Files
enginex-ascend-910-vllm/csrc/scripts/opgen/template/add/op_kernel/add_example.h
Sun Ruoxi 7f8a1b1f7a init v0.23.0
Signed-off-by: Sun Ruoxi <sunruoxi@4paradigm.com>
2026-08-27 15:11:51 +08:00

117 lines
3.9 KiB
C++
Raw Blame History

This file contains invisible Unicode characters

This file contains invisible Unicode characters that are indistinguishable to humans but may be processed differently by a computer. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

/**
 * 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 add_example.h
* \brief
*/
#ifndef ADD_EXAMPLE_H
#define ADD_EXAMPLE_H
#include "kernel_operator.h"
#include "kernel_tiling/kernel_tiling.h"
#include "add_example_tiling_data.h"
#include "add_example_tiling_key.h"
namespace NsAddExample {
using namespace AscendC;
constexpr int32_t BUFFER_NUM = 2;
template <typename T>
class AddExample
{
public:
__aicore__ inline AddExample(){};
__aicore__ inline void Init(GM_ADDR x, GM_ADDR y, GM_ADDR z, const AddExampleTilingData* tilingData);
__aicore__ inline void Process();
private:
__aicore__ inline void CopyIn(int32_t progress);
__aicore__ inline void CopyOut(int32_t progress);
__aicore__ inline void Compute(const int32_t dataLength);
private:
TPipe pipe;
TQue<QuePosition::VECIN, BUFFER_NUM> inputQueueX;
TQue<QuePosition::VECIN, BUFFER_NUM> inputQueueY;
TQue<QuePosition::VECOUT, BUFFER_NUM> outputQueueZ;
GlobalTensor<T> inputGMX;
GlobalTensor<T> inputGMY;
GlobalTensor<T> outputGMZ;
int64_t blockLength_ = 0;
int64_t tileNum_ = 0;
uint32_t tileLength_ = 0;
};
template <typename T>
__aicore__ inline void AddExample<T>::Init(GM_ADDR x, GM_ADDR y, GM_ADDR z, const AddExampleTilingData* tilingData)
{
blockLength_ = tilingData->totalLength / AscendC::GetBlockNum();
tileNum_ = tilingData->tileNum;
tileLength_ = blockLength_ / tileNum_ / BUFFER_NUM;
inputGMX.SetGlobalBuffer((__gm__ T*)x + blockLength_ * AscendC::GetBlockIdx(), blockLength_);
inputGMY.SetGlobalBuffer((__gm__ T*)y + blockLength_ * AscendC::GetBlockIdx(), blockLength_);
outputGMZ.SetGlobalBuffer((__gm__ T*)z + blockLength_ * AscendC::GetBlockIdx(), blockLength_);
pipe.InitBuffer(inputQueueX, BUFFER_NUM, tileLength_ * sizeof(T));
pipe.InitBuffer(inputQueueY, BUFFER_NUM, tileLength_ * sizeof(T));
pipe.InitBuffer(outputQueueZ, BUFFER_NUM, tileLength_ * sizeof(T));
}
template <typename T>
__aicore__ inline void AddExample<T>::CopyIn(int32_t progress)
{
AscendC::LocalTensor<T> xLocal = inputQueueX.AllocTensor<T>();
AscendC::LocalTensor<T> yLocal = inputQueueY.AllocTensor<T>();
AscendC::DataCopy(xLocal, inputGMX[progress * tileLength_], tileLength_);
AscendC::DataCopy(yLocal, inputGMY[progress * tileLength_], tileLength_);
inputQueueX.EnQue(xLocal);
inputQueueY.EnQue(yLocal);
}
template <typename T>
__aicore__ inline void AddExample<T>::CopyOut(int32_t progress)
{
AscendC::LocalTensor<T> zLocal = outputQueueZ.DeQue<T>();
AscendC::DataCopy(outputGMZ[progress * tileLength_], zLocal, tileLength_);
outputQueueZ.FreeTensor(zLocal);
}
template <typename T>
__aicore__ inline void AddExample<T>::Compute(int32_t progress)
{
AscendC::LocalTensor<T> xLocal = inputQueueX.DeQue<T>();
AscendC::LocalTensor<T> yLocal = inputQueueY.DeQue<T>();
AscendC::LocalTensor<T> zLocal = outputQueueZ.AllocTensor<T>();
AscendC::Add(zLocal, xLocal, yLocal, tileLength_);
outputQueueZ.EnQue<T>(zLocal);
inputQueueX.FreeTensor(xLocal);
inputQueueY.FreeTensor(yLocal);
}
template <typename T>
__aicore__ inline void AddExample<T>::Process()
{
int32_t loopCount = tileNum_ * BUFFER_NUM;
for (int32_t i = 0; i < loopCount; i++) {
CopyIn(i);
Compute(i);
CopyOut(i);
}
}
} // namespace NsAddExample
#endif // ADD_EXAMPLE_H