Line data Source code
1 : /**
2 : * Copyright (c) 2026 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 : #ifndef HOST_CPU_URMA_CHANNEL_H
11 : #define HOST_CPU_URMA_CHANNEL_H
12 :
13 : #include "../channel.h"
14 : #include <mutex>
15 : #include "urma_types.h"
16 :
17 : // Orion
18 : #include "../../../../../../src/legacy/ascend950/unified_platform/resource/socket/socket.h"
19 : #include "../../sockets/socket_mgr.h"
20 : #include "ub_mem_transport.h"
21 : #include "host_ub_connection.h"
22 : #include "ub_local_notify.h"
23 : #include "aicpu_res_package_helper.h"
24 : #include "hcomm_adapter_hccp.h"
25 :
26 : namespace hcomm {
27 :
28 : class HostCpuUrmaChannel : public Channel {
29 : public:
30 : HostCpuUrmaChannel(EndpointHandle endpointHandle, const HcommChannelDesc &channelDesc);
31 : ~HostCpuUrmaChannel() override;
32 :
33 : HcclResult Init() override;
34 : HcclResult GetNotifyNum(uint32_t *notifyNum) const override;
35 : HcclResult GetRemoteMems(uint32_t *memNum, CommMem **remoteMem, char ***memInfos) override;
36 : ChannelStatus GetStatus() override;
37 0 : const HcommChannelDesc& GetChannelDesc() const override { return channelDesc_; }
38 :
39 : // 数据面接口
40 : HcclResult NotifyRecord(const uint32_t remoteNotifyIdx) override;
41 : HcclResult NotifyWait(const uint32_t localNotifyIdx, const uint32_t timeout) override;
42 : HcclResult WriteWithNotify(void *dst, const void *src, const uint64_t len, uint32_t remoteNotifyIdx) override;
43 : HcclResult Write(void *dst, const void *src, uint64_t len) override;
44 : HcclResult Read(void *dst, const void *src, uint64_t len) override;
45 : HcclResult ChannelFence() override;
46 :
47 : HcclResult Clean() override;
48 : HcclResult Resume() override;
49 :
50 : private:
51 : HcclResult ParseInputParam();
52 : HcclResult StartListen();
53 : HcclResult BuildSocket();
54 : HcclResult BuildConnection();
55 : HcclResult BuildUbMemTransport();
56 : HcclResult GetLocSeg(const void *addr, const size_t size, u64 *seg);
57 : HcclResult UrmaPostJettySendWr(urma_opcode_t opcode, void *dst, const void *src, uint64_t len);
58 : HcclResult GetSplitNum(uint64_t len, uint64_t maxJettyWrDataLen, uint64_t &splitNum);
59 : HcclResult GetLocalAndRemoteSeg(urma_opcode_t opcode, void *dst, const void *src, uint64_t len, u64 &localSeg, u64 &remoteSeg);
60 :
61 : private:
62 : // --------------------- 入参 ---------------------
63 : EndpointHandle endpointHandle_;
64 : HcommChannelDesc channelDesc_;
65 :
66 : // --------------------- 转换参数 ---------------------
67 : EndpointDesc localEp_{};
68 : EndpointDesc remoteEp_{};
69 :
70 : // --------------------- 具体成员 ---------------------
71 : Hccl::Socket* socket_{nullptr};
72 : const Hccl::SocketConfig* socketConfig_{nullptr};
73 : RdmaHandle rdmaHandle_{nullptr};
74 : std::unique_ptr<Hccl::UbMemTransport> memTransport_{nullptr};
75 : Hccl::BaseMemTransport::Attribution attr_{};
76 : Hccl::BaseMemTransport::CommonLocRes commonRes_{};
77 : std::vector<std::unique_ptr<Hccl::HostUbConnection>> connections_{};
78 : std::unique_ptr<Hccl::Socket> serverSocket_{nullptr};
79 :
80 : urma_jfc_t jfc_{};
81 : urma_jetty_t jetty_{};
82 : urma_target_jetty_t tjetty_{};
83 : uint32_t wqeNum_{0};
84 : bool fenceFlag_{false};
85 :
86 : std::mutex jfcMutex_;
87 : std::mutex fenceMutex_;
88 : bool Onetime_{false};
89 : DevBaseAttr devBaseAttr_{};
90 : uint32_t devicePhyId_{};
91 : };
92 :
93 : } // namespace hcomm
94 :
95 : #endif // HOST_CPU_URMA_CHANNEL_H
|