/**
 * Copyright (c) 2025 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 "aclnn_real.h"
#include "real.h"
#include "aclnn_kernels/contiguous.h"
#include "opdev/op_log.h"
#include "opdev/op_dfx.h"
#include "opdev/common_types.h"
#include "opdev/data_type_utils.h"
#include "opdev/make_op_executor.h"
#include "opdev/platform.h"
#include "aclnn_kernels/common/op_error_check.h"
#include "op_api/aclnn_check.h"

using namespace op;

#ifdef __cplusplus
extern "C" {
#endif

// Multi-SOC dtype support matrix（在本算子各 SOC 间一致性约束）：
// - ascend950 (regbase / arch35): AICORE 走本仓 op_kernel/arch35/real_apt（FLOAT/FLOAT16/COMPLEX32/COMPLEX64）；
//   COMPLEX128 通过 RealAiCpu 走AICPU 实现（DOUBLE 输出）。
// - ascend910B / ascend910_93: AICORE 同上；COMPLEX128 通过 RealAiCpu 走 canndev AICPU 实现（DOUBLE 输出）。
// - 其它 (legacy 910): 全部走 RealAiCpu（COMPLEX64/COMPLEX128）。
static const std::initializer_list<DataType> ASCEND910B_DTYPE_SUPPORT_LIST = {
    DataType::DT_FLOAT, DataType::DT_FLOAT16, DataType::DT_COMPLEX32, DataType::DT_COMPLEX64, DataType::DT_COMPLEX128};

// Generated By CANNBot: ascend950 / arch35 (regbase). COMPLEX128 已确认在本 SOC 支持，走 AI CPU：
// l0 IsAiCoreSupport 的 AICORE 表不含 COMPLEX128，经 RealAiCpu 由 ADD_TO_LAUNCHER_LIST_AICPU 走 canndev
// AICPU 实现（DOUBLE 输出），与 op_def AICore 配置（不含 COMPLEX128）一致。
static const std::initializer_list<DataType> ASCEND950_DTYPE_SUPPORT_LIST = {
    DataType::DT_FLOAT, DataType::DT_FLOAT16, DataType::DT_COMPLEX32, DataType::DT_COMPLEX64, DataType::DT_COMPLEX128};

static const std::initializer_list<DataType> OTHER_DTYPE_SUPPORT_LIST = {DataType::DT_COMPLEX64,
                                                                         DataType::DT_COMPLEX128};

static const std::initializer_list<std::pair<DataType, DataType>> DTYPE_SUPPORT_LIST = {
    {DataType::DT_COMPLEX64, DataType::DT_FLOAT},
    {DataType::DT_COMPLEX32, DataType::DT_FLOAT16},
    {DataType::DT_COMPLEX128, DataType::DT_DOUBLE},
    {DataType::DT_FLOAT16, DataType::DT_FLOAT16},
    {DataType::DT_FLOAT, DataType::DT_FLOAT}};

static bool IsPrivateFormat(ge::Format format)
{
    return format == ge::FORMAT_NC1HWC0 || format == ge::FORMAT_FRACTAL_Z || format == ge::FORMAT_NDC1HWC0 ||
           format == ge::FORMAT_FRACTAL_Z_3D || format == ge::FORMAT_FRACTAL_NZ || format == ge::FORMAT_NC1HWC0_C04;
}

static const std::initializer_list<DataType>& GetDtypeSupportList()
{
    auto npuArch = GetCurrentPlatformInfo().GetCurNpuArch();
    if (IsRegBase(npuArch)) {
        return ASCEND950_DTYPE_SUPPORT_LIST;
    }
    if (GetCurrentPlatformInfo().GetSocVersion() == SocVersion::ASCEND910B ||
        GetCurrentPlatformInfo().GetSocVersion() == SocVersion::ASCEND910_93) {
        return ASCEND910B_DTYPE_SUPPORT_LIST;
    } else {
        return OTHER_DTYPE_SUPPORT_LIST;
    }
}

static bool SpecialDtypeTensorNeedPassthrough(const aclTensor* self)
{
    if (self->GetDataType() == DataType::DT_FLOAT || self->GetDataType() == DataType::DT_FLOAT16) {
        return true;
    }

    return false;
}

static bool CheckNotNull(const aclTensor* self, const aclTensor* out)
{
    OP_CHECK_NULL(self, return false);
    OP_CHECK_NULL(out, return false);

    return true;
}

static bool CheckDtypeValid(const aclTensor* self)
{
    // 检查self的数据类型是否在支持列表内
    auto supportList = GetDtypeSupportList();
    OP_CHECK_DTYPE_NOT_SUPPORT(self, supportList, return false);

    return true;
}

static bool CheckShape(const aclTensor* self, const aclTensor* out)
{
    // 所有算子的维度都不能超过8
    OP_CHECK_MAX_DIM(self, ACLNN_MAX_SHAPE_RANK, return false);

