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 INDEPENDENT_OP_RESOURCE_MANAGER_H
      12              : #define INDEPENDENT_OP_RESOURCE_MANAGER_H
      13              : 
      14              : #include <memory>
      15              : #include <string>
      16              : #include <unordered_set>
      17              : #include <atomic>
      18              : #include "hccl/hccl_res.h"
      19              : #include "hccl_independent_common.h"
      20              : #include "comm_mem_manager.h"
      21              : #include "comm_engine_res_manager.h"
      22              : #include "rank_graph.h"
      23              : #include "comm_config_pub.h"
      24              : #include "independent_op_context_manager.h"
      25              : #include "channel_manager.h"
      26              : #include "aicpu_init_param.h"
      27              : 
      28              : namespace hccl {
      29              : constexpr int32_t  HCCL_COMM_ENGINE_CONFIG_NOT_SET = -1;
      30              : constexpr uint32_t HCCL_COMM_THREADNUM_CONFIG_NOT_SET = 0xffffffff;
      31              : constexpr uint32_t HCCL_COMM_NOTIFY_NUM_PER_THREAD_CONFIG_NOT_SET = 0xffffffff;
      32              : 
      33              : class IndependentOp {
      34              : public:
      35              :     IndependentOp();
      36          604 :     ~IndependentOp() = default;
      37              : 
      38              :     // 初始化资源管理器
      39              :     HcclResult SetIndependentOpConfig(const CommConfig &commConfig, const RankTable_t &rankTable,
      40              :         const HcclTopoAttr &topoAttr, const aclrtBinHandle binHandle, HDCommunicateParams &kfcControlTransferH2DParams,
      41              :         HDCommunicateParams &kfcStatusTransferD2HParams, CCLBufferManager &bufferManager);
      42              :     HcclResult SetChannelCallbacks(const ChannelManagerCallbacks& channelCallbacks);
      43              : 
      44              :     // 获取配置信息
      45              :     u32 GetThreadNum() const { return threadNum_; }
      46              :     u32 GetNotifyNumPerThread() const { return notifyNumPerThread_; }
      47              : 
      48              :     bool GetAicpuCommState(); // 是否线程安全
      49              :     void SetAicpuCommState(bool aicpuCommState);
      50              : 
      51            7 :     inline CommMemMgr& GetCommMemMgr() {
      52            7 :         return commMemMgr_;
      53              :     }
      54              : 
      55           22 :     inline CommEngineResMgr& GetCommEngineResMgr() {
      56           22 :         return engineResMgr_;
      57              :     }
      58              : 
      59           52 :     inline ContextManager& GetContextManager() {
      60           52 :         return contextMgr_;
      61              :     }
      62              : 
      63          402 :     inline ChannelManager& GetChannelManager() {
      64          402 :         return channelMgr_;
      65              :     }
      66              : 
      67              :     // 自定义算子Aicpu通信域公共初始化
      68              :     HcclResult KernelLaunchAicpuCommInit();
      69              : 
      70              : private:
      71              :     // config内容
      72              :     int32_t commEngine_ = -1;
      73              :     u32 threadNum_ = 0;
      74              :     u32 notifyNumPerThread_ = 0;
      75              :     u64 cclBufferSize_ = 0;
      76              :     std::string commId_;
      77              :     aclrtBinHandle binHandle_ = nullptr;
      78              : 
      79              :     // 管理器
      80              :     CommMemMgr commMemMgr_;
      81              :     CommEngineResMgr engineResMgr_;
      82              :     ContextManager contextMgr_;
      83              :     ChannelManager channelMgr_;
      84              : 
      85              :     bool isAicpuCommInit_ = false;
      86              :     CommAicpuParam commAicpuParam_{};
      87              : };
      88              : 
      89              : }  // namespace hccl
      90              : #endif