* 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.
*/
* \file digamma.h
* \brief
*/
#ifndef LIB_MATH_DIGAMMA_H
#define LIB_MATH_DIGAMMA_H
#if defined(__NPU_ARCH__) && (__NPU_ARCH__ == 2201 || __NPU_ARCH__ == 2002 || __NPU_ARCH__ == 3101 || __NPU_ARCH__ == 5102)
#include "kernel_tensor.h"
#include "kernel_pop_stack_buffer.h"
#include "kernel_tiling/kernel_tiling.h"
#include "../../../impl/adv_api/detail/math/digamma/digamma_common_impl.h"
namespace AscendC {
#pragma begin_pipe(V)
* @ingroup Digamma
* @brief Computes the logarithmic derivative of the gamma function on input. f(x) = digamma(x)
* @tparam T: Input and output data types, half or float.
* @tparam isReuseSrc: Whether temporary variables can reuse the input memory.
* @param [out] dstTensor: output LocalTensor
* @param [in] srcTensor: input LocalTensor
* @param [in] sharedTmpBuffer: input local temporary Tensor
* @param [in] calCount: amount of data to be calculated
*/
template <typename T, bool isReuseSource = false>
__aicore__ inline void Digamma(LocalTensor<T>& dstTensor, const LocalTensor<T>& srcTensor,
LocalTensor<uint8_t>& sharedTmpBuffer, const uint32_t calCount)
{
if ASCEND_IS_AIC {
return;
}
DigammaCompute<T, isReuseSource>(dstTensor, srcTensor, sharedTmpBuffer, calCount);
}
* @ingroup Digamma
* @brief Computes the logarithmic derivative of the gamma function on input. f(x) = digamma(x)
* @tparam T: Input and output data types, half or float.
* @tparam isReuseSrc: Whether temporary variables can reuse the input memory.
* @param [out] dstTensor: output LocalTensor
* @param [in] srcTensor: input LocalTensor
* @param [in] calCount: amount of data to be calculated
*/
template <typename T, bool isReuseSource = false>
__aicore__ inline void Digamma(LocalTensor<T>& dstTensor, const LocalTensor<T>& srcTensor, const uint32_t calCount)
{
if ASCEND_IS_AIC {
return;
}
LocalTensor<uint8_t> tmp;
const bool ret = PopStackBuffer<uint8_t, TPosition::LCM>(tmp);
ASCENDC_ASSERT((ret), { KERNEL_LOG(KERNEL_ERROR, "PopStackBuffer Error!"); });
DigammaCompute<T, isReuseSource>(dstTensor, srcTensor, tmp, calCount);
}
#pragma end_pipe
}
#endif
#endif