/* * Copyright (c) 2022 Huawei Device Co., Ltd. * Licensed under the Apache License, Version 2.0 (the "License"); * you may not use this file except in compliance with the License. * You may obtain a copy of the License at * * http://www.apache.org/licenses/LICENSE-2.0 * * Unless required by applicable law or agreed to in writing, software * distributed under the License is distributed on an "AS IS" BASIS, * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. * See the License for the specific language governing permissions and * limitations under the License. */ #include "mul_builder.h" #include "frameworks/native/transform.h" #include "frameworks/native/validation.h" #include "frameworks/native/ops_registry.h" namespace OHOS { namespace NeuralNetworkRuntime { namespace Ops { static const int INPUT_NUM = 2; static const int OUTPUT_NUM = 1; static const int SCALE_LENGTH = 1; static const std::string OP_NAME = "Mul"; MulBuilder::MulBuilder() {} MulBuilder::~MulBuilder() {} OH_NN_ReturnCode MulBuilder::SetActivationType(std::shared_ptr tensor) { tensor->IdentifyOpParameter(); if (tensor->GetElementCount() != SCALE_LENGTH) { LOGE("[Mul] Mul SetActivationType failed. The shape of activation should be scaler."); return OH_NN_INVALID_PARAMETER; } if (tensor->GetDataType() != OH_NN_INT8) { LOGE("[Mul] Mul SetActivationType failed. The activation should be type OH_NN_INT8."); return OH_NN_INVALID_PARAMETER; } void* buffer = tensor->GetBuffer(); if (buffer == nullptr) { LOGE("[Mul] SetActivationType failed, the activationType passed a empty buffer."); return OH_NN_INVALID_PARAMETER; } int8_t* fuseData = static_cast(buffer); if (!OHOS::NeuralNetworkRuntime::Validation::ValidateFuseType(static_cast(*fuseData))) { LOGE("[Mul] Mul SetActivationType failed. Fuse activation type is invalid"); return OH_NN_INVALID_PARAMETER; } auto fuseType = (OH_NN_FuseType)(*fuseData); m_activationType = NNToMS::TransfromFusionType(fuseType); return OH_NN_SUCCESS; } OH_NN_ReturnCode MulBuilder::Build(const std::vector& paramsIndex, const std::vector& inputsIndex, const std::vector& outputsIndex, const std::vector>& allTensors) { if (m_isBuild) { LOGE("[Mul] Mul build failed. operation has been build, cannot build again."); return OH_NN_OPERATION_FORBIDDEN; } OH_NN_ReturnCode returnCode = CheckIOIndex(inputsIndex, outputsIndex, allTensors, INPUT_NUM, OUTPUT_NUM); if (returnCode != OH_NN_SUCCESS) { LOGE("[Mul] Mul build failed. Passed invalid input or output index of Mul operation index."); return returnCode; } m_inputsIndex = inputsIndex; m_outputsIndex = outputsIndex; for (int i : paramsIndex) { std::shared_ptr tensor = allTensors[i]; switch (tensor->GetType()) { case OH_NN_MUL_ACTIVATION_TYPE: returnCode = SetActivationType(tensor); break; default: LOGE("[Mul] Parameter Type is invalid, type=%d", tensor->GetType()); return OH_NN_INVALID_PARAMETER; } if (returnCode != OH_NN_SUCCESS) { LOGE("[Mul] Mul build failed. Passed invalid param."); return returnCode; } } // The quantization type of the first output determinies that of the operator. SetQuantType(outputsIndex, allTensors); m_name = OP_NAME; m_isBuild = true; return OH_NN_SUCCESS; } LiteGraphPrimitvePtr MulBuilder::GetPrimitive() { if (!m_isBuild) { LOGE("[Mul] Mul GetPrimitive failed. Cannot get primitive before call build."); return {nullptr, DestroyLiteGraphPrimitive}; } void* primitive = mindspore::lite::MindIR_MulFusion_CreatePrimitive(m_activationType); LiteGraphPrimitvePtr graphPrimitivePtr(primitive, DestroyLiteGraphPrimitive); return graphPrimitivePtr; } REGISTER_OPS(MulBuilder, OH_NN_OPS_MUL); } // namespace Ops } // namespace NeuralNetworkRuntime } // namespace OHOS