镜像启动torch.npu.init()报错

一、问题现象（附报错日志上下文）：

打开日志后显示：
bash-5.1$ ASCEND_SLOG_PRINT_TO_STDOUT=1 python /home/ma-user/work/pangu_rl_code/code/init.py
[INFO] ATRACE(173034,python):2025-09-01-22:48:01.876.618 [trace_attr.c:189](tid:173034) attr init success, timeout=0ms.
[INFO] ATRACE(173034,python):2025-09-01-22:48:01.876.676 [trace_recorder.c:121](tid:173034) use root path: /home/ma-user/ascend
[ERROR] ATRACE(173034,python):2025-09-01-22:48:01.877.990 [stacktrace_dumper.c:209](tid:173034) can not get realpath, path=/usr/local/Ascend/ascend-toolkit/8.1.RC1/aarch64-linux/lib64/../bin/asc_dumper, strerr=Operation not permitted.
[ERROR] ATRACE(173034,python):2025-09-01-22:48:01.878.003 [stacktrace_dumper.c:238](tid:173034) get bin path failed, ret=-1.
[ERROR] ATRACE(173034,python):2025-09-01-22:48:01.878.012 [trace_core.c:47](tid:173034) init stacktrace dumper failed, ret=-1.
[EVENT] PROFILING(173034,python):2025-09-01-22:48:01.881.738 [msprof_callback_impl.cpp:358] >>> (tid:173034) Started to register profiling ctrl callback.
[EVENT] PROFILING(173034,python):2025-09-01-22:48:01.881.932 [msprof_callback_impl.cpp:365] >>> (tid:173034) Started to register profiling hash id callback.
[INFO] PROFILING(173034,python):2025-09-01-22:48:01.881.944 [prof_atls_plugin.cpp:117] (tid:173034) RegisterProfileCallback, callback type is 7
[EVENT] PROFILING(173034,python):2025-09-01-22:48:01.881.951 [msprof_callback_impl.cpp:372] >>> (tid:173034) Started to register profiling enable host freq callback.
[INFO] PROFILING(173034,python):2025-09-01-22:48:01.881.957 [prof_atls_plugin.cpp:117] (tid:173034) RegisterProfileCallback, callback type is 8
[INFO] RUNTIME(173034,python):2025-09-01-22:48:01.883.368 [task_fail_callback_data_manager.cc:58] 173034 TaskFailCallBackManager: Constructor.
[INFO] HCCL(173034,python):2025-09-01-22:48:01.953.756 [adapter_rts.cc:2720] [173034][adapter_rts.cc][CallBackInitRts] g_deviceType [6] g_deviceLogicId [-1] g_devicePhyId [-1]
/usr/local/python3.10.17/lib/python3.10/site-packages/torch_npu/utils/collect_env.py:58: UserWarning: Warning: The /usr/local/Ascend/ascend-toolkit/latest owner does not match the current owner.
  warnings.warn(f"Warning: The {path} owner does not match the current owner.")
/usr/local/python3.10.17/lib/python3.10/site-packages/torch_npu/utils/collect_env.py:58: UserWarning: Warning: The /usr/local/Ascend/ascend-toolkit/8.1.RC1/aarch64-linux/ascend_toolkit_install.info owner does not match the current owner.
  warnings.warn(f"Warning: The {path} owner does not match the current owner.")
/usr/local/python3.10.17/lib/python3.10/site-packages/torch_npu/utils/_path_manager.py:67: UserWarning: Permission mismatch: The owner of /usr/local/python3.10.17/lib/python3.10/site-packages/torch_npu/lib/libop_plugin_atb.so does not match.
  warnings.warn(f"Permission mismatch: The owner of {path} does not match.")
[INFO] PROFILING(173034,python):2025-09-01-22:48:02.409.725 [prof_atls_plugin.cpp:210] (tid:173034) Module[0] register callback of ctrl handle.
/usr/local/python3.10.17/lib/python3.10/site-packages/torch_npu/utils/collect_env.py:58: UserWarning: Warning: The /usr/local/Ascend/ascend-toolkit/latest owner does not match the current owner.
  warnings.warn(f"Warning: The {path} owner does not match the current owner.")
/usr/local/python3.10.17/lib/python3.10/site-packages/torch_npu/utils/collect_env.py:58: UserWarning: Warning: The /usr/local/Ascend/ascend-toolkit/8.1.RC1/aarch64-linux/ascend_toolkit_install.info owner does not match the current owner.
  warnings.warn(f"Warning: The {path} owner does not match the current owner.")
[INFO] PROFILING(173034,python):2025-09-01-22:48:02.567.466 [prof_atls_plugin.cpp:210] (tid:173034) Module[61] register callback of ctrl handle.
[INFO] PROFILING(173034,python):2025-09-01-22:48:02.605.294 [prof_atls_plugin.cpp:210] (tid:173034) Module[45] register callback of ctrl handle.
[INFO] GE(173034,python):2025-09-01-22:48:02.778.254 [op_tiling_manager.cc:109]173034 ~FuncPerfScope:[GEPERFTRACE] The time cost of OpTilingManager::LoadSo is [172905] micro seconds.
[INFO] PROFILING(173034,python):2025-09-01-22:48:02.826.646 [prof_atls_plugin.cpp:210] (tid:173034) Module[6] register callback of ctrl handle.
[INFO] RUNTIME(173034,python):2025-09-01-22:48:02.835.246 [runtime_keeper.cc:149] 173034 CreateRuntimeImpl: Open libruntime_v100.so success
[INFO] RUNTIME(173034,python):2025-09-01-22:48:02.835.314 [runtime.cc:5060] 173034 GetVisibleDevices: ASCEND_RT_VISIBLE_DEVICES param was not set
[INFO] PROFILING(173034,python):2025-09-01-22:48:02.837.501 [prof_atls_plugin.cpp:210] (tid:173034) Module[7] register callback of ctrl handle.
[INFO] IDEDD(173034,python):2025-09-01-22:48:02.838.001 [adx_dump_record.cpp:629][tid:173256] start dump thread, remote dump record temp path : /home/ma-user/work/.
[INFO] IDEDD(173034,python):2025-09-01-22:48:02.838.173 [adx_server_manager.cpp:203][tid:173257] Run Server(0) Process
[INFO] IDEDD(173034,python):2025-09-01-22:48:02.838.244 [hdc_comm_opt.cpp:59][tid:173257] open device[0] server[13]
[INFO] DRV(173034,python):2025-09-01-22:48:02.838.349 [drv_log_user.c:623][ascend][curpid:173034,173257][drv][hdc][share_log_read_in_single_module]Server create success.(dev_id=0; fid=0; service_type="service_dump"; pid=1889254)
[INFO] IDEDD(173034,python):2025-09-01-22:48:02.838.365 [hdc_api.cpp:119][tid:173257] logDevId 0 create HDC server successfully
[INFO] IDEDD(173034,python):2025-09-01-22:48:02.838.403 [hdc_comm_opt.cpp:59][tid:173257] open device[1] server[13]
[INFO] DRV(173034,python):2025-09-01-22:48:02.838.524 [drv_log_user.c:623][ascend][curpid:173034,173257][drv][hdc][share_log_read_in_single_module]Server create success.(dev_id=1; fid=0; service_type="service_dump"; pid=1889254)
[INFO] IDEDD(173034,python):2025-09-01-22:48:02.838.532 [hdc_api.cpp:119][tid:173257] logDevId 1 create HDC server successfully
[INFO] IDEDD(173034,python):2025-09-01-22:48:02.838.544 [hdc_comm_opt.cpp:59][tid:173257] open device[2] server[13]
[INFO] DRV(173034,python):2025-09-01-22:48:02.838.673 [drv_log_user.c:623][ascend][curpid:173034,173257][drv][hdc][share_log_read_in_single_module]Server create success.(dev_id=2; fid=0; service_type="service_dump"; pid=1889254)
[INFO] IDEDD(173034,python):2025-09-01-22:48:02.838.681 [hdc_api.cpp:119][tid:173257] logDevId 2 create HDC server successfully
[INFO] IDEDD(173034,python):2025-09-01-22:48:02.838.690 [hdc_comm_opt.cpp:59][tid:173257] open device[3] server[13]
[INFO] DRV(173034,python):2025-09-01-22:48:02.838.810 [drv_log_user.c:623][ascend][curpid:173034,173257][drv][hdc][share_log_read_in_single_module]Server create success.(dev_id=3; fid=0; service_type="service_dump"; pid=1889254)
[INFO] IDEDD(173034,python):2025-09-01-22:48:02.838.820 [hdc_api.cpp:119][tid:173257] logDevId 3 create HDC server successfully
[INFO] IDEDD(173034,python):2025-09-01-22:48:02.838.829 [hdc_comm_opt.cpp:59][tid:173257] open device[4] server[13]
[INFO] DRV(173034,python):2025-09-01-22:48:02.838.946 [drv_log_user.c:623][ascend][curpid:173034,173257][drv][hdc][share_log_read_in_single_module]Server create success.(dev_id=4; fid=0; service_type="service_dump"; pid=1889254)
[INFO] IDEDD(173034,python):2025-09-01-22:48:02.838.953 [hdc_api.cpp:119][tid:173257] logDevId 4 create HDC server successfully
[INFO] IDEDD(173034,python):2025-09-01-22:48:02.838.961 [hdc_comm_opt.cpp:59][tid:173257] open device[5] server[13]
[INFO] DRV(173034,python):2025-09-01-22:48:02.839.079 [drv_log_user.c:623][ascend][curpid:173034,173257][drv][hdc][share_log_read_in_single_module]Server create success.(dev_id=5; fid=0; service_type="service_dump"; pid=1889254)
[INFO] IDEDD(173034,python):2025-09-01-22:48:02.839.086 [hdc_api.cpp:119][tid:173257] logDevId 5 create HDC server successfully
[INFO] IDEDD(173034,python):2025-09-01-22:48:02.839.095 [hdc_comm_opt.cpp:59][tid:173257] open device[6] server[13]
[INFO] DRV(173034,python):2025-09-01-22:48:02.839.225 [drv_log_user.c:623][ascend][curpid:173034,173257][drv][hdc][share_log_read_in_single_module]Server create success.(dev_id=6; fid=0; service_type="service_dump"; pid=1889254)
[INFO] IDEDD(173034,python):2025-09-01-22:48:02.839.233 [hdc_api.cpp:119][tid:173257] logDevId 6 create HDC server successfully
[INFO] IDEDD(173034,python):2025-09-01-22:48:02.839.242 [hdc_comm_opt.cpp:59][tid:173257] open device[7] server[13]
[INFO] DRV(173034,python):2025-09-01-22:48:02.839.375 [drv_log_user.c:623][ascend][curpid:173034,173257][drv][hdc][share_log_read_in_single_module]Server create success.(dev_id=7; fid=0; service_type="service_dump"; pid=1889254)
[INFO] IDEDD(173034,python):2025-09-01-22:48:02.839.383 [hdc_api.cpp:119][tid:173257] logDevId 7 create HDC server successfully
[INFO] PROFILING(173034,python):2025-09-01-22:48:02.842.985 [prof_atls_plugin.cpp:210] (tid:173034) Module[48] register callback of ctrl handle.
