/**
 * Copyright (c) 2024 Huawei Technologies Co., Ltd.
 * This file is a part of the CANN Open Software.
 * Licensed under CANN Open Software License Agreement Version 1.0 (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 xor.h
 * \brief
 */
#ifndef LIB_XOR_XOR_H
#define LIB_XOR_XOR_H

#include "kernel_tensor.h"
#include "kernel_operator_intf.h"
#include "kernel_pop_stack_buffer.h"
#include "../../impl/math/xor/xor_common_impl.h"
#if ASCENDC_CPU_DEBUG
#include "kernel_log.h"
#include <type_traits>
#endif

#if __CCE_AICORE__ >= 200

namespace AscendC {
#pragma begin_pipe(V)
/*
 * @brief Xor Computes the element-wise logical XOR of the given input tensors. Zeros are treated as False and nonzeros
 * are treated as True. Mathematical formulas: 0^0=0；0^1=1；1^0=1；1^1=0
 * @ingroup xor
 * @param [out] dstTensor, output LocalTensor
 * @param [in] srcTensor0, input LocalTensor
 * @param [in] srcTensor1, input LocalTensor
 * @param [in] sharedTmpBuffer, input local temporary Tensor
 * @param [in] calCount, amount of input data to be calculated
 */
template <typename T, bool isReuseSource = false>
__aicore__ inline void Xor(const LocalTensor<T>& dstTensor, const LocalTensor<T> &src0Tensor,
    const LocalTensor<T> &src1Tensor, const LocalTensor<uint8_t>& sharedTmpBuffer, const uint32_t calCount)
{
    // Only for AI Vector Core.
    if ASCEND_IS_AIC {
        return;
    }

    uint32_t tmpBufferSize = sharedTmpBuffer.GetSize() / sizeof(T);
    uint32_t stackSize = tmpBufferSize / ONE_BLK_SIZE * ONE_BLK_SIZE; // integer multiple of 32 bytes
#if ASCENDC_CPU_DEBUG
    ASCENDC_ASSERT(((QuePosition)dstTensor.GetPosition() == TPosition::VECIN ||
                    (QuePosition)dstTensor.GetPosition() == TPosition::VECOUT ||
                    (QuePosition)dstTensor.GetPosition() == TPosition::VECCALC), {
        KERNEL_LOG(KERNEL_ERROR,
            "dst position not support, just support position is VECIN, VECOUT, VECCALC.");
    });
    bool result = (calCount <= src0Tensor.GetSize()) &&  (calCount <= src1Tensor.GetSize())
        && (calCount <= dstTensor.GetSize() && (calCount > 0));
    ASCENDC_ASSERT(result, { KERNEL_LOG(KERNEL_ERROR, "calCount must be less than or equal to src/dst tensor"); });
    result = (std::is_same<T, int16_t>::value) || (std::is_same<T, uint16_t>::value);
    ASCENDC_ASSERT(result, { KERNEL_LOG(KERNEL_ERROR, "type must be int16_t or unt16_t"); });
    ASCENDC_ASSERT((tmpBufferSize > 0), { KERNEL_LOG(KERNEL_ERROR, "tmpBufferSize must > 0!"); });
    ASCENDC_ASSERT((stackSize > 0), { KERNEL_LOG(KERNEL_ERROR, "stackSize must > 0!"); });
#endif
    const uint32_t round = calCount / stackSize;
    const uint32_t tail = calCount % stackSize;
    LocalTensor<T> tmpTensor = sharedTmpBuffer.ReinterpretCast<T>();
    SetMaskCount();
    SetVectorMask<T>(0, stackSize);
    uint32_t offset = 0;
    for (uint32_t i = 0; i < round; i++) {
        XorCalcSimplified(dstTensor[offset], src0Tensor[offset], src1Tensor[offset], tmpTensor);
        offset = offset + stackSize;
    }

    if (tail != 0) {
        SetVectorMask<T>(0, tail);
        XorCalcSimplified(dstTensor[offset], src0Tensor[offset], src1Tensor[offset], tmpTensor);
    }

    SetMaskNorm();
    SetVectorMask<T>(FULL_MASK, FULL_MASK);
}

/*
 * @brief Xor Computes the element-wise logical XOR of the given input tensors. Zeros are treated as False and nonzeros
 * are treated as True. Mathematical formulas: 0^0=0；0^1=1；1^0=1；1^1=0
 * @ingroup xor
 * @param [out] dstTensor, output LocalTensor
 * @param [in] srcTensor0, input LocalTensor
 * @param [in] srcTensor1, input LocalTensor
 * @param [in] sharedTmpBuffer, input local temporary Tensor
 */
template <typename T, bool isReuseSource = false>
__aicore__ inline void Xor(const LocalTensor<T>& dstTensor, const LocalTensor<T> &src0Tensor,
    const LocalTensor<T> &src1Tensor, const LocalTensor<uint8_t>& sharedTmpBuffer)
{
#if ASCENDC_CPU_DEBUG
    bool result = (src0Tensor.GetSize() == src1Tensor.GetSize());
    ASCENDC_ASSERT(result, { KERNEL_LOG(KERNEL_ERROR, "operands must be equal in size"); });
#endif
    Xor<T, isReuseSource>(dstTensor, src0Tensor, src1Tensor, sharedTmpBuffer, src0Tensor.GetSize());
}

/*
 * @brief Xor Computes the element-wise logical XOR of the given input tensors. Zeros are treated as False and nonzeros
 * are treated as True. Mathematical formulas: 0^0=0；0^1=1；1^0=1；1^1=0
 * @ingroup xor
 * @param [out] dstTensor, output LocalTensor
 * @param [in] srcTensor0, input LocalTensor
 * @param [in] srcTensor1, input LocalTensor
 * @param [in] calCount, amount of input data to be calculated
 */
template <typename T, bool isReuseSource = false>
__aicore__ inline void Xor(const LocalTensor<T> &dstTensor, const LocalTensor<T> &src0Tensor,
    const LocalTensor<T> &src1Tensor, const uint32_t calCount)
{
    LocalTensor<uint8_t> sharedTmpBuffer;
    bool ans = PopStackBuffer<uint8_t, TPosition::LCM>(sharedTmpBuffer);
    ASCENDC_ASSERT((ans), { KERNEL_LOG(KERNEL_ERROR, "PopStackBuffer Error!"); });

    Xor<T, isReuseSource>(dstTensor, src0Tensor, src1Tensor, sharedTmpBuffer, calCount);
}

/*
 * @brief Xor Computes the element-wise logical XOR of the given input tensors. Zeros are treated as False and nonzeros
 * are treated as True. Mathematical formulas: 0^0=0；0^1=1；1^0=1；1^1=0
 * @ingroup xor
 * @param [out] dstTensor, output LocalTensor
 * @param [in] srcTensor0, input LocalTensor
 * @param [in] srcTensor1, input LocalTensor
 */
template <typename T, bool isReuseSource = false>
__aicore__ inline void Xor(const LocalTensor<T> &dstTensor, const LocalTensor<T> &src0Tensor,
    const LocalTensor<T> &src1Tensor)
{
#if ASCENDC_CPU_DEBUG
    bool result = (src0Tensor.GetSize() == src1Tensor.GetSize());
    ASCENDC_ASSERT(result, { KERNEL_LOG(KERNEL_ERROR, "operands must be equal in size"); });
#endif
    Xor<T, isReuseSource>(dstTensor, src0Tensor, src1Tensor, src0Tensor.GetSize());
}
#pragma end_pipe
}  // namespace AscendC

#endif

#endif  // LIB_XOR_XOR_H