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              : #ifndef HCCLV2_RANK_INFO_DETECT_CLIENT_H
      12              : #define HCCLV2_RANK_INFO_DETECT_CLIENT_H
      13              : 
      14              : #include "socket.h"
      15              : #include "new_rank_info.h"
      16              : #include "rank_table_info.h"
      17              : #include "json_parser.h"
      18              : #include "internal_exception.h"
      19              : #include "timeout_exception.h"
      20              : #include "socket_exception.h"
      21              : #include "socket_agent.h"
      22              : #include "root_handle_v2.h"
      23              : 
      24              : namespace Hccl {
      25              : 
      26              : #ifdef OPEN_HCCL_TEST
      27              : const u32 WAIT_ERROR_BROADCAST_TIME = 1;
      28              : #else
      29              : const u32 WAIT_ERROR_BROADCAST_TIME = 20;
      30              : #endif
      31              : 
      32              : class RankInfoDetectClient {
      33              : public:
      34           22 :     RankInfoDetectClient(u32 devPhyId, u32 rankSize, u32 rankId, const std::shared_ptr<Socket> &clientSocket)
      35           22 :         : devPhyId_(devPhyId), rankSize_(rankSize), rankId_(rankId), clientSocket_(clientSocket), socketAgent_(clientSocket.get())
      36              :     {
      37           22 :     }
      38              :     ~RankInfoDetectClient();
      39              : 
      40              :     void Setup(RankTableInfo &rankTable);
      41              : 
      42              : private:
      43              :     u32                             devPhyId_{0};
      44              :     u32                             rankSize_{0};
      45              :     u32                             rankId_{0};
      46              :     std::shared_ptr<Socket>         clientSocket_{nullptr};
      47              :     u32                             currentStep_{0};
      48              :     RankTableInfo                   rankTable_{};
      49              :     SocketAgent                     socketAgent_;
      50              :     std::mutex hostSocketLock_;
      51              :     std::shared_ptr<Socket> hostSocket_ = nullptr;
      52              :     
      53              :     void SetupHostListenPort(u32 devLogicId, u32 devPhyId, const IpAddress &hostIp, uint32_t &hostPort);
      54              :     void SocketTearDown(u32 devPhyId);
      55              :     void Connect();
      56              :     void CheckStatus();
      57              :     void SendAgentIdAndRankSize();
      58              :     void SendLocalRankTable(const RankTableInfo &localRankTable);
      59              :     void ConstructRankTable(RankTableInfo &localRankTable);
      60              :     void VerifyRankTable();
      61              :     void RecvRankTable();
      62              :     void RecvRankTableMsg(vector<char> &rankInfoMsg);
      63              :     void ParseRankTable(vector<char> &rankInfoMsg);
      64              :     void GetLocalRankTableJson(const nlohmann::json &parseJson, nlohmann::json &localRankTableJson);
      65              :     void GetLocalDevInfoJson(const nlohmann::json &parseJson, nlohmann::json &localDevInfoJson);
      66              :     void ConstructSingleRank(RankTableInfo &localRankTable);
      67              :     HcclResult GetLocalTlsStatus(TlsStatus &tlsStatus) const;
      68              :     HcclResult VerifyTlsConsistency() const;
      69              :     void GenerateTlsStatusStr(std::string &tlsStatusStr, const std::vector<u32> &tlsStatusRanks) const;
      70              :     void ReportTlsConfigurationError(
      71              :         const std::string &tlsInconsistentTlsType, const std::string &tlsEnableRankStr,
      72              :         const std::string &tlsDisableRankStr, const std::string &tlsUnknownRankStr) const;
      73              :     void TearDown();
      74              :     void HostListenPortDetect(NewRankInfo &rankInfo);
      75              : };
      76              : 
      77              : } // namespace Hccl
      78              : #endif // HCCLV2_RANK_INFO_DETECT_CLIENT_H