* 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 clamp.h
* \brief
*/
#if !defined(__ASCENDC_INCLUDE_INTERNAL_HEADERS__)
#define __ASCENDC_INCLUDE_INTERNAL_HEADERS__
#define __UNDEF_ASCENDC_INCLUDE_INTERNAL_HEADERS_CLAMP_H__
#endif
#ifndef LIB_MATH_CLAMP_H
#define LIB_MATH_CLAMP_H
#include "kernel_tensor.h"
#if defined(__NPU_ARCH__) && (__NPU_ARCH__ == 2002 || __NPU_ARCH__ == 2201)
#include "../../../impl/adv_api/detail/math/clamp/clamp_common_impl.h"
#elif defined(__NPU_ARCH__) && (__NPU_ARCH__ == 3510 || __NPU_ARCH__ == 5102)
#include "../../../impl/adv_api/detail/math/clamp/clamp_c310_impl.h"
#endif
#include "kernel_pop_stack_buffer.h"
#if defined(__NPU_ARCH__) && (__NPU_ARCH__ == 2201 || __NPU_ARCH__ == 2002 || __NPU_ARCH__ == 3510 || __NPU_ARCH__ == 5102)
namespace AscendC {
* \brief This function implements replaces numbers greater than scalar with scalar
* (e.g. ClampMax(2) means to replace numbers greater than 2 with 2 ). For details about the interface description, see
* https://pytorch.org/docs/stable/generated/torch.clamp.html.
*
* \note support data type: half and float
*
* \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
*/
* ClampMax *
* ************************************************************************************************* */
#pragma begin_pipe(V)
template <typename T, bool isReuseSource = false>
__aicore__ inline void ClampMax(const LocalTensor<T>& dstTensor, const LocalTensor<T>& srcTensor,
const LocalTensor<uint8_t>& sharedTmpBuffer, const T scalar, const uint32_t calCount)
{
if ASCEND_IS_AIC {
return;
}
ClampMaxImpl<T, isReuseSource>(dstTensor, srcTensor, sharedTmpBuffer, scalar, calCount);
}
template <typename T, bool isReuseSource = false>
__aicore__ inline void ClampMax(const LocalTensor<T>& dstTensor, const LocalTensor<T>& srcTensor, const T scalar,
const uint32_t calCount)
{
if ASCEND_IS_AIC {
return;
}
LocalTensor<uint8_t> sharedTmpBuffer;
bool ret = PopStackBuffer<uint8_t, TPosition::LCM>(sharedTmpBuffer);
ASCENDC_ASSERT((ret), { KERNEL_LOG(KERNEL_ERROR, "PopStackBuffer Error!"); });
ClampMaxImpl<T, isReuseSource>(dstTensor, srcTensor, sharedTmpBuffer, scalar, calCount);
}
* \brief This function implements replace numbers less than scalar with scalar
* (e.g. ClampMin(2) means to replace numbers less than 2 with 2 ). For details about the interface description, see
* https://pytorch.org/docs/stable/generated/torch.clamp.html.
*
* \note support data type: half and float
*
* \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
*/
* ClampMin *
* ************************************************************************************************* */
template <typename T, bool isReuseSource = false>
__aicore__ inline void ClampMin(const LocalTensor<T>& dstTensor, const LocalTensor<T>& srcTensor,
const LocalTensor<uint8_t>& sharedTmpBuffer, const T scalar, const uint32_t calCount)
{
if ASCEND_IS_AIC {
return;
}
ClampMinImpl<T, isReuseSource>(dstTensor, srcTensor, sharedTmpBuffer, scalar, calCount);
}
template <typename T, bool isReuseSource = false>
__aicore__ inline void ClampMin(const LocalTensor<T>& dstTensor, const LocalTensor<T>& srcTensor, const T scalar,
const uint32_t calCount)
{
if ASCEND_IS_AIC {
return;
}
LocalTensor<uint8_t> sharedTmpBuffer;
bool ret = PopStackBuffer<uint8_t, TPosition::LCM>(sharedTmpBuffer);
ASCENDC_ASSERT((ret), { KERNEL_LOG(KERNEL_ERROR, "PopStackBuffer Error!"); });
ClampMinImpl<T, isReuseSource>(dstTensor, srcTensor, sharedTmpBuffer, scalar, calCount);
}
#if defined(__NPU_ARCH__) && (__NPU_ARCH__ == 3510 || __NPU_ARCH__ == 5102)
* \brief This function implements replaces numbers greater than min with min and less than max with max
* (e.g. Clamp(2, 5) means to replace numbers greater than 2 with 2 and less than 5 with 5). For details about the interface description, see
* https://pytorch.org/docs/stable/generated/torch.clamp.html.
*
* \param [out] dstTensor, output LocalTensor
* \param [in] srcTensor, input LocalTensor
* \param [in] min, input the min value
* \param [in] max, input the max value
* \param [in] count, amount of data to be calculated
*/
* Clamp *
* ***************************************************************************************************/
template <const ClampConfig& config = DEFAULT_CLAMP_CONFIG, typename T, typename U, typename S>
__aicore__ inline void Clamp(const LocalTensor<T>& dst, const LocalTensor<T>& src, const U& min, const S& max,
const uint32_t count)
{
ClampImpl<config, T, U, S>(dst, src, min, max, count);
}
#endif
#pragma end_pipe
}
#endif
#endif
#if defined(__UNDEF_ASCENDC_INCLUDE_INTERNAL_HEADERS_CLAMP_H__)
#undef __ASCENDC_INCLUDE_INTERNAL_HEADERS__
#undef __UNDEF_ASCENDC_INCLUDE_INTERNAL_HEADERS_CLAMP_H__
#endif
