rk1808节点启动失败

bobby_jiang 发表于 2020-6-4 16:37
pc上把程序跑起来啊，跑起来就有log输出了

节点起不来，报主题帖里面的错误

raul 发表于 2020-6-4 17:16
麻烦提供下你的rknn模型，我们debug看下。

个人觉得和模型无关，先给你看一下运行官方demo报错情况吧，使用的demo是http://repo.rock-chips.com/rk1808/mobilenet_v1.tar.gz
--> config model
done
--> Loading model
done
--> Building model
W The target_platform is not set in config, using default target platform rk1808.
W:tensorflow:From /home/johnny/.local/lib/python3.5/site-packages/tensorflow/python/ops/control_flow_ops.py:3632: colocate_with (from tensorflow.python.framework.ops) is deprecated and will be removed in a future version.
Instructions for updating:
Colocations handled automatically by placer.
W:tensorflow:From /home/johnny/.local/lib/python3.5/site-packages/tensorflow/python/ops/control_flow_ops.py:1941: py_func (from tensorflow.python.ops.script_ops) is deprecated and will be removed in a future version.
Instructions for updating:
tf.py_func is deprecated in TF V2. Instead, use
    tf.py_function, which takes a python function which manipulates tf eager
    tensors instead of numpy arrays. It's easy to convert a tf eager tensor to
    an ndarray (just call tensor.numpy()) but having access to eager tensors
    means `tf.py_function`s can use accelerators such as GPUs as well as
    being differentiable using a gradient tape.
   
done
--> Export RKNN model
done
--> Init runtime environment
E Execute ntp_transfer_proxy devices failed, ret code: 1. /home/johnny/.local/lib/python3.5/site-packages/rknn/3rdparty/platform-tools/ntp/linux-x86_64/npu_transfer_proxy: /usr/lib/x86_64-linux-gnu/libstdc++.so.6: version `GLIBCXX_3.4.22' not found (required by /home/johnny/.local/lib/python3.5/site-packages/rknn/3rdparty/platform-tools/ntp/linux-x86_64/npu_transfer_proxy)

E Catch exception when init runtime!
E Traceback (most recent call last):
E   File "rknn/api/rknn_base.py", line 1042, in rknn.api.rknn_base.RKNNBase.init_runtime
E   File "rknn/api/rknn_runtime.py", line 168, in rknn.api.rknn_runtime.RKNNRuntime.__init__
E   File "rknn/api/rknn_platform_utils.py", line 272, in rknn.api.rknn_platform_utils.start_ntp_or_adb
E   File "rknn/api/rknn_platform_utils.py", line 136, in rknn.api.rknn_platform_utils.weather_device_use_ntb
E Exception: Init runtime environment failed!
E Current device id is: None
E Devices connected:
E Execute ntp_transfer_proxy devices failed, ret code: 1. /home/johnny/.local/lib/python3.5/site-packages/rknn/3rdparty/platform-tools/ntp/linux-x86_64/npu_transfer_proxy: /usr/lib/x86_64-linux-gnu/libstdc++.so.6: version `GLIBCXX_3.4.22' not found (required by /home/johnny/.local/lib/python3.5/site-packages/rknn/3rdparty/platform-tools/ntp/linux-x86_64/npu_transfer_proxy)

Traceback (most recent call last):
  File "test.py", line 66, in <module>
    ret = rknn.init_runtime(target='rk1808')
  File "/home/johnny/.local/lib/python3.5/site-packages/rknn/api/rknn.py", line 378, in init_runtime
    adb_devices, ntb_devices = self.rknn_base.list_devices()
  File "rknn/api/rknn_base.py", line 1006, in rknn.api.rknn_base.RKNNBase.list_devices
  File "rknn/api/rknn_platform_utils.py", line 328, in rknn.api.rknn_platform_utils.get_ntb_devices
RuntimeError

raul 发表于 2020-6-5 08:59
这个错和前面的错是不一样的呀。这个错是因为gcc版本不够的原因导致的。要把gcc升级到5.5.0 ...

那我们先升级一下gcc，你把邮箱给我一下吧，模型发你邮箱

bobby_jiang 发表于 2020-6-5 10:50
bobby.jiang@rock-chip.com

已发，请注意查收

raul 发表于 2020-6-5 08:59
这个错和前面的错是不一样的呀。这个错是因为gcc版本不够的原因导致的。要把gcc升级到5.5.0 ...

我们不方便更新到gcc5.5.0，会干扰到其他进程。你们的官方demo是python脚本，有没有c++的提供一下？

bobby_jiang 发表于 2020-6-5 11:22
wiki上本来就有C++的，你认真看一下

我们已经解决问题了，根本原因还是出在npu_transfer_proxy上。

在http://repo.rock-chips.com/rk180 ... proxy/linux-x86_64/下载npu_transfer_proxy。
首先启动可执行文件npu_transfer_proxy，接着启动节点，即可成功。
我们又进行了优化，利用systemd实现npu_transfer_proxy开机自启动，避免每次手动启动。

最近麻烦你们了，十分感谢。也希望我们的解决方案能帮到遇到同样问题的人。