &  tensors, const Window &  window, const ThreadInfo &  info  )

overridevirtual

Execute the kernel on the passed window.

Warning

If is_parallelisable() returns false then the passed window must be equal to window()

Note

The window has to be a region within the window returned by the window() method

The width of the window has to be a multiple of num_elems_processed_per_iteration().

Parameters

[in]	tensors	A vector containing the tensors to operate on.
[in]	window	Region on which to execute the kernel. (Must be a region of the window returned by window())
[in]	info	Info about executing thread and CPU.

Reimplemented from ICPPKernel.

Definition at line 72 of file CpuWinogradConv2dKernel.cpp.

{
    ARM_COMPUTE_UNUSED(window);
    const ITensor *dst_nhwc                  = tensors.get_const_tensor(TensorType::ACL_DST);
    const ITensor *winograd_output_transform = tensors.get_const_tensor(TensorType::ACL_SRC_0);
    const ITensor *biases                    = tensors.get_const_tensor(TensorType::ACL_SRC_1);
    const ITensor *workspace                 = tensors.get_tensor(TensorType::ACL_INT);
 
    const unsigned int width_idx             = 1;
    const unsigned int height_idx            = 2;
    const unsigned int batch_idx             = 3;
    const int          element_size_in_bytes = dst_nhwc->info()->element_size();
    const auto         dst_strides           = dst_nhwc->info()->strides_in_bytes();
 
    const size_t out_row_stride   = dst_strides[height_idx] / element_size_in_bytes;
    const size_t out_col_stride   = dst_strides[width_idx] / element_size_in_bytes;
    const size_t out_batch_stride = dst_strides[batch_idx] / element_size_in_bytes;
    const auto   wout_transf_ptr  = reinterpret_cast<const void *>(
        winograd_output_transform->buffer() + winograd_output_transform->info()->offset_first_element_in_bytes());
    auto dst_nhwc_ptr =
        reinterpret_cast<void *>(dst_nhwc->buffer() + dst_nhwc->info()->offset_first_element_in_bytes());
    void *biases_data_ptr = nullptr;
    if (biases != nullptr)
    {
        biases_data_ptr = reinterpret_cast<void *>(biases->buffer() + biases->info()->offset_first_element_in_bytes());
    }
 
    // Output transform
    _winograd_impl.output_transform->execute(_conv_args, wout_transf_ptr, _winograd_impl.winograd_spec, biases_data_ptr,
                                             dst_nhwc_ptr, out_batch_stride, out_row_stride, out_col_stride,
                                             workspace->buffer(), info.thread_id, _nthreads);
}

References arm_compute::ACL_DST, arm_compute::ACL_INT, arm_compute::ACL_SRC_0, arm_compute::ACL_SRC_1, ARM_COMPUTE_UNUSED, ITensor::buffer(), ITensorInfo::element_size(), ITensorPack::get_const_tensor(), ITensorPack::get_tensor(), arm_compute::cpu::height_idx, ITensor::info(), arm_compute::test::validation::info, ITensorInfo::offset_first_element_in_bytes(), ITensorInfo::strides_in_bytes(), arm_compute::cpu::width_idx, IKernel::window(), and arm_compute::test::validation::reference::winograd_output_transform().

---

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

• src/cpu/kernels/CpuWinogradConv2dKernel.h
• src/cpu/kernels/CpuWinogradConv2dKernel.cpp