[EVENT] PROFILING(173034,python):2025-09-01-22:48:02.843.068 [msprof_callback_impl.cpp:110] >>> (tid:173034) MsprofCtrlCallback called, type: 255
[EVENT] PROFILING(173034,python):2025-09-01-22:48:02.843.461 [ai_drv_dev_api.cpp:333] >>> (tid:173034) Succeeded to DrvGetApiVersion version: 0x72317
[INFO] ATRACE(173034,python):2025-09-01-22:48:02.845.866 [atrace_client_core.c:160](tid:173034) start to receive device[0] trace log.
[INFO] TDT(173034,python):2025-09-01-22:48:02.845.957 [client_manager.cpp:435][IsHostEnvironment][tid:173034] [TsdClient] logicDeviceId is [0], hostaicpunum[0]
[INFO] TDT(173034,python):2025-09-01-22:48:02.845.970 [client_manager.cpp:464][GetClientRunMode][tid:173034] runningMode:0
[INFO] TDT(173034,python):2025-09-01-22:48:02.845.977 [client_manager.cpp:127][GetInstance][tid:173034] [ClientManager] Current mode:1
[INFO] TDT(173034,python):2025-09-01-22:48:02.845.998 [process_mode_base.cpp:149][GetLogLevel][tid:173034] [TsdClient] get ASCEND_GLOBAL_LOG_LEVEL [] ASCEND_GLOBAL_EVENT_ENABLE [] ASCEND_MODULE_LOG_LEVEL []
[INFO] TDT(173034,python):2025-09-01-22:48:02.846.076 [process_mode_manager.cpp:71][OpenProcess][tid:173034] [ProcessModeManager] enter into open process deviceId[0] rankSize[0]
[INFO] TDT(173034,python):2025-09-01-22:48:02.846.540 [process_mode_manager.cpp:403][InitTsdClient][tid:173034] [TsdClient] deviceId[0] begin to init hdc client
[INFO] TDT(173034,python):2025-09-01-22:48:02.846.651 [version_verify.cpp:36][SetVersionInfo][tid:173034] VersionVerify: send client version to server
[INFO] TDT(173034,python):2025-09-01-22:48:02.846.663 [version_verify.cpp:52][SetVersionInfo][tid:173034] send feature_info:{msg_type:35, features:{check before send aicpu package}}
[INFO] TDT(173034,python):2025-09-01-22:48:02.846.671 [version_verify.cpp:52][SetVersionInfo][tid:173034] send feature_info:{msg_type:37, features:{check before send open qs message}}
[INFO] TDT(173034,python):2025-09-01-22:48:02.846.679 [version_verify.cpp:52][SetVersionInfo][tid:173034] send feature_info:{msg_type:64, features:{get check code retry}}
[INFO] TDT(173034,python):2025-09-01-22:48:02.847.324 [version_verify.cpp:68][PeerVersionCheck][tid:173034] VersionVerify: Check client version info, server[1230], client[1230]
[INFO] TDT(173034,python):2025-09-01-22:48:02.847.335 [version_verify.cpp:89][ParseVersionInfo][tid:173034] VersionVerify: pass client version info success
[INFO] TDT(173034,python):2025-09-01-22:48:02.847.341 [hdc_client.cpp:275][CheckHdcConnection][tid:173034] Service[2] create hdc success
[INFO] TDT(173034,python):2025-09-01-22:48:02.847.351 [version_verify.cpp:122][SpecialFeatureCheck][tid:173034] VersionVerify: new type[35], supported
[INFO] TDT(173034,python):2025-09-01-22:48:02.847.385 [process_mode_manager.cpp:627][GetDeviceCheckCodeOnce][tid:173034] [TsdClient][deviceId=0] [sessionId=1] wait package info respond
[INFO] TDT(173034,python):2025-09-01-22:48:02.847.511 [version_verify.cpp:122][SpecialFeatureCheck][tid:173034] VersionVerify: new type[64], supported
[INFO] TDT(173034,python):2025-09-01-22:48:02.847.586 [process_mode_manager.cpp:403][InitTsdClient][tid:173034] [TsdClient] deviceId[0] begin to init hdc client
[INFO] TDT(173034,python):2025-09-01-22:48:02.847.656 [version_verify.cpp:36][SetVersionInfo][tid:173034] VersionVerify: send client version to server
[INFO] TDT(173034,python):2025-09-01-22:48:02.847.666 [version_verify.cpp:52][SetVersionInfo][tid:173034] send feature_info:{msg_type:35, features:{check before send aicpu package}}
[INFO] TDT(173034,python):2025-09-01-22:48:02.847.675 [version_verify.cpp:52][SetVersionInfo][tid:173034] send feature_info:{msg_type:37, features:{check before send open qs message}}
