下载
中文
注册

调试PyTorch接口调用的算子

展示如何使用msDebug工具来上板调试一个PyTorch接口调用的add算子,该add算子可实现两个向量相加并输出结果的功能。

前提条件

  • 单击Link获取样例工程,为进行算子调试做准备。
    下载代码样例时,需执行以下命令指定分支版本。
    git clone https://gitee.com/ascend/samples.git -b 8.0.RC3
  • 已参考Ascend Extension for PyTorch 配置与安装,完成PyTorch框架和torch_npu插件的安装。
  • 参考使用前准备完成相关环境变量配置。

操作步骤

  1. 参考创建算子工程完成算子工程的创建。
  2. 参考算子开发完成算子开发。
  3. 参考算子编译部署,完成算子的编译部署。

    编辑op_kernel/CMakeLists.txt文件,增加编译选项-O0 -g。

    add_ops_compile_options(ALL OPTIONS -O0 -g )
  4. 进入PyTorch接入工程,使用PyTorch调用方式调用AddCustom算子工程,并按照指导完成编译。
    下载代码样例时,需执行以下命令指定分支版本。
    git clone https://gitee.com/ascend/samples.git -b 8.0.RC3
    1
    2
    3
    4
    PytorchInvocation
    ├── op_plugin_patch         
    ├── run_op_plugin.sh      //  5.执行样例时,需要使用
    └── test_ops_custom.py    //  步骤6启动工具时,需要使用
    
  5. 执行样例,样例执行过程中会自动生成测试数据,然后运行pytorch样例,最后检验运行结果。
     1
     2
     3
     4
     5
     6
     7
     8
     9
    10
    11
    12
    13
    14
    15
    16
    17
    18
    19
    20
    21
    22
    23
    24
    25
    26
    bash run_op_plugin.sh
    -- CMAKE_CCE_COMPILER: /usr/local/Ascend/ascend-toolkit/latest/toolkit/tools/ccec_compiler/bin/ccec
    -- CMAKE_CURRENT_LIST_DIR: ${INSTALL_DIR}/AddKernelInvocation/cmake/Modules
    -- ASCEND_PRODUCT_TYPE:
      Ascendxxxyy
    -- ASCEND_CORE_TYPE:
      VectorCore
    -- ASCEND_INSTALL_PATH:
      /usr/local/Ascend/ascend-toolkit/latest
    -- The CXX compiler identification is GNU 10.3.1
    -- Detecting CXX compiler ABI info
    -- Detecting CXX compiler ABI info - done
    -- Check for working CXX compiler: /usr/bin/c++ - skipped
    -- Detecting CXX compile features
    -- Detecting CXX compile features - done
    -- Configuring done
    -- Generating done
    -- Build files have been written to: ${INSTALL_DIR}/AddKernelInvocation/build
    Scanning dependencies of target add_npu
    ...
    [100%] Built target add_npu
    ${INSTALL_DIR}/AddKernelInvocation
    ...
    INFO: Ascend C Add Custom SUCCESS
    ...
    INFO: Ascend C Add Custom  in torch.compile graph SUCCESS
    
  6. 手动导入算子调试信息。
    • ${INSTALL_DIR}请替换为CANN软件安装后文件存储路径。若安装的Ascend-cann-toolkit软件包,以root安装举例,则安装后文件存储路径为:/usr/local/Ascend/ascend-toolkit/latest。
    • 在安装昇腾AI处理器的服务器执行npu-smi info命令进行查询,获取Chip Name信息。实际配置值为AscendChip Name,例如Chip Name取值为xxxyy,实际配置值为Ascendxxxyy
    (msdebug) export 
    LAUNCH_KERNEL_PATH=${INSTALL_DIR}/opp/vendors/customize/op_impl/ai_core/tbe/kernel/SOC_VERSION/add_custom/AddCustom_1e04ee05ab491cc5ae9c3d5c9ee8950b.o
  7. 启动msDebug工具拉起Python程序,进入调试界面。
    1
    2
    3
    4
    5
    msdebug python3 test_ops_custom.py
    (msdebug) target create "python3"
    Current executable set to '/home/mindstudio/miniconda3/envs/py37/bin/python3' (aarch64).
    (msdebug) settings set -- target.run-args  "test_ops_custom.py"
    (msdebug)
    
  8. 设置断点。
    根据指定源码文件与对应行号,在核函数中设置NPU断点。
    1
    2
    (msdebug) b add_custom.cpp:60
    Breakpoint 1: where = AddCustom_1e04ee05ab491cc5ae9c3d5c9ee8950b.o`::AddCustom_1e04ee05ab491cc5ae9c3d5c9ee8950b_1(uint8_t *, uint8_t *, uint8_t *, uint8_t *, uint8_t *) + 9912 [inlined] KernelAdd::Compute(int) + 3400 at add_custom.cpp:60:9, address = 0x00000000000026b8
    
  9. 运行程序,等待直到命中断点。
     1
     2
     3
     4
     5
     6
     7
     8
     9
    10
    11
    12
    13
    14
    15
    16
    17
    (msdebug) r
    Process 197189 launched: '/home/miniconda3/envs/py38/bin/python3' (aarch64)
    Process 197189 stopped and restarted: thread 1 received signal: SIGCHLD
    ...
    [Launch of Kernel anonymous on Device 0]
    Process 197189 stopped
    [Switching to focus on Kernel anonymous, CoreId 8, Type aiv]
    * thread #1, name = 'python3', stop reason = breakpoint 2.1
        frame #0: 0x00000000000026b8 AddCustom_1e04ee05ab491cc5ae9c3d5c9ee8950b.o`::AddCustom_1e04ee05ab491cc5ae9c3d5c9ee8950b_1(uint8_t *, uint8_t *, uint8_t *, uint8_t *, uint8_t *) [inlined] KernelAdd::Compute(this=0x000000000020efb8, progress=1) at add_custom.cpp:60:9
       57              LocalTensor<DTYPE_Y> yLocal = inQueueY.DeQue<DTYPE_Y>();
       58              LocalTensor<DTYPE_Z> zLocal = outQueueZ.AllocTensor<DTYPE_Z>();
       59              Add(zLocal, xLocal, yLocal, this->tileLength);
    -> 60              outQueueZ.EnQue<DTYPE_Z>(zLocal);
       61              inQueueX.FreeTensor(xLocal);
       62              inQueueY.FreeTensor(yLocal);
       63          }
    (msdebug)
    

    其他调试操作可参考导入调试信息内存与变量打印调试信息展示核切换等,与其操作一致。

  10. 删除断点,具体操作请参见删除断点
  11. 调试完以后,执行q命令并输入Y或y结束调试。
    (msdebug) q
    Quitting LLDB will kill one or more processes. Do you really want to proceed: [Y/n] y