* Copyright (c) 2025-2026 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 "onnx_common.h"
#include "math/reduce_max/op_graph/reduce_max_proto.h"
using namespace ge;
using ge::Operator;
namespace domi {
using NodeProto = ge::onnx::NodeProto;
static Status ParseParamsReduceMax(const Message* op_src, ge::Operator& op_dest)
{
const NodeProto* node = dynamic_cast<const NodeProto*>(op_src);
if (node == nullptr) {
OP_LOGE(GetOpName(op_dest).c_str(), "Dynamic cast op_src to NodeProto failed.");
return FAILED;
}
std::vector<int> axes = {};
bool keep_dims = true;
for (const auto& attr : node->attribute()) {
if (attr.name() == "axes" && attr.type() == ge::onnx::AttributeProto::INTS) {
for (int i = 0; i < attr.ints_size(); i++) {
axes.push_back(attr.ints(i));
}
} else if (attr.name() == "keepdims" && attr.type() == ge::onnx::AttributeProto::INT) {
keep_dims = (attr.i() == 1);
}
if (attr.name() == "noop_with_empty_axes" && attr.type() == ge::onnx::AttributeProto::INT && attr.i() == 1) {
OP_LOGW(GetOpName(op_dest).c_str(), "Only support noop_with_empty_axes=0, but 1 is obtained now");
}
}
int num = axes.size();
std::vector<int64_t> dims = {};
if (num != 0) {
dims.push_back(num);
} else {
dims.push_back(0);
}
ge::Tensor tensor = Vec2Tensor(axes, dims, ge::DT_INT32, ge::FORMAT_NCHW);
op_dest.SetAttr("name", node->name());
op_dest.SetAttr("axes", tensor);
op_dest.SetAttr("keep_dims", keep_dims);
op_dest.SetAttr("noop_with_empty_axes", 0);
const int input = 2;
const int output = 1;
op_dest.DynamicInputRegister("x", input);
op_dest.DynamicOutputRegister("y", output);
op_dest.SetAttr("original_type", "ai.onnx::11::ReduceMax");
return SUCCESS;
}
static Status ParseOpToGraphReduceMax(const ge::Operator& op, Graph& graph)
{
std::string ori_name;
if (op.GetAttr("name", ori_name) != SUCCESS) {
OP_LOGE(GetOpName(op).c_str(), "get name from op failed.");
return FAILED;
}
auto data0 = op::Data((ori_name + "_data0").c_str()).set_attr_index(0);
ge::Tensor axes;
if (op.GetAttr("axes", axes) != SUCCESS) {
OP_LOGE(GetOpName(op).c_str(), "get axes from op failed");
return FAILED;
}
auto data1 = op::Const((ori_name + "_data1").c_str()).set_attr_value(axes);
auto reducemax = op::ReduceMax((ori_name + "_ReduceMax").c_str()).set_input_x(data0).set_input_axes(data1);
bool keep_dims = false;
if (op.GetAttr("keep_dims", keep_dims) != SUCCESS) {
OP_LOGE(GetOpName(op).c_str(), "get keep_dims from op failed");
return FAILED;
}
reducemax.set_attr_keep_dims(keep_dims);
int noop_with_empty_axes = 0;
if (op.GetAttr("noop_with_empty_axes", noop_with_empty_axes) != SUCCESS) {
OP_LOGE(GetOpName(op).c_str(), "get noop_with_empty_axes from op failed");
return FAILED;
}
reducemax.set_attr_noop_with_empty_axes(noop_with_empty_axes);
std::vector<ge::Operator> inputs{data0};
std::vector<std::pair<ge::Operator, std::vector<size_t> > > outputs;
outputs.emplace_back(reducemax, std::vector<std::size_t>{0});
graph.SetInputs(inputs).SetOutputs(outputs);
return SUCCESS;
}
static Status ParseParamsReduceMax13(const Message* op_src, ge::Operator& op_dest)
{
const NodeProto* node = dynamic_cast<const NodeProto*>(op_src);
if (node == nullptr) {
OP_LOGE("ReduceMax13", "Dynamic cast op_src to NodeProto failed.");
return FAILED;
}
op_dest.SetAttr("original_type", "ai.onnx::13::ReduceMax");
int input_size = node->input_size();
bool keep_dims = true;
int noop_with_empty_axes = 0;
for (const auto& attr : node->attribute()) {
if (attr.name() == "keepdims" && attr.type() == ge::onnx::AttributeProto::INT) {
keep_dims = (attr.i() == 1);
} else if (attr.name() == "noop_with_empty_axes" && attr.type() == ge::onnx::AttributeProto::INT) {
noop_with_empty_axes = attr.i();
}
}
std::vector<int> axes = {};
std::vector<int64_t> dims = {0};
ge::Tensor axes_tensor = Vec2Tensor(axes, dims, ge::DT_INT32, ge::FORMAT_NCHW);
op_dest.SetAttr("axes", axes_tensor);
op_dest.SetAttr("name", node->name());
op_dest.SetAttr("input_size", input_size);
op_dest.SetAttr("keep_dims", keep_dims);
op_dest.SetAttr("noop_with_empty_axes", noop_with_empty_axes);
return SUCCESS;
}
namespace {
struct ReduceMax13Prop {
std::string ori_name;
bool keep_dims = false;
int input_num = 1;
int empty_axes = 0;
};
Status GetProperty(const Operator& op, ReduceMax13Prop& prop)
{
if (op.GetAttr("name", prop.ori_name) != SUCCESS) {
OP_LOGE(GetOpName(op).c_str(), "get name from op failed.");
return FAILED;
}
if (op.GetAttr("keep_dims", prop.keep_dims) != SUCCESS) {
OP_LOGE(GetOpName(op).c_str(), "get keep_dims from op failed");
return FAILED;
}
if (op.GetAttr("input_size", prop.input_num) != SUCCESS) {
OP_LOGE(GetOpName(op).c_str(), "get input_num from op failed");
return FAILED;
}
if (op.GetAttr("noop_with_empty_axes", prop.empty_axes) != SUCCESS) {
OP_LOGE(GetOpName(op).c_str(), "get attribute noop_with_empty_axes failed");
return FAILED;
}
return SUCCESS;
}
}
static Status ParseOpToGraphReduceMax13(const Operator& op, Graph& graph)
{
ReduceMax13Prop prop;
if (GetProperty(op, prop) != SUCCESS) {
return FAILED;
}
auto data0 = op::Data((prop.ori_name + "_data0").c_str()).set_attr_index(0);
int num_input = 2;
if (prop.input_num == 1) {
ge::Tensor axes;
if (op.GetAttr("axes", axes) != SUCCESS) {
OP_LOGE(GetOpName(op).c_str(), "get axes from op failed");
return FAILED;
}
auto data1 = op::Const((prop.ori_name + "_data1").c_str()).set_attr_value(axes);
auto reducemax = op::ReduceMax((prop.ori_name + "_ReduceMax").c_str())
.set_input_x(data0)
.set_input_axes(data1)
.set_attr_keep_dims(prop.keep_dims)
.set_attr_noop_with_empty_axes(prop.empty_axes);
std::vector<Operator> inputs{data0};
std::vector<std::pair<Operator, std::vector<size_t> > > output_indexs;
output_indexs.emplace_back(reducemax, vector<std::size_t>{0});
graph.SetInputs(inputs).SetOutputs(output_indexs);
} else if (prop.input_num == num_input) {
auto data1 = op::Data((prop.ori_name + "_data1").c_str()).set_attr_index(1);
auto reducemax13 = op::ReduceMax((prop.ori_name + "_ReduceMax").c_str())
.set_input_x(data0)
.set_input_axes(data1)
.set_attr_keep_dims(prop.keep_dims)
.set_attr_noop_with_empty_axes(prop.empty_axes);
std::vector<Operator> inputs{data0, data1};
std::vector<std::pair<Operator, std::vector<size_t> > > output_indexs;
output_indexs.emplace_back(reducemax13, vector<std::size_t>{0});
graph.SetInputs(inputs).SetOutputs(output_indexs);
} else {
OP_LOGE(GetOpName(op).c_str(), "Input num or set attr is error");
return FAILED;
}
return SUCCESS;
}
REGISTER_CUSTOM_OP("PartitionedCall")
.FrameworkType(ONNX)
.OriginOpType({ge::AscendString("ai.onnx::8::ReduceMax"),
ge::AscendString("ai.onnx::9::ReduceMax"),
ge::AscendString("ai.onnx::10::ReduceMax"),
ge::AscendString("ai.onnx::11::ReduceMax"),
ge::AscendString("ai.onnx::12::ReduceMax")})
.ParseParamsFn(ParseParamsReduceMax)
.ParseOpToGraphFn(ParseOpToGraphReduceMax)
.ImplyType(ImplyType::TVM);
REGISTER_CUSTOM_OP("ReduceMax")
.FrameworkType(ONNX)
.OriginOpType({ge::AscendString("ai.onnx::13::ReduceMax"),
ge::AscendString("ai.onnx::14::ReduceMax"),
ge::AscendString("ai.onnx::15::ReduceMax"),
ge::AscendString("ai.onnx::16::ReduceMax"),
ge::AscendString("ai.onnx::17::ReduceMax"),
ge::AscendString("ai.onnx::18::ReduceMax")})
.ParseParamsFn(ParseParamsReduceMax13)
.ParseOpToGraphFn(ParseOpToGraphReduceMax13)
.ImplyType(ImplyType::TVM);
}