[INFO] TDT(173034,python):2025-09-01-22:48:02.847.683 [version_verify.cpp:52][SetVersionInfo][tid:173034] send feature_info:{msg_type:64, features:{get check code retry}}
[INFO] TDT(173034,python):2025-09-01-22:48:02.848.095 [version_verify.cpp:68][PeerVersionCheck][tid:173034] VersionVerify: Check client version info, server[1230], client[1230]
[INFO] TDT(173034,python):2025-09-01-22:48:02.848.104 [version_verify.cpp:89][ParseVersionInfo][tid:173034] VersionVerify: pass client version info success
[INFO] TDT(173034,python):2025-09-01-22:48:02.848.110 [hdc_client.cpp:275][CheckHdcConnection][tid:173034] Service[2] create hdc success
[INFO] TDT(173034,python):2025-09-01-22:48:02.848.119 [process_mode_manager.cpp:458][ConstructOpenMsg][tid:173034] [TsdClient] tsd get process sign successfully, procpid[1889254] signSize[48]
[INFO] TDT(173034,python):2025-09-01-22:48:02.848.128 [version_verify.cpp:114][SpecialFeatureCheck][tid:173034] VersionVerify: previous type[6], supported
[INFO] TDT(173034,python):2025-09-01-22:48:02.848.146 [process_mode_manager.cpp:88][OpenProcess][tid:173034] [ProcessModeManager] deviceId[0] sessionId[1] rankSize[0], wait sub process start respond
[INFO] DRV(173034,python):2025-09-01-22:48:03.039.316 [devmm_svm.c:100][ascend][curpid:173034,173034][drv][devmm][devmm_setup_device]DrvMemDeviceOpen. (devid=0)
[INFO] DRV(173034,python):2025-09-01-22:48:03.041.176 [devmm_virt_interface.c:394][ascend][curpid:173034,173034][drv][devmm][devmm_ioctl_init_process]Init_process details. (hostpid=173034)
[INFO] DRV(173034,python):2025-09-01-22:48:03.041.515 [drv_log_user.c:623][ascend][curpid:173034,173034][drv][devmm][share_log_read_in_single_module]Setup device succeeded. (logical_devid=0; devid=0; vfid=0; hostpid=1889254; devpid=21951)
[INFO] DRV(173034,python):2025-09-01-22:48:03.041.530 [devmm_svm.c:85][ascend][curpid:173034,173034][drv][devmm][devmm_set_device_info]Device info. (devid=0; dvpp_size=17179869184; support_bar_mem=1; support_dev_read_only=1; support_dev_mem_map_host=1; support_bar_huge_mem=1; is_alloced_by_malloc=0; host_rw_dev_ro=1; double_pgtable_offset=228698418577408)
[INFO] TDT(173034,python):2025-09-01-22:48:03.041.767 [stub_process_mode_nowin.cpp:60][ProcessQueueForMdc][tid:173034] [TsdClient] it is unnecessary of current mode[0] to grant queue auth to aicpusd
[INFO] TDT(173034,python):2025-09-01-22:48:03.041.779 [stub_process_mode_nowin.cpp:99][OpenInHost][tid:173034] enter into OpenInHost deviceid[0]
[INFO] TDT(173034,python):2025-09-01-22:48:03.041.802 [client_manager.cpp:435][IsHostEnvironment][tid:173034] [TsdClient] logicDeviceId is [0], hostaicpunum[0]
[INFO] TDT(173034,python):2025-09-01-22:48:03.041.812 [stub_process_mode_nowin.cpp:103][OpenInHost][tid:173034] host cpu not support
[INFO] TDT(173034,python):2025-09-01-22:48:03.041.820 [process_mode_manager.cpp:121][OpenProcess][tid:173034] [TsdClient][deviceId=0] [sessionId=1] start hccp and computer process success
[INFO] DRV(173034,python):2025-09-01-22:48:03.045.852 [devmm_svm.c:100][ascend][curpid:173034,173034][drv][devmm][devmm_setup_device]DrvMemDeviceOpen. (devid=0)
[INFO] DRV(173034,python):2025-09-01-22:48:03.045.873 [devmm_svm.c:85][ascend][curpid:173034,173034][drv][devmm][devmm_set_device_info]Device info. (devid=0; dvpp_size=17179869184; support_bar_mem=1; support_dev_read_only=1; support_dev_mem_map_host=1; support_bar_huge_mem=1; is_alloced_by_malloc=0; host_rw_dev_ro=1; double_pgtable_offset=228698418577408)
[INFO] RUNTIME(173034,python):2025-09-01-22:48:03.045.941 [raw_device.cc:346] 173034 Init: isAddrFlat:0
[INFO] TDT(173034,python):2025-09-01-22:48:03.046.856 [tsd_client.cpp:189][TsdCapabilityGet][tid:173034] TsdCapabilityGet Begin.
