#include <Half.hpp>
#include <aclCommon/ArmComputeTensorUtils.hpp>
#include <cl/OpenClTimer.hpp>
#include <backendsCommon/CpuTensorHandle.hpp>
#include <armnn/Utils.hpp>
#include <arm_compute/runtime/CL/CLTensor.h>
#include <arm_compute/runtime/IFunction.h>
#include <sstream>
Go to the source code of this file.
|
template<typename T > |
void | CopyArmComputeClTensorData (arm_compute::CLTensor &dstTensor, const T *srcData) |
|
auto | SetClStridedSliceData (const std::vector< int > &m_begin, const std::vector< int > &m_end, const std::vector< int > &m_stride) |
|
auto | SetClSliceData (const std::vector< unsigned int > &m_begin, const std::vector< unsigned int > &m_size) |
|
void | InitializeArmComputeClTensorData (arm_compute::CLTensor &clTensor, const ConstCpuTensorHandle *handle) |
|
RuntimeException | WrapClError (const cl::Error &clError, const CheckLocation &location) |
|
void | RunClFunction (arm_compute::IFunction &function, const CheckLocation &location) |
|
◆ ARMNN_SCOPED_PROFILING_EVENT_CL
#define ARMNN_SCOPED_PROFILING_EVENT_CL |
( |
|
name | ) |
|
Value: name, \
OpenClTimer instrument that times all OpenCl kernels executed between calls to Start() and Stop()...
GPU Execution: OpenCL: ArmCompute.
#define ARMNN_SCOPED_PROFILING_EVENT_WITH_INSTRUMENTS(backendId,...)
Definition at line 20 of file ClWorkloadUtils.hpp.
Referenced by armnn::CopyArmComputeClTensorData(), ClConstantWorkload::Execute(), ClPadWorkload::Execute(), ClConvertFp16ToFp32Workload::Execute(), ClAdditionWorkload::Execute(), ClConvertFp32ToFp16Workload::Execute(), ClSubtractionWorkload::Execute(), ClQuantizeWorkload::Execute(), ClRsqrtWorkload::Execute(), ClActivationWorkload::Execute(), ClLstmFloatWorkload::Execute(), ClAbsWorkload::Execute(), ClDequantizeWorkload::Execute(), ClPreluWorkload::Execute(), ClFloorFloatWorkload::Execute(), ClReshapeWorkload::Execute(), ClResizeWorkload::Execute(), ClSoftmaxFloatWorkload::Execute(), ClInstanceNormalizationWorkload::Execute(), ClSpaceToDepthWorkload::Execute(), ClStackWorkload::Execute(), ClMaximumWorkload::Execute(), ClMinimumWorkload::Execute(), ClBatchToSpaceNdWorkload::Execute(), ClNormalizationFloatWorkload::Execute(), ClArgMinMaxWorkload::Execute(), ClGreaterWorkload< T >::Execute(), ClSliceWorkload::Execute(), ClL2NormalizationFloatWorkload::Execute(), ClSoftmaxUint8Workload::Execute(), ClMeanWorkload::Execute(), ClDepthToSpaceWorkload::Execute(), ClMultiplicationWorkload::Execute(), ClConcatWorkload::Execute(), ClSpaceToBatchNdWorkload::Execute(), ClQuantizedLstmWorkload::Execute(), ClStridedSliceWorkload::Execute(), ClDivisionFloatWorkload::Execute(), ClSplitterWorkload::Execute(), ClPooling2dWorkload::Execute(), ClBatchNormalizationFloatWorkload::Execute(), ClDepthwiseConvolutionWorkload::Execute(), ClConvolution2dWorkload::Execute(), ClFullyConnectedWorkload::Execute(), ClPermuteWorkload::Execute(), and ClTransposeConvolution2dWorkload::Execute().