21.02
|
#include <backendsCommon/Workload.hpp>
#include <aclCommon/ArmComputeTensorUtils.hpp>
#include <neon/NeonTensorHandle.hpp>
#include <neon/NeonTimer.hpp>
#include <backendsCommon/CpuTensorHandle.hpp>
#include <armnn/Utils.hpp>
#include <Half.hpp>
Go to the source code of this file.
Namespaces | |
armnn | |
Copyright (c) 2021 ARM Limited and Contributors. | |
Macros | |
#define | ARMNN_SCOPED_PROFILING_EVENT_NEON(name) |
Functions | |
template<typename T > | |
void | CopyArmComputeTensorData (arm_compute::Tensor &dstTensor, const T *srcData) |
void | InitializeArmComputeTensorData (arm_compute::Tensor &tensor, const ConstCpuTensorHandle *handle) |
auto | SetNeonStridedSliceData (const std::vector< int > &m_begin, const std::vector< int > &m_end, const std::vector< int > &m_stride) |
auto | SetNeonSliceData (const std::vector< unsigned int > &m_begin, const std::vector< unsigned int > &m_size) |
template<typename DataType , typename PayloadType > | |
DataType * | GetOutputTensorData (unsigned int idx, const PayloadType &data) |
#define ARMNN_SCOPED_PROFILING_EVENT_NEON | ( | name | ) |
Definition at line 17 of file NeonWorkloadUtils.hpp.
Referenced by NeonConstantWorkload::Execute(), NeonConvertBf16ToFp32Workload::Execute(), NeonConvertFp16ToFp32Workload::Execute(), NeonFillWorkload::Execute(), NeonConvertFp32ToBf16Workload::Execute(), NeonConvertFp32ToFp16Workload::Execute(), NeonPadWorkload::Execute(), NeonQuantizeWorkload::Execute(), NeonExpWorkload::Execute(), NeonNegWorkload::Execute(), NeonAbsWorkload::Execute(), NeonFloorFloatWorkload::Execute(), NeonRsqrtWorkload::Execute(), NeonLogicalNotWorkload::Execute(), NeonPreluWorkload::Execute(), NeonResizeWorkload::Execute(), NeonGatherWorkload::Execute(), NeonStackWorkload::Execute(), NeonActivationWorkload::Execute(), NeonMeanWorkload::Execute(), NeonArgMinMaxWorkload::Execute(), NeonReduceWorkload::Execute(), NeonInstanceNormalizationWorkload::Execute(), NeonLogicalAndWorkload::Execute(), NeonLogicalOrWorkload::Execute(), NeonMaximumWorkload::Execute(), NeonDivisionWorkload::Execute(), NeonSliceWorkload::Execute(), NeonComparisonWorkload::Execute(), NeonDepthToSpaceWorkload::Execute(), NeonDequantizeWorkload::Execute(), NeonSpaceToDepthWorkload::Execute(), NeonAdditionWorkload::Execute(), NeonMultiplicationWorkload::Execute(), NeonSplitterWorkload::Execute(), NeonReshapeWorkload::Execute(), NeonConcatWorkload::Execute(), NeonNormalizationFloatWorkload::Execute(), NeonSubtractionWorkload::Execute(), NeonL2NormalizationFloatWorkload::Execute(), NeonStridedSliceWorkload::Execute(), NeonSpaceToBatchNdWorkload::Execute(), NeonLogSoftmaxWorkload::Execute(), NeonSoftmaxWorkload::Execute(), NeonBatchToSpaceNdWorkload::Execute(), NeonPooling2dWorkload::Execute(), NeonDetectionPostProcessWorkload::Execute(), NeonBatchNormalizationWorkload::Execute(), NeonPermuteWorkload::Execute(), NeonMinimumWorkload::Execute(), NeonFullyConnectedWorkload::Execute(), NeonTransposeWorkload::Execute(), NeonDepthwiseConvolutionWorkload::Execute(), NeonTransposeConvolution2dWorkload::Execute(), and NeonConvolution2dWorkload::Execute().