Issue with DLIB - cudnnGetConvolutionForwardWorkspaceSize() - CUDNN_STATUS_NOT_SUPPORTED

Hi,
i am new with dlib and the face detection modules.
Need to get php dlib module working for which a requirement is to install DLib as shared library so ad to complie it manually with apparent successes via:

git clone https://github.com/davisking/dlib.git
cd dlib/dlib
mkdir build
cd build
cmake -DBUILD_SHARED_LIBS=ON …
make
sudo make install
sudo ldconfig

Then when run a php/python/cpp test via utility: https://github.com/matiasdelellis/pdlib-min-test-suite
Always getting the flowing error for all three methods:

Error while calling cudnnGetConvolutionForwardWorkspaceSize( context(), descriptor(data), (const cudnnFilterDescriptor_t)filter_handle, (const cudnnConvolutionDescriptor_t)conv_handle, descriptor(dest_desc), (cudnnConvolutionFwdAlgo_t)forward_algo, &forward_workspace_size_in_bytes) in file /home/nvjet/dlib/dlib/cuda/cudnn_dlibapi.cpp:1026. code: 9, reason: CUDNN_STATUS_NOT_SUPPORTED

Looking into https://docs.nvidia.com/deeplearning/cudnn/release-notes/rel_8.html#rel_8 found that this potentialy should have been resoved?

The functions cudnnGetConvolutionForwardAlgorithm_v7() and cudnnGetConvolutionForwardWorkspaceSize() may return CUDNN_STATUS_SUCCESS while the execution of the same convolution returns CUDNN_STATUS_NOT_SUPPORTED. Similar issues may also happen for convolutionBackwardData() and convolutionBackwardFilter(). This issue is present in cuDNN 7.2.2 library and later versions. This has been fixed in
cuDNN 8.0.0.

Current machine build details:
Using Jetpack 4.4.1 [L4T 32.4.4]
image

Hi,

CUDNN_STATUS_NOT_SUPPORTED indicates the requested usage is out of our support scope.
Please check the detail information of cudnnGetConvolutionForwardWorkspaceSize here:


Thanks.