* Copyright (c) 2026 Huawei Technologies Co., Ltd.
* This program is free software, you can redistribute it and/or modify it under the terms and conditions of
* CANN Open Software License Agreement Version 2.0 (the "License").
* Please refer to the License for details. You may not use this file except in compliance with the License.
* THIS SOFTWARE IS PROVIDED ON AN "AS IS" BASIS, WITHOUT WARRANTIES OF ANY KIND, EITHER EXPRESS OR IMPLIED,
* INCLUDING BUT NOT LIMITED TO NON-INFRINGEMENT, MERCHANTABILITY, OR FITNESS FOR A PARTICULAR PURPOSE.
* See LICENSE in the root of the software repository for the full text of the License.
*/
#include "math/add/op_api/add.h"
#include "opdev/aicpu/aicpu_task.h"
#include "opdev/make_op_executor.h"
#include "opdev/op_def.h"
#include "opdev/op_dfx.h"
#include "opdev/op_executor.h"
#include "opdev/op_log.h"
#include "opdev/shape_utils.h"
using namespace op;
namespace l0op {
OP_TYPE_REGISTER(Addcmul);
static const std::initializer_list<op::DataType> AICORE_DTYPE_SUPPORT_LIST = {
op::DataType::DT_FLOAT, op::DataType::DT_FLOAT16, op::DataType::DT_BF16, op::DataType::DT_INT32};
static bool IsAiCoreSupport(const aclTensor* self)
{
return CheckType(self->GetDataType(), AICORE_DTYPE_SUPPORT_LIST);
}
static const aclTensor* AddcmulAiCore(
const aclTensor* self, const aclTensor* tensor1, const aclTensor* tensor2, const aclTensor* value,
const aclTensor* out, aclOpExecutor* executor)
{
L0_DFX(AddcmulAiCore, self, tensor1, tensor2, out);
auto ret = ADD_TO_LAUNCHER_LIST_AICORE(Addcmul, OP_INPUT(self, tensor1, tensor2, value), OP_OUTPUT(out));
OP_CHECK(
ret == ACL_SUCCESS, OP_LOGE(ACLNN_ERR_INNER_NULLPTR, "AddcmulAiCore ADD_TO_LAUNCHER_LIST_AICORE failed."),
return nullptr);
return out;
}
static const aclTensor* AddcmulAiCpu(
const aclTensor* self, const aclTensor* tensor1, const aclTensor* tensor2, const aclTensor* value, aclTensor* out,
aclOpExecutor* executor)
{
L0_DFX(AddcmulAiCpu, self, tensor1, tensor2, out);
static internal::AicpuTaskSpace space("Addcmul");
auto ret =
ADD_TO_LAUNCHER_LIST_AICPU(Addcmul, OP_ATTR_NAMES(), OP_INPUT(self, tensor1, tensor2, value), OP_OUTPUT(out));
OP_CHECK(
ret == ACL_SUCCESS, OP_LOGE(ACLNN_ERR_INNER_NULLPTR, "AddcmulAiCpu ADD_TO_LAUNCHER_LIST_AICPU failed."),
return nullptr);
return out;
}
static bool InferShape(
const aclTensor* self, const aclTensor* tensor1, const aclTensor* tensor2, op::Shape& broadcastShape)
{
if (!BroadcastInferShape(tensor1->GetViewShape(), tensor2->GetViewShape(), broadcastShape)) {
return false;
}
if (!BroadcastInferShape(self->GetViewShape(), broadcastShape, broadcastShape)) {
return false;
}
return true;
}
const aclTensor* Addcmul(
const aclTensor* self, const aclTensor* tensor1, const aclTensor* tensor2, const aclTensor* value,
aclOpExecutor* executor)
{
op::Shape broadcastShape;
if (!InferShape(self, tensor1, tensor2, broadcastShape)) {
OP_LOGE(ACLNN_ERR_PARAM_INVALID, "Broadcast failed.");
return nullptr;
}
auto addcmulOut = executor->AllocTensor(broadcastShape, self->GetDataType());
if (IsAiCoreSupport(self)) {
return AddcmulAiCore(self, tensor1, tensor2, value, addcmulOut, executor);
} else {
return AddcmulAiCpu(self, tensor1, tensor2, value, addcmulOut, executor);
}
}
}