* 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 swish_3510_impl.h
* \brief
*/
#if !defined(__ASCENDC_INCLUDE_INTERNAL_HEADERS__)
#pragma message( \
"impl/adv_api/detail/activation/swish/swish_3510_impl.h is an internal header file and must not be used directly. Functions or variables defined in this file may be removed in the future. Please use \"#include \"adv_api/activation/swish.h\"\" and use public functions or variables defined in interface headers files.")
#define __ASCENDC_INCLUDE_INTERNAL_HEADERS__
#define __UNDEF_ASCENDC_INCLUDE_INTERNAL_HEADERS_SWISH_C310_IMPL_H__
#endif
#ifndef IMPL_MATH_SWISH_SWISH_C310_IMPL_H
#define IMPL_MATH_SWISH_SWISH_C310_IMPL_H
#include "kernel_tensor.h"
#include "kernel_basic_intf.h"
#include "../../common/check.h"
namespace AscendC {
namespace Internal {
template <typename T>
__simd_vf__ inline void SwishComputeVF(
__ubuf__ T* dst, __ubuf__ T* src, uint32_t count, const T scalarValue, const uint16_t repeatTimes)
{
constexpr uint32_t oneRepElm = static_cast<uint32_t>(GetVecLen() / sizeof(T));
Reg::RegTensor<T> srcVreg;
Reg::RegTensor<T> vreg0;
Reg::RegTensor<T> dstVreg;
Reg::MaskReg mask;
for (uint16_t i = 0; i < repeatTimes; ++i) {
mask = Reg::UpdateMask<T>(count);
Reg::LoadAlign(srcVreg, src + i * oneRepElm);
Reg::Muls(vreg0, srcVreg, scalarValue, mask);
Reg::Exp(vreg0, vreg0, mask);
Reg::Adds(vreg0, vreg0, 1.0f, mask);
Reg::Div(dstVreg, srcVreg, vreg0, mask);
Reg::StoreAlign(dst + i * oneRepElm, dstVreg, mask);
}
}
}
template <typename T, bool isReuseSource = false>
__aicore__ inline void SwishCompute(
const LocalTensor<T>& dstLocal, const LocalTensor<T>& srcLocal, const uint32_t count, const T scalarValue)
{
if ASCEND_IS_AIC {
return;
}
static_assert(SupportType<T, half, float>(), "Swish only support half/float data type on current device!");
CheckTensorPosition(dstLocal, "dstLocal", "VECIN, VECOUT, VECCALC");
CheckTensorPosition(srcLocal, "srcLocal", "VECIN, VECOUT, VECCALC");
CheckCalCount(count, "count", dstLocal, "dstLocal", "Swish");
CheckCalCount(count, "count", srcLocal, "srcLocal", "Swish");
float negOne = -1.0;
constexpr uint32_t oneRepElm = static_cast<uint32_t>(GetVecLen() / sizeof(T));
uint16_t repeatTimes = static_cast<uint16_t>(CeilDivision(count, oneRepElm));
const T scalar = static_cast<T>(negOne * static_cast<float>(scalarValue));
Internal::SwishComputeVF<T>(
(__ubuf__ T*)dstLocal.GetPhyAddr(), (__ubuf__ T*)srcLocal.GetPhyAddr(), count, scalar, repeatTimes);
}
}
#endif
#if defined(__UNDEF_ASCENDC_INCLUDE_INTERNAL_HEADERS_SWISH_C310_IMPL_H__)
#undef __ASCENDC_INCLUDE_INTERNAL_HEADERS__
#undef __UNDEF_ASCENDC_INCLUDE_INTERNAL_HEADERS_SWISH_C310_IMPL_H__
#endif