/**
 * 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 cosh.h
 * \brief Defines a series of interface used to do elementwise math cosh calculation.
 * Formula:  Cosh(x) = (e^x+e^(-x))/2 = e^(x-ln2) + 0.25/(e^(x-ln2))
 */
#ifndef LIB_MATH_COSH_H
#define LIB_MATH_COSH_H
#if __CCE_AICORE__ == 200 || __CCE_AICORE__ == 220

#include "kernel_tensor.h"
#include "../../impl/math/cosh/cosh_common_impl.h"

namespace AscendC {
#pragma begin_pipe(V)
/*!
 * \ingroup Cosh
 * \brief compute Cosh elementwisely
 * \tparam T: half/float
 * \tparam isReuseSource: whether allows API to modify source data, usually for performance reason,
 * this parameter is reserved, please use the default value.
 * \param [out] dstTensor: output LocalTensor
 * \param [in] srcTensor: input LocalTensor
 * \param [in] sharedTmpBuffer: extra temporary shared space used for intermediate values among calculation process,
 * whose required space size should refer to corresponding tiling API, which is defined at cosh_tiling.h.
 * Generally, the more space you allocate, the better performance you will achieve, and the performance
 * reaches peak when buffer size is maximum(calculated by tiling function). Moreover, it is not guaranteed
 * that the shared space will be cleared after usage, the data could be anything.
 * \param [in] calCount: the number of elements to be processed.
 * \note src/dst Tensor must be 32B align, and it doesn't allow src/dst/sharedTmpBuffer tensor address overlap.
 */
template<typename T, bool isReuseSource = false>
__aicore__ inline void Cosh(const LocalTensor<T>& dstTensor, const LocalTensor<T>& srcTensor,
    const LocalTensor<uint8_t>& sharedTmpBuffer)
{
    Cosh<T, isReuseSource>(dstTensor, srcTensor, sharedTmpBuffer, srcTensor.GetSize());
}

/*!
 * \ingroup Cosh
 * \brief compute Cosh elementwisely
 * \tparam T: half/float
 * \tparam isReuseSource: whether allows API to modify source data, usually for performance reason,
 * this parameter is reserved, please use the default value.
 * \param [out] dstTensor: output LocalTensor
 * \param [in] srcTensor: input LocalTensor
 * \param [in] sharedTmpBuffer: extra temporary shared space used for intermediate values among calculation process,
 * whose required space size should refer to corresponding tiling API, which is defined at cosh_tiling.h.
 * Generally, the more space you allocate, the better performance you will achieve, and the performance
 * reaches peak when buffer size is maximum(calculated by tiling function). Moreover, it is not guaranteed
 * that the shared space will be cleared after usage, the data could be anything.
 * \param [in] calCount: the number of elements to be processed.
 * \note src/dst Tensor must be 32B align, and it doesn't allow src/dst/sharedTmpBuffer tensor address overlap.
 */
template <typename T, bool isReuseSource = false>
__aicore__ inline void Cosh(const LocalTensor<T>& dstTensor, const LocalTensor<T>& srcTensor,
    const LocalTensor<uint8_t>& sharedTmpBuffer, const uint32_t calCount)
{
    CoshImpl(dstTensor, srcTensor, sharedTmpBuffer, calCount);
}

/*!
 * \ingroup Cosh
 * \brief compute Cosh elementwisely
 * \tparam T: half/float
 * \tparam isReuseSource: whether allows API to modify source data, usually for performance reason,
 * this parameter is reserved, please use the default value.
 * \param [out] dstTensor: output LocalTensor
 * \param [in] srcTensor: input LocalTensor
 * \note src/dst Tensor must be 32B align, and it doesn't allow src/dst/sharedTmpBuffer tensor address overlap.
 */
template <typename T, bool isReuseSource = false>
__aicore__ inline void Cosh(const LocalTensor<T>& dstTensor, const LocalTensor<T>& srcTensor)
{
    Cosh<T, isReuseSource>(dstTensor, srcTensor, srcTensor.GetSize());
}

/*!
 * \ingroup Cosh
 * \brief compute Cosh elementwisely
 * \tparam T: half/float
 * \tparam isReuseSource: whether allows API to modify source data, usually for performance reason,
 * this parameter is reserved, please use the default value.
 * \param [out] dstTensor: output LocalTensor
 * \param [in] srcTensor: input LocalTensor
 * \param [in] calCount: the number of elements to be processed.
 * \note src/dst Tensor must be 32B align, and it doesn't allow src/dst/sharedTmpBuffer tensor address overlap.
 */
template <typename T, bool isReuseSource = false>
__aicore__ inline void Cosh(const LocalTensor<T>& dstTensor, const LocalTensor<T>& srcTensor, const uint32_t calCount)
{
    CoshImpl(dstTensor, srcTensor, calCount);
}
#pragma end_pipe
} // namespace AscendC
#endif

#endif // LIB_MATH_COSH_H