Failing TestReferenceMPIDForce and unexpected MD test run

[wwilla7]

Hi,

I am trying to build the mpid plugin in a fresh conda environment but I am getting some unexpected results and I am hoping to get some help.

I installed the MPID plugin in a conda environment with the following command:

conda create -n mpid-test openmm=7.5 cudatoolkit=11.2 swig=3 mdtraj -c conda-forge -y

Following all instructions on the documentation, I am getting the following output with cmake:

-- The C compiler identification is GNU 7.5.0
-- The CXX compiler identification is GNU 7.5.0
-- Check for working C compiler: /usr/bin/cc
-- Check for working C compiler: /usr/bin/cc -- works
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Detecting C compile features
-- Detecting C compile features - done
-- Check for working CXX compiler: /usr/bin/g++
-- Check for working CXX compiler: /usr/bin/g++ -- works
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Looking for pthread.h
-- Looking for pthread.h - found
-- Performing Test CMAKE_HAVE_LIBC_PTHREAD
-- Performing Test CMAKE_HAVE_LIBC_PTHREAD - Failed
-- Looking for pthread_create in pthreads
-- Looking for pthread_create in pthreads - not found
-- Looking for pthread_create in pthread
-- Looking for pthread_create in pthread - found
-- Found Threads: TRUE  
-- Configuring done
-- Generating done

I'm not very familiar with the pthread so I'm not sure if the above "Failed" and "not found" will be fatal errors.
4 out of 5 tests passed, with the TestReferenceMPIDForce failing. This confused me since all CUDA tests can pass. I ran the reference test separately and got the following output:

TestReferenceMPIDForce running test...
exception: Assertion failure at TestReferenceMPIDForce.cpp:934.  Expected -nan, found -nan
FAIL - ERROR.  Test failed.

Regardless of the failed test, I finished installing the plugin. make PythonInstall gave me the following warning message:

/home/wwilla/anaconda3/envs/mpid-test/include/swig/OpenMMSwigHeaders.i:2179: Warning 314: 'None' is a python keyword, renaming to '_None'
/home/wwilla/anaconda3/envs/mpid-test/include/swig/OpenMMSwigHeaders.i:505: Warning 453: Can't apply (OpenMM::Context &OUTPUT). No typemaps are defined.
/home/wwilla/anaconda3/envs/mpid-test/include/swig/OpenMMSwigHeaders.i:506: Warning 453: Can't apply (std::vector< double,std::allocator< double > > &OUTPUT). No typemaps are defined.
/home/wwilla/anaconda3/envs/mpid-test/include/swig/OpenMMSwigHeaders.i:510: Warning 453: Can't apply (OpenMM::Context &OUTPUT). No typemaps are defined.
/home/wwilla/anaconda3/envs/mpid-test/include/swig/OpenMMSwigHeaders.i:513: Warning 453: Can't apply (OpenMM::Context &OUTPUT). No typemaps are defined.
/home/wwilla/anaconda3/envs/mpid-test/include/swig/OpenMMSwigHeaders.i:553: Warning 453: Can't apply (std::vector< double,std::allocator< double > > &OUTPUT). No typemaps are defined.
/home/wwilla/anaconda3/envs/mpid-test/include/swig/OpenMMSwigHeaders.i:558: Warning 453: Can't apply (OpenMM::Context &OUTPUT). No typemaps are defined.
/home/wwilla/anaconda3/envs/mpid-test/include/swig/OpenMMSwigHeaders.i:581: Warning 453: Can't apply (Vec3 &OUTPUT). No typemaps are defined.
/home/wwilla/anaconda3/envs/mpid-test/include/swig/OpenMMSwigHeaders.i:582: Warning 453: Can't apply (Vec3 &OUTPUT). No typemaps are defined.
/home/wwilla/anaconda3/envs/mpid-test/include/swig/OpenMMSwigHeaders.i:583: Warning 453: Can't apply (Vec3 &OUTPUT). No typemaps are defined.
/home/wwilla/anaconda3/envs/mpid-test/include/swig/OpenMMSwigHeaders.i:641: Warning 453: Can't apply (std::vector< double,std::allocator< double > > &OUTPUT). No typemaps are defined.
...

I am not very familiar with swig either so I cannot interpret these warnings properly. The PythonInstall was built regardless of those warnings. So I went ahead and ran the run_ethane.py example. There was no explicit error in the MD run, but the system was blowing up. The initial density was 0.92 g/ml but ended up with 0.02 g/ml after 0.5ns of simulation. I am wondering what has caused these unexpected results.

I was able to build the plugin before and the ethane test run was correct in the old build. I must have missed something in the new build and I will need some help diagnosing the problem. The cmake version is 3.16.3. I can provide more log files as needed, thank you!