/**
 * 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.
 */


/**
 * @file xlog1py.cpp
 * @brief ACLNN L0 API 实现 - xlog1py 算子
 *
 * z = x * log1p(y), 特殊规则: x == 0 → z = 0
 * 2入1出 broadcast，支持 FLOAT16, FLOAT, BF16
 */

#include "xlog1py.h"
#include "opdev/op_log.h"
#include "opdev/op_dfx.h"
#include "opdev/shape_utils.h"
#include "opdev/make_op_executor.h"
#include "opdev/platform.h"

using namespace op;

namespace l0op {

OP_TYPE_REGISTER(Xlog1py);

static const std::initializer_list<op::DataType> AICORE_DTYPE_SUPPORT_LIST = {
    DataType::DT_FLOAT, DataType::DT_FLOAT16, DataType::DT_BF16
};

static bool IsAiCoreSupport(const aclTensor* x, const aclTensor* y)
{
    auto npuArch = GetCurrentPlatformInfo().GetCurNpuArch();
    OP_CHECK(npuArch == NpuArch::DAV_3510,
             OP_LOGE(ACLNN_ERR_PARAM_INVALID,
                     "Xlog1py only supported on ascend950 (DAV_3510). npuArch=%d.",
                     static_cast<int>(npuArch)),
             return false);
    OP_CHECK(CheckType(x->GetDataType(), AICORE_DTYPE_SUPPORT_LIST) &&
             CheckType(y->GetDataType(), AICORE_DTYPE_SUPPORT_LIST),
             OP_LOGE(ACLNN_ERR_PARAM_INVALID,
                     "Xlog1py dtype not supported: x_dtype=%d, y_dtype=%d.",
                     static_cast<int>(x->GetDataType()),
                     static_cast<int>(y->GetDataType())),
             return false);
    return true;
}

static bool Xlog1pyInferShape(const op::Shape& xShape, const op::Shape& yShape, op::Shape& outShape)
{
    OP_CHECK(BroadcastInferShape(xShape, yShape, outShape),
             OP_LOGE(ACLNN_ERR_PARAM_INVALID, "Xlog1py shape broadcast failed."),
             return false);
    return true;
}

static const aclTensor* Xlog1pyAiCore(const aclTensor* x, const aclTensor* y,
                                       const aclTensor* out, aclOpExecutor* executor)
{
    L0_DFX(Xlog1pyAiCore, x, y, out);

    auto ret = ADD_TO_LAUNCHER_LIST_AICORE(Xlog1py,
        OP_INPUT(x, y), OP_OUTPUT(out));
    OP_CHECK(ret == ACLNN_SUCCESS,
             OP_LOGE(ACLNN_ERR_INNER_NULLPTR, "Xlog1pyAiCore failed."),
             return nullptr);
    return out;
}

const aclTensor* Xlog1py(const aclTensor* x, const aclTensor* y, aclOpExecutor* executor)
{
    Shape outShape;
    const aclTensor* out = nullptr;

    OP_CHECK(Xlog1pyInferShape(x->GetViewShape(), y->GetViewShape(), outShape),
             OP_LOGE(ACLNN_ERR_PARAM_INVALID, "Xlog1py infer shape failed."),
             return nullptr);

    out = executor->AllocTensor(outShape, x->GetDataType());

    OP_CHECK(IsAiCoreSupport(x, y),
             OP_LOGE(ACLNN_ERR_PARAM_INVALID, "Xlog1py IsAiCoreSupport check failed."),
             return nullptr);

    return Xlog1pyAiCore(x, y, out, executor);
}

} // namespace l0op