[INFO] TDT(173034,python):2025-09-01-22:48:03.046.868 [process_mode_manager.cpp:1042][CapabilityGet][tid:173034] [ProcessModeManager] enter into CapabilityGet process deviceId[0] type[0]
[INFO] TDT(173034,python):2025-09-01-22:48:03.046.879 [version_verify.cpp:114][SpecialFeatureCheck][tid:173034] VersionVerify: previous type[38], supported
[INFO] TDT(173034,python):2025-09-01-22:48:03.046.907 [process_mode_manager.cpp:1065][CapabilityGet][tid:173034] [TsdClient][deviceId=0] [sessionId=1] [type=0]send capability success 
[INFO] TDT(173034,python):2025-09-01-22:48:03.047.866 [process_mode_msg_parse.cpp:170][PidQosMsgProc][tid:173034] [TsdClient] PidQosMsgProc recvMsg realDeviceId[0] msgType[39] localDevId[0] rspCode[0],pidqos[1]
[INFO] TDT(173034,python):2025-09-01-22:48:03.047.876 [process_mode_manager.cpp:1070][CapabilityGet][tid:173034] [TsdClient][logicDeviceId_=0][type=0][pidQos=1][ret=0]recv capability respond finish
[INFO] RUNTIME(173034,python):2025-09-01-22:48:03.047.888 [engine.cc:78] 173034 Engine: Constructor.
[INFO] RUNTIME(173034,python):2025-09-01-22:48:03.047.896 [stars_engine.cc:51] 173034 StarsEngine: Constructor.
[INFO] RUNTIME(173034,python):2025-09-01-22:48:03.061.272 [npu_driver.cc:6306] 173260 GetDeviceStatus: GetDeviceStatus status=1.
[INFO] RUNTIME(173034,python):2025-09-01-22:48:03.061.330 [stars_engine.cc:1910] 173260 HBMRasProducer: read fault event not support or drv no resouces
[INFO] ATRACE(173034,python):2025-09-01-22:48:03.064.739 [tracer_mgr_operate.c:66](tid:173034) create object RUNTIME_ATRACE_DEV0_TS0 success, exitSave(false), noLock(false).
[INFO] TDT(173034,python):2025-09-01-22:48:03.064.774 [client_manager.cpp:196][SetProfilingCallback][tid:173034] [TsdClient] set profiling callback success
[ERROR] FE(173034,python):2025-09-01-22:48:03.065.444 [platform_info.cpp:969]173034 RealPath:"path /usr/local/Ascend/ascend-toolkit/8.1.RC1/aarch64-linux/lib64/../data/platform_config is not exist."
[ERROR] FE(173034,python):2025-09-01-22:48:03.065.457 [platform_info.cpp:1222]173034 InitRuntimePlatformInfos:"File path[] is not valid."
[ERROR] ASCENDCL(173034,python):2025-09-01-22:48:03.065.474 [device.cpp:70]173034 UpdatePlatformInfoWithDevice: [Init][PlatformInfo]init runtime platform info failed, SocVersion = Ascend910B4
[INFO] GE(173034,python):2025-09-01-22:48:03.067.293 [ge_api.cc:207][status:INIT]173034 GEInitializeImpl:GEInitialize start
[INFO] GE(173034,python):2025-09-01-22:48:03.067.316 [ge_api.cc:221]173034 GEInitializeImpl:[GEPERFTRACE] The time cost of GEInitialize::GEAPICheckSupportedGlobalOptions is [5] micro seconds.
[INFO] GE(173034,python):2025-09-01-22:48:03.076.428 [ge_api.cc:236]173034 GEInitializeImpl:[GEPERFTRACE] The time cost of GEInitialize::CheckOptionsValid is [36] micro seconds.
[INFO] GE(173034,python):2025-09-01-22:48:03.077.118 [ge_api.cc:251]173034 GEInitializeImpl:[GEPERFTRACE] The time cost of GEInitialize::ManagerInitialize is [1] micro seconds.
[EVENT] PROFILING(173034,python):2025-09-01-22:48:03.079.570 [msprof_callback_impl.cpp:110] >>> (tid:173034) MsprofCtrlCallback called, type: 255
[EVENT] PROFILING(173034,python):2025-09-01-22:48:03.079.588 [ai_drv_dev_api.cpp:333] >>> (tid:173034) Succeeded to DrvGetApiVersion version: 0x72317
[INFO] GE(173034,python):2025-09-01-22:48:03.079.634 [ge_api.cc:270]173034 GEInitializeImpl:[GEPERFTRACE] The time cost of GEInitialize::InitProfiling is [79] micro seconds.
