[gmx-users] energy components zero in double precision?

[David Mobley]

All,

I'm having a perplexing problem(?) with a double-precision version of
GROMACS 3.3 I just compiled. I'm using it to reprocess a single frame
of a trr file using mdrun -rerun and a particular mdp file, topology,
and gro, and looking at the resulting energies. For some reason, my
coulomb-14 and LJ-14 energies are zero when I use the double precision
version, but if I do EXACTLY the same test using the single precision
version, coulomb-14 and LJ-14 energies are nonzero.

Does anyone have any idea what might be going on? Should I be
submitting a bugzilla, or is there something I should be
troubleshooting regarding my compilation? There didn't seem to be any
problems, and GROMACS runs without errors in both cases...

Thanks,
David