<style type="text/css">
        <!--
                @page { margin: 0.79in }
                PRE.cjk { font-family: "Droid Sans Fallback", monospace }
                P { margin-bottom: 0.08in }
        -->
        </style>

<pre class="western">Dear Gromacs developers:
I&#39;m trying to run a QM-MM single point energy calculation on a structure (a small molecule with 50 water molecