2

I am trying to install a open-source software "openpose" for which I needed to install cuda, cudnn and nvidia drivers. Output of nvidia-smi is :

+-----------------------------------------------------------------------------+
| NVIDIA-SMI 440.59       Driver Version: 440.59       CUDA Version: 10.2     |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
|===============================+======================+======================|
|   0  GeForce 940MX       Off  | 00000000:01:00.0 Off |                  N/A |
| N/A   47C    P8    N/A /  N/A |    107MiB /  2004MiB |      7%      Default |
+-------------------------------+----------------------+----------------------+

+-----------------------------------------------------------------------------+
| Processes:                                                       GPU Memory |
|  GPU       PID   Type   Process name                             Usage      |
|=============================================================================|
|    0      1513      G   /usr/lib/xorg/Xorg                            63MiB |
|    0      1698      G   /usr/bin/gnome-shell                          41MiB |
+-----------------------------------------------------------------------------+

And output of cat /usr/include/cudnn.h | grep CUDNN_MAJOR -A 2 gives:

#define CUDNN_MAJOR 7
#define CUDNN_MINOR 6
#define CUDNN_PATCHLEVEL 5

After successfull installations of all the above softwares and libraries, I finally ran openpose with:

./build/examples/openpose/openpose.bin --video examples/media/video.avi

But the output was:

Starting OpenPose demo...
Configuring OpenPose...
Starting thread(s)...
Auto-detecting all available GPUs... Detected 1 GPU(s), using 1 of them starting at GPU 0.
F0214 01:02:35.327615  3433 cudnn_conv_layer.cpp:53] Check failed: status == CUDNN_STATUS_SUCCESS (1 vs. 0)  CUDNN_STATUS_NOT_INITIALIZED
*** Check failure stack trace: ***
    @     0x7fabb8f390cd  google::LogMessage::Fail()
    @     0x7fabb8f3af33  google::LogMessage::SendToLog()
    @     0x7fabb8f38c28  google::LogMessage::Flush()
    @     0x7fabb8f3b999  google::LogMessageFatal::~LogMessageFatal()
    @     0x7fabb89459d3  caffe::CuDNNConvolutionLayer<>::LayerSetUp()
    @     0x7fabb8a42308  caffe::Net<>::Init()
    @     0x7fabb8a441e0  caffe::Net<>::Net()
    @     0x7fabbaa2ccaa  op::NetCaffe::initializationOnThread()
    @     0x7fabbaa500a1  op::addCaffeNetOnThread()
    @     0x7fabbaa51518  op::PoseExtractorCaffe::netInitializationOnThread()
    @     0x7fabbaa57163  op::PoseExtractorNet::initializationOnThread()
    @     0x7fabbaa4be61  op::PoseExtractor::initializationOnThread()
    @     0x7fabbaa46a51  op::WPoseExtractor<>::initializationOnThread()
    @     0x7fabbaa8aff1  op::Worker<>::initializationOnThreadNoException()
    @     0x7fabbaa8b120  op::SubThread<>::initializationOnThread()
    @     0x7fabbaa8d2d8  op::Thread<>::initializationOnThread()
    @     0x7fabbaa8d4a7  op::Thread<>::threadFunction()
    @     0x7fabba32566f  (unknown)
    @     0x7fabb9a476db  start_thread
    @     0x7fabb9d8088f  clone
Aborted

I have gone through a lot of online discussions but could not figure out how to resolve this.

talonmies
  • 70,661
  • 34
  • 192
  • 269
Sunny
  • 51
  • 1
  • 7

1 Answers1

6

I have been having the same problem with CUDNN.

Although not ideal, I have been running without CUDNN. In cmake-gui uncheck USE_CUDNN and then compile. When running openpose I have also had to reduce -net_resolution.

For example: ./build/examples/openpose/openpose.bin -net_resolution 256x192 The greater the resolution the slower the FPS though.

Max
  • 61
  • 1
  • 1
  • i guess the last resort is to launch tensorflow from inside the docker container – ivan866 Aug 16 '20 at 18:10
  • openpose uses pytorch/caffe, not tensorflow ;). One way is to downgrade driver + cuda + cudnn combo to match openpose requirements. – yǝsʞǝla Feb 11 '21 at 02:27