error while loading shared libraries: libmkl_intel_lp64.so

[rozentill]

Hi,

Thank you for releasing the codes. I met problem with marching cube at last of the demo, the error is:

./isosurface/computeMarchingCubes: error while loading shared libraries: libmkl_intel_lp64.so: cannot open shared object file: No such file or directory

However I've tried many methods to reinstall mkl, add this file to lib path and even move it to the same folder with computeMarchingCubes, neither of these methods work.

Any idea on solving this would be appreciated.

Thanks!

[walsvid]

I think you can modify LIB_PATH to correct mkl path. The LIB_PATH provided by the author is set to author's path, which is not correct. However, the author has provided a link library available in this github repo, which is in the current folder path.

Maybe you can try this: ./isosurface/tbb/tbb2018_20180822oss/lib/intel64/gcc4.7

[qiaogh97]

I met same problem. My Linux server has no surface to display, there is no /opt/intel/lib/intel64 or /opt/intel/mkl/lib/intel64/ path. Any other code to convert numpy/result.dist to result.obj?

[qiaogh97]

I solve this problem.
Step1 : Save result.npy in demo.py fuction create_obj()
https://github.com/laughtervv/DISN/blob/ddb432b8b92c76eaa59c2d2a014770966378c824/demo/demo.py#L356
np.save('result.npy', pred_sdf_val)
Step 2: Install PyMcubes
pip install --upgrade PyMCubes
https://github.com/pmneila/PyMCubes
Step 3: Convert result.npy to result.obj using PyMcubes
You can refer to my code.
https://github.com/qiao1025566574/DISN/blob/master/demo/numpyToOBJ.py

[qiaogh97]

This is effect picture of result.obj

[bharadwajdhornala]

Hi @qiao1025566574 @walsvid @Xharlie I am trying this code on CPU version. I have made the GPU to CPU changes. I am stuck at changing LIB_PATH. Can you please specify what each means in the given statement.
nohup python -u demo/demo.py --cam_est --log_dir checkpoint/SDF_DISN --cam_log_dir cam_est/checkpoint/cam_DISN --img_feat_twostream --sdf_res 256 &> log/create_sdf.log &
Thanks!

[bharadwajdhornala]

Hi @walsvid. Thanks for your fast reply. export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:./isosurface/:/home/DISN/isosurface/tbb/tbb2018_20180822oss/lib/intel64/gcc4.7:/opt/intel/lib/intel64:/opt/intel/mkl/lib/intel64:/usr/local/lib64:/usr/local/lib:/usr/local/cuda/lib64 What does :/opt/intel/lib/intel64:/opt/intel/mkl/lib/intel64:/usr/local/lib64:/usr/local/lib:/usr/local/cuda/lib64 are for?

[walsvid]

Hi @bharadwajdhornala , it just means check which common paths contain the required so files.

[shravanc]

I think you can modify LIB_PATH to correct mkl path. The LIB_PATH provided by the author is set to author's path, which is not correct. However, the author has provided a link library available in this github repo, which is in the current folder path.

Maybe you can try this: ./isosurface/tbb/tbb2018_20180822oss/lib/intel64/gcc4.7

I had to give an extra path '/opt/nvidia/nsight-systems/2019.5.2/host-linux-x64/' and it worked fine.

This is my complete LIB_PATH content:
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:./isosurface/:/home/shravan/noplacelike/original/DISN-1/isosurface/tbb/tbb2018_20180822oss/lib/intel64/gcc4.7:/opt/intel/lib/intel64:/opt/intel/mkl/lib/intel64:/usr
/local/lib64:/usr/local/lib:/usr/local/cuda/lib64:/opt/nvidia/nsight-systems/2019.5.2/host-linux-x64/

[georged4s]

Thanks to @qiao1025566574 for providing an alternative to the original method. I have made slight changes based on his suggestion and seems to fix the issue. Everything works nicely afterwards. In the demo.py script, I have changed the create_obj function as follow:

def create_obj(pred_sdf_val, sdf_params, dir, i):
    import mcubes
    new_dim = int(pred_sdf_val.shape[0] ** (1.0 / 3)) + 1
    u = pred_sdf_val.reshape(new_dim, new_dim, new_dim)
    vertices, triangles = mcubes.marching_cubes(u, 0)
    mcubes.export_obj(vertices, triangles, "demo/result.obj")

[zhangjunhust]

can anyone help me? I have changed the LIB PATH to:
./zj/DISN/isosurface/tbb/tbb2018_20180822oss/lib/intel64/gcc4.7;
It still got error when I ran the demo.py :
./isosurface/computeMarchingCubes: error while loading shared libraries: libtbb.so.2: cannot open shared object file: No such file or directory

Why?