* 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.
*/
#if !defined(__ASCENDC_INCLUDE_INTERNAL_HEADERS__)
#pragma message( \
"impl/adv_api/detail/math/trunc/trunc_common_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/math/trunc.h\"\" and use public functions or variables defined in interface headers files.")
#define __ASCENDC_INCLUDE_INTERNAL_HEADERS__
#define __UNDEF_ASCENDC_INCLUDE_INTERNAL_HEADERS_MATH_TRUNC_TRUNC_COMMON_IMPL_H__
#endif
#ifndef IMPL_MATH_TRUNC_TRUNC_COMMON_IMPL_H
#define IMPL_MATH_TRUNC_TRUNC_COMMON_IMPL_H
#include "kernel_basic_intf.h"
#include "kernel_tensor.h"
#include "../../common/check.h"
#ifdef ASCENDC_CPU_DEBUG
#include "../../api_check/kernel_check/math/trunc/trunc_check.h"
#endif
#include "../../api_check/kernel_api_check.h"
#if defined(__NPU_ARCH__) && __NPU_ARCH__ == 2002
#include "trunc_v200_impl.h"
#elif defined(__NPU_ARCH__) && __NPU_ARCH__ == 2201
#include "trunc_v220_impl.h"
#endif
namespace AscendC {
__aicore__ inline void TruncCompute(
const LocalTensor<half>& dstTensor, const LocalTensor<half>& srcTensor, const LocalTensor<uint8_t>& tmpTensor)
{
const LocalTensor<float> floatTmpTensor = tmpTensor.ReinterpretCast<float>();
Cast<float, half, false>(
floatTmpTensor, srcTensor, RoundMode::CAST_NONE, MASK_PLACEHOLDER, 1,
{1, 1, DEFAULT_REPEAT_STRIDE, HALF_DEFAULT_REPEAT_STRIDE});
PipeBarrier<PIPE_V>();
TruncCastForTrunc(floatTmpTensor, floatTmpTensor, tmpTensor);
Cast<half, float, false>(
dstTensor, floatTmpTensor, RoundMode::CAST_NONE, MASK_PLACEHOLDER, 1,
{1, 1, HALF_DEFAULT_REPEAT_STRIDE, DEFAULT_REPEAT_STRIDE});
PipeBarrier<PIPE_V>();
}
__aicore__ inline void TruncCompute(
const LocalTensor<float>& dstTensor, const LocalTensor<float>& srcTensor, const LocalTensor<uint8_t>& tmpTensor)
{
TruncCastForTrunc(dstTensor, srcTensor, tmpTensor);
}
template <typename T, bool isReuseSource = false>
__aicore__ inline void TruncImpl(
const LocalTensor<T>& dstTensor, const LocalTensor<T>& srcTensor, const LocalTensor<uint8_t>& sharedTmpBuffer,
const uint32_t calCount)
{
if ASCEND_IS_AIC {
return;
}
CHECK_FUNC_HIGHLEVEL_API(Trunc, (T, isReuseSource), (dstTensor, srcTensor, sharedTmpBuffer, calCount));
uint32_t tmpBufferSize = sharedTmpBuffer.GetSize();
uint32_t splitCount = tmpBufferSize / sizeof(T);
constexpr uint32_t TRUNC_HALF_CALC_PROCEDURE = 2;
if constexpr (sizeof(T) == sizeof(half)) {
splitCount = splitCount / TRUNC_HALF_CALC_PROCEDURE / ONE_BLK_SIZE * ONE_BLK_SIZE;
} else {
splitCount = splitCount / ONE_BLK_SIZE * ONE_BLK_SIZE;
}
CheckTmpBufferSize(splitCount, 0, tmpBufferSize);
uint32_t loopCount = calCount / splitCount;
uint32_t calcTail = calCount % splitCount;
SetMaskCount();
SetVectorMask<half, MaskMode::COUNTER>(0, splitCount);
for (uint32_t i = 0; i < loopCount; ++i) {
TruncCompute(dstTensor[i * splitCount], srcTensor[i * splitCount], sharedTmpBuffer);
}
if (calcTail > 0) {
SetVectorMask<half>(0, calcTail);
TruncCompute(dstTensor[loopCount * splitCount], srcTensor[loopCount * splitCount], sharedTmpBuffer);
}
SetMaskNorm();
ResetMask();
}
template <typename T, bool isReuseSource = false>
__aicore__ inline void TruncImpl(
const LocalTensor<T>& dstTensor, const LocalTensor<T>& srcTensor, 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!"); });
TruncImpl<T, isReuseSource>(dstTensor, srcTensor, sharedTmpBuffer, calCount);
}
}
#endif
#if defined(__UNDEF_ASCENDC_INCLUDE_INTERNAL_HEADERS_MATH_TRUNC_TRUNC_COMMON_IMPL_H__)
#undef __ASCENDC_INCLUDE_INTERNAL_HEADERS__
#undef __UNDEF_ASCENDC_INCLUDE_INTERNAL_HEADERS_MATH_TRUNC_TRUNC_COMMON_IMPL_H__
#endif