[INFO] GE(173034,python):2025-09-01-22:48:03.080.289 [ge_api.cc:280]173034 GEInitializeImpl:[GEPERFTRACE] The time cost of GEInitialize::InitPreparation is [476] micro seconds.
[INFO] GE(173034,python):2025-09-01-22:48:03.080.305 [gelib.cc:113]173034 Initialize:[GEPERFTRACE] GE Init Start
[ERROR] FE(173034,python):2025-09-01-22:48:03.080.433 [platform_info.cpp:969]173034 RealPath:"path /usr/local/Ascend/ascend-toolkit/8.1.RC1/aarch64-linux/lib64/../data/platform_config is not exist."
[ERROR] FE(173034,python):2025-09-01-22:48:03.080.442 [platform_info.cpp:1002]173034 InitializePlatformInfo:"File path[] is not valid."
[ERROR] FE(173034,python):2025-09-01-22:48:03.080.560 [platform_info.cpp:969]173034 RealPath:"path /usr/local/Ascend/ascend-toolkit/8.1.RC1/aarch64-linux/lib64/../data/platform_config is not exist."
[ERROR] FE(173034,python):2025-09-01-22:48:03.080.570 [platform_info.cpp:1002]173034 InitializePlatformInfo:"File path[] is not valid."
[ERROR] GE(173034,python):2025-09-01-22:48:03.080.584 [platform_info_util.cc:38]173034 GetJitCompileDefaultValue: ErrorNo: 4294967295(failed) [INIT][SYS]Initialize platform info failed.
[INFO] GE(173034,python):2025-09-01-22:48:03.080.599 [gelib.cc:179]173034 Initialize:[GEPERFTRACE] The time cost of GELib::InitializePrepare is [286] micro seconds.
[INFO] GE(173034,python):2025-09-01-22:48:03.080.780 [gelib.cc:337]173034 SystemInitialize:Infer init GELib success, device id :-1 
[INFO] GE(173034,python):2025-09-01-22:48:03.080.790 [gelib.cc:209]173034 InnerInitialize:[GEPERFTRACE] The time cost of InnerInitialize::SystemInitialize is [13] micro seconds.
[ERROR] FE(173034,python):2025-09-01-22:48:03.080.912 [platform_info.cpp:969]173034 RealPath:"path /usr/local/Ascend/ascend-toolkit/8.1.RC1/aarch64-linux/lib64/../data/platform_config is not exist."
[ERROR] FE(173034,python):2025-09-01-22:48:03.080.921 [platform_info.cpp:1002]173034 InitializePlatformInfo:"File path[] is not valid."
[ERROR] GE(173034,python):2025-09-01-22:48:03.080.931 [model_helper.cc:1558]173034 GetPlatformInfo: ErrorNo: 4294967295(failed) [INIT][SYS]Initialize platform info failed.
[ERROR] GE(173034,python):2025-09-01-22:48:03.080.940 [model_helper.cc:1466]173034 GetHardwareInfo: ErrorNo: 4294967295(failed) [INIT][SYS]Get platform info failed, device id: 0, soc_version: Ascend910B4.
[ERROR] GE(173034,python):2025-09-01-22:48:03.080.950 [gelib.cc:217]173034 InnerInitialize: ErrorNo: 4294967295(failed) [INIT][SYS][Get][Hardware]GeLib initial: Get hardware info failed.
[ERROR] GE(173034,python):2025-09-01-22:48:03.080.959 [gelib.cc:183]173034 Initialize: ErrorNo: 4294967295(failed) [INIT][SYS][Init][GeLib]GeLib initial failed.
[INFO] GE(173034,python):2025-09-01-22:48:03.081.009 [ge_api.cc:286]173034 GEInitializeImpl:[GEPERFTRACE] The time cost of GEInitialize::GELibInitialize is [705] micro seconds.
[ERROR] GE(173034,python):2025-09-01-22:48:03.081.019 [ge_api.cc:288]173034 GEInitializeImpl: ErrorNo: 1343229953(GEInitialize Failed.) [INIT][SYS][Init][GELib]Failed, error code = 4294967295
[ERROR] GE(173034,python):2025-09-01-22:48:03.081.029 [ge_api.cc:370]173034 GEInitialize: ErrorNo: 4294967295(failed) [INIT][SYS][GEInit] GEInitialize failed, error code:4294967295.
[ERROR] ASCENDCL(173034,python):2025-09-01-22:48:03.081.054 [local_compiler.cpp:76]173034 Init: [INIT][SYS][Initialize][Ge]GEInitialize failed. ge result = 4294967295
[ERROR] ASCENDCL(173034,python):2025-09-01-22:48:03.081.079 [op_compile_service.cpp:73]173034 SetCompileStrategy: [INIT][SYS][Init][Compiler]Init compiler failed
[ERROR] ASCENDCL(173034,python):2025-09-01-22:48:03.081.101 [op_compile_processor.cpp:67]173034 Init: [INIT][SYS][Set][Options]OpCompileProcessor init failed!
[ERROR] ASCENDCL(173034,python):2025-09-01-22:48:03.081.121 [op_compile_processor.cpp:111]173034 SetCompileOpt: [INIT][SYS][Init][Env]init env failed!
[INFO] RUNTIME(173034,python):2025-09-01-22:48:03.081.176 [api_impl.cc:7464] 173034 PeekLastErr: level=0 err=0.
Traceback (most recent call last):
  File "/home/ma-user/work/pangu_rl_code/code/init.py", line 4, in <module>
    x = torch.randn(2, 2).npu()
  File "/usr/local/python3.10.17/lib/python3.10/site-packages/torch/utils/backend_registration.py", line 147, in wrap_tensor_to
    device_idx = _normalization_device(custom_backend_name, device)
  File "/usr/local/python3.10.17/lib/python3.10/site-packages/torch/utils/backend_registration.py", line 97, in _normalization_device
    return _get_current_device_index()
  File "/usr/local/python3.10.17/lib/python3.10/site-packages/torch/utils/backend_registration.py", line 91, in _get_current_device_index
    return getattr(getattr(torch, custom_backend_name), _get_device_index)()
  File "/usr/local/python3.10.17/lib/python3.10/site-packages/torch_npu/npu/utils.py", line 84, in current_device
    torch_npu.npu._lazy_init()
  File "/usr/local/python3.10.17/lib/python3.10/site-packages/torch_npu/npu/__init__.py", line 242, in _lazy_init
    torch_npu._C._npu_init()
RuntimeError: SetPrecisionMode:build/CMakeFiles/torch_npu.dir/compiler_depend.ts:156 NPU function error: at_npu::native::AclSetCompileopt(aclCompileOpt::ACL_PRECISION_MODE, precision_mode), error code is 500001
[ERROR] 2025-09-01-22:48:03 (PID:173034, Device:0, RankID:-1) ERR00100 PTA call acl api failed
[Error]: The internal ACL of the system is incorrect.
        Rectify the fault based on the error information in the ascend log.
EH9999: Inner Error!
EH9999: [PID: 173034] 2025-09-01-22:48:03.065.496 [Init][PlatformInfo]init runtime platform info failed, SocVersion = Ascend910B4[FUNC:ReportInnerError][FILE:log_inner.cpp][LINE:145]
        TraceBack (most recent call last):
       GELib::InnerInitialize failed.[FUNC:Initialize][FILE:gelib.cc][LINE:184]
       GEInitialize failed.[FUNC:GEInitialize][FILE:ge_api.cc][LINE:371]
       [Initialize][Ge]GEInitialize failed. ge result = 4294967295[FUNC:ReportCallError][FILE:log_inner.cpp][LINE:161]
       [Init][Compiler]Init compiler failed[FUNC:ReportInnerError][FILE:log_inner.cpp][LINE:145]
       [Set][Options]OpCompileProcessor init failed![FUNC:ReportInnerError][FILE:log_inner.cpp][LINE:145]

[INFO] IDEDD(173034,python):2025-09-01-22:48:04.395.178 [adx_server_manager.cpp:236][tid:173257] server manager stop
[INFO] RUNTIME(173034,python):2025-09-01-22:48:04.925.326 [runtime.cc:1540] 173034 ~Runtime: deconstruct runtime
[INFO] RUNTIME(173034,python):2025-09-01-22:48:04.925.369 [runtime.cc:1547] 173034 ~Runtime: wait monitor success, use=0.
[INFO] ATRACE(173034,python):2025-09-01-22:48:04.927.746 [tracer_mgr_operate.c:248](tid:173034) destroy object RUNTIME_ATRACE_DEV0_TS0, exitSave(false).
[INFO] RUNTIME(173034,python):2025-09-01-22:48:04.936.312 [stars_engine.cc:58] 173034 ~StarsEngine: Destructor.
[INFO] RUNTIME(173034,python):2025-09-01-22:48:04.936.328 [engine.cc:84] 173034 ~Engine: Destructor.