    // self和out的shape必须一致
    OP_CHECK_SHAPE_NOT_EQUAL(self, out, return false);

    return true;
}

static bool CheckFormat(const aclTensor* self, const aclTensor* out)
{
    const auto curArch = GetCurrentPlatformInfo().GetCurNpuArch();
    if (!IsRegBase(curArch)) {
        return true;
    }
    OP_CHECK(!::IsPrivateFormat(self->GetStorageFormat()),
             OP_LOGE(ACLNN_ERR_PARAM_INVALID, "Real does not support private input format on RegBase."), return false);
    OP_CHECK(!::IsPrivateFormat(out->GetStorageFormat()),
             OP_LOGE(ACLNN_ERR_PARAM_INVALID, "Real does not support private output format on RegBase."), return false);
    return true;
}

static bool CheckInOutDtypeMatch(const aclTensor* self, const aclTensor* out)
{
    auto found = std::find(DTYPE_SUPPORT_LIST.begin(), DTYPE_SUPPORT_LIST.end(),
                           std::pair<DataType, DataType>(self->GetDataType(), out->GetDataType()));
    if (found != DTYPE_SUPPORT_LIST.end()) {
        return true;
    }
    OP_LOGE(ACLNN_ERR_PARAM_INVALID, "when dtype of input is [%s] output is [%s],the param is invalid.",
            ToString(self->GetDataType()).GetString(), ToString(out->GetDataType()).GetString());
    return false;
}

static aclnnStatus CheckParams(const aclTensor* self, const aclTensor* out)
{
    // 1. 检查参数是否为空指针
    CHECK_RET(CheckNotNull(self, out), ACLNN_ERR_PARAM_NULLPTR);

    // 2. 检查输入的数据类型是否在API支持的数据类型范围内，需要根据api定义校验
    CHECK_RET(CheckDtypeValid(self), ACLNN_ERR_PARAM_INVALID);

    // 3. 检查输入输出数据类型是否匹配
    CHECK_RET(CheckInOutDtypeMatch(self, out), ACLNN_ERR_PARAM_INVALID);

    // 4. 检查shape是否满足约束
    CHECK_RET(CheckShape(self, out), ACLNN_ERR_PARAM_INVALID);

    CHECK_RET(CheckFormat(self, out), ACLNN_ERR_PARAM_INVALID);

    return ACLNN_SUCCESS;
}

aclnnStatus aclnnRealGetWorkspaceSize(const aclTensor* self, aclTensor* out, uint64_t* workspaceSize,
                                      aclOpExecutor** executor)
{
    OP_CHECK_COMM_INPUT(workspaceSize, executor);

    L2_DFX_PHASE_1(aclnnReal, DFX_IN(self), DFX_OUT(out));

    // 固定写法，创建OpExecutor
    auto uniqueExecutor = CREATE_EXECUTOR();
    CHECK_RET(uniqueExecutor.get() != nullptr, ACLNN_ERR_INNER_CREATE_EXECUTOR);

    // 固定写法，参数检查
    auto ret = CheckParams(self, out);
    CHECK_RET(ret == ACLNN_SUCCESS, ret);

    // 空Tensor处理
    if (self->IsEmpty()) {
        *workspaceSize = 0;
        uniqueExecutor.ReleaseTo(executor);
        return ACLNN_SUCCESS;
    }

    // self如果非连续，需要转换
    auto selfContiguous = l0op::Contiguous(self, uniqueExecutor.get());
    CHECK_RET(selfContiguous != nullptr, ACLNN_ERR_INNER_NULLPTR);

    // FLOAT和FLOAT16类型，直接拷贝透传
    if (SpecialDtypeTensorNeedPassthrough(selfContiguous)) {
        auto viewCopyResult = l0op::ViewCopy(selfContiguous, out, uniqueExecutor.get());
        *workspaceSize = uniqueExecutor->GetWorkspaceSize();
        uniqueExecutor.ReleaseTo(executor);
        return (viewCopyResult != nullptr) ? ACLNN_SUCCESS : ACLNN_ERR_INNER_NULLPTR;
    }

    // 调用l0算子Real进行计算
    auto realResult = l0op::Real(selfContiguous, uniqueExecutor.get());
    CHECK_RET(realResult != nullptr, ACLNN_ERR_INNER_NULLPTR);

    // 如果出参out是非连续Tensor，需要把计算完的连续Tensor转非连续
    auto viewCopyResult = l0op::ViewCopy(realResult, out, uniqueExecutor.get());
    CHECK_RET(viewCopyResult != nullptr, ACLNN_ERR_INNER_NULLPTR);

    // 固定写法，获取计算过程中需要使用的workspace大小
    *workspaceSize = uniqueExecutor->GetWorkspaceSize();
    uniqueExecutor.ReleaseTo(executor);
    return ACLNN_SUCCESS;
}

aclnnStatus aclnnReal(void* workspace, uint64_t workspaceSize, aclOpExecutor* executor, aclrtStream stream)
{
    L2_DFX_PHASE_2(aclnnReal);

    return CommonOpExecutorRun(workspace, workspaceSize, executor, stream);
}

#ifdef __cplusplus
}
#endif