Line data    Source code

       1              : /**
       2              :  * Copyright (c) 2025 Huawei Technologies Co., Ltd.
       3              :  * This program is free software, you can redistribute it and/or modify it under the terms and conditions of
       4              :  * CANN Open Software License Agreement Version 2.0 (the "License").
       5              :  * Please refer to the License for details. You may not use this file except in compliance with the License.
       6              :  * THIS SOFTWARE IS PROVIDED ON AN "AS IS" BASIS, WITHOUT WARRANTIES OF ANY KIND, EITHER EXPRESS OR IMPLIED,
       7              :  * INCLUDING BUT NOT LIMITED TO NON-INFRINGEMENT, MERCHANTABILITY, OR FITNESS FOR A PARTICULAR PURPOSE.
       8              :  * See LICENSE in the root of the software repository for the full text of the License.
       9              :  */
      10              :  
      11              : #include "all_gather_ahc.h"
      12              : #include "alg_template_register.h"
      13              : 
      14              : namespace hccl {
      15              :  
      16            0 : AllGatherAHC::AllGatherAHC(const HcclDispatcher dispatcher)
      17            0 :     : AllGatherAHCBase(dispatcher)
      18              : {
      19            0 : }
      20              :  
      21            0 : AllGatherAHC::~AllGatherAHC()
      22              : {
      23            0 : }
      24              :  
      25            0 : HcclResult AllGatherAHC::DisposeSubGroups(const u32 rank)
      26              : {
      27            0 :     CommAHCBaseInfo::DisposeSubGroups(rank, globalSubGroups_, level0SubGroups_, level1SubGroups_);
      28            0 :     return HCCL_SUCCESS;
      29              : }
      30              :  
      31            0 : HcclResult AllGatherAHC::CommAHCInfoInit()
      32              : {
      33            0 :     commAHCBaseInfo_.reset(new (std::nothrow) CommAHCAlignInfo(level0SubGroups_));
      34            0 :     CHK_SMART_PTR_NULL(commAHCBaseInfo_);
      35            0 :     CHK_RET(commAHCBaseInfo_->Init(AHCOpType::AHC_OP_TYPE_ALLGATHER, ahcAlgOption_));
      36            0 :     return HCCL_SUCCESS;
      37              : }
      38              :  
      39            0 : HcclResult AllGatherAHC::RunInterAllGather(u32 rank, const std::vector<LINK> &links,
      40              :         const std::unique_ptr<CommAHCBaseInfo> &commAHCBaseInfo)
      41              : {
      42              :     // 获取当前rank的组间rank
      43            0 :     HCCL_INFO("[AllGatherAHC][RunInterAllGather] begin inter AllGather rank[%u]", rank);
      44              :  
      45            0 :     u32 interRank = commAHCBaseInfo->GetInterRank(0, rank);
      46              :  
      47              :     // 创建执行算子实列
      48            0 :     std::unique_ptr<AlgTemplateBase> tempAlg;
      49            0 :     commAHCBaseInfo->GetInterAlgTemplateOpInstance(AHCOpType::AHC_OP_TYPE_ALLGATHER, tempAlg, dispatcher_, reduceAttr_,
      50            0 :         extendFlag_, ahcExtendPreparePara_);
      51              :  
      52            0 :     std::vector<std::vector<Slice>> interSlicesVector;
      53            0 :     std::vector<std::vector<LINK>> interLinksVector;
      54            0 :     std::vector<u32>                logicCardList;
      55            0 :     CHK_RET(commAHCBaseInfo->CalcInterSlicesAndLinks(rank, DataUnitSize(dataType_), count_, links, interLinksVector, interSlicesVector, logicCardList));
      56              :  
      57            0 :     HCCL_DEBUG("[AllGatherAHC][RunInterAllGather] run inst rank[%u] interRank[%u]",
      58              :         rank, interRank);
      59              :  
      60            0 :     for (u32 i = 0; i < logicCardList.size(); i++) {
      61            0 :         std::vector<Slice> interSlices = interSlicesVector[i];
      62            0 :         std::vector<LINK> interLinks = interLinksVector[i];
      63            0 :         if (interLinks.size() <= 1) {
      64            0 :             continue;
      65              :         }
      66            0 :         CHK_RET(RunInstance(interRank, interLinks, interSlices, tempAlg, AHCOpType::AHC_OP_TYPE_ALLGATHER));
      67            0 :     }
      68              :  
      69            0 :     HCCL_DEBUG("[AllGatherAHC][RunInterAllGather] end inter AllGather rank[%u]", rank);
      70              :  
      71            0 :     return HCCL_SUCCESS;
      72            0 : }
      73              : REGISTER_TEMPLATE(TemplateType::TEMPLATE_ALL_GATHER_AHC, AllGatherAHC);
      74              : }