[INFO] DRV(173034,python):2025-09-01-22:48:04.936.912 [devmm_svm.c:2036][ascend][curpid:173034,173034][drv][devmm][drvMemDeviceClose]DrvMemDeviceClose. (devid=0)
[INFO] DRV(173034,python):2025-09-01-22:48:04.936.930 [svm_mem_statistics.c:136][ascend][curpid:173034,173034][drv][devmm][_svm_mem_stats_show]HOST_MEM Mem stats (Bytes). (module_name=RUNTIME; module_id=7; current_alloced_size=0; alloced_peak_size=11341824; alloc_cnt=3; free_cnt=3)
[INFO] DRV(173034,python):2025-09-01-22:48:04.936.941 [svm_mem_statistics.c:146][ascend][curpid:173034,173034][drv][devmm][_svm_mem_stats_show]HOST_MEM Cached_size:12582912Bytes
[INFO] DRV(173034,python):2025-09-01-22:48:04.936.949 [svm_mem_statistics.c:136][ascend][curpid:173034,173034][drv][devmm][_svm_mem_stats_show]MEM_DEV_SMALL_HBM dev0 Mem stats (Bytes). (module_name=RUNTIME; module_id=7; current_alloced_size=0; alloced_peak_size=77824; alloc_cnt=4; free_cnt=4)
[INFO] DRV(173034,python):2025-09-01-22:48:04.936.958 [svm_mem_statistics.c:146][ascend][curpid:173034,173034][drv][devmm][_svm_mem_stats_show]MEM_DEV_SMALL_HBM dev0 Cached_size:2097152Bytes
[INFO] DRV(173034,python):2025-09-01-22:48:04.936.965 [svm_mem_statistics.c:136][ascend][curpid:173034,173034][drv][devmm][_svm_mem_stats_show]MEM_DEV_SMALL_DDR dev0 Mem stats (Bytes). (module_name=RUNTIME; module_id=7; current_alloced_size=0; alloced_peak_size=16384; alloc_cnt=1; free_cnt=1)
[INFO] DRV(173034,python):2025-09-01-22:48:04.936.974 [svm_mem_statistics.c:146][ascend][curpid:173034,173034][drv][devmm][_svm_mem_stats_show]MEM_DEV_SMALL_DDR dev0 Cached_size:2097152Bytes
[INFO] DRV(173034,python):2025-09-01-22:48:04.936.984 [svm_mem_statistics.c:136][ascend][curpid:173034,173034][drv][devmm][_svm_mem_stats_show]MEM_DEV_HUGE_HBM dev0 Mem stats (Bytes). (module_name=UNKNOWN; module_id=0; current_alloced_size=0; alloced_peak_size=2097152; alloc_cnt=1; free_cnt=1)
[INFO] DRV(173034,python):2025-09-01-22:48:04.936.991 [svm_mem_statistics.c:136][ascend][curpid:173034,173034][drv][devmm][_svm_mem_stats_show]MEM_DEV_HUGE_HBM dev0 Mem stats (Bytes). (module_name=RUNTIME; module_id=7; current_alloced_size=0; alloced_peak_size=20971520; alloc_cnt=6; free_cnt=6)
[INFO] DRV(173034,python):2025-09-01-22:48:04.936.999 [svm_mem_statistics.c:146][ascend][curpid:173034,173034][drv][devmm][_svm_mem_stats_show]MEM_DEV_HUGE_HBM dev0 Cached_size:23068672Bytes
[INFO] DRV(173034,python):2025-09-01-22:48:04.937.006 [svm_mem_statistics.c:136][ascend][curpid:173034,173034][drv][devmm][_svm_mem_stats_show]MEM_DEV_HUGE_DDR dev0 Mem stats (Bytes). (module_name=RUNTIME; module_id=7; current_alloced_size=0; alloced_peak_size=6291456; alloc_cnt=2; free_cnt=2)
[INFO] DRV(173034,python):2025-09-01-22:48:04.937.014 [svm_mem_statistics.c:146][ascend][curpid:173034,173034][drv][devmm][_svm_mem_stats_show]MEM_DEV_HUGE_DDR dev0 Cached_size:6291456Bytes
[INFO] DRV(173034,python):2025-09-01-22:48:04.937.020 [svm_mem_statistics.c:136][ascend][curpid:173034,173034][drv][devmm][_svm_mem_stats_show]MEM_DEV_HUGE_P2P_HBM dev0 Mem stats (Bytes). (module_name=RUNTIME; module_id=7; current_alloced_size=0; alloced_peak_size=4194304; alloc_cnt=1; free_cnt=1)
[INFO] DRV(173034,python):2025-09-01-22:48:04.937.028 [svm_mem_statistics.c:146][ascend][curpid:173034,173034][drv][devmm][_svm_mem_stats_show]MEM_DEV_HUGE_P2P_HBM dev0 Cached_size:4194304Bytes
[INFO] RUNTIME(173034,python):2025-09-01-22:48:05.092.428 [task_fail_callback_data_manager.cc:63] 173034 ~TaskFailCallBackManager: Destructor.
[INFO] ATRACE(173034,python):2025-09-01-22:48:05.301.417 [stacktrace_signal.c:188](tid:173034) unregister all signo sigaction, can not capture signal.
[INFO] ATRACE(173034,python):2025-09-01-22:48:06.361.300 [atrace_client_core.c:127](tid:173259) atrace receive thread exited, devId=0.

二、软件版本:
-- CANN 版本 (CANN 8.1.RC1):  
--Pytorch 版本(torch: 2.5.1, torch_npu: 2.5.1.post1.dev20250619):
--Python 版本 (python 3.10.17):
--操作系统版本 (4.19.90-vhulk2211.3.0.h1543.eulerosv2r10.aarch64):

三、测试步骤：
执行下面程序会报错
import torch
import torch_npu

x = torch.randn(2, 2).npu()
y = torch.randn(2, 2).npu()
z = x.mm(y)

print(z)

Ascend/pytorch
暂停

内容风险标识

评论 (4)

Ascend/pytorch暂停 .gitee-modal { width: 500px !important; }

内容风险标识