- changed status to resolved
VdW interactions in Parallel (OpenMP)
Issue #10
resolved
The steric VdW interactions between elements do not seem to work correctly with any more than 1 processor. I believe I've narrowed the problematic function down to 'do_interaction' in the vdw_solver and it's children where the programs seems to register an overlap between elements when there clearly isn't one. Until this is addressed, I've commented out the pragma's surrounding the for loop, which enables everything to work correctly, albeit slowly
Comments (2)
-
reporter -
Ben himself found that this was caused because the function in tet_a_tet had global variables, and it was fixed long ago. I am closing this issue.
- Log in to comment
Fixed by making the volume interaction tools have local variables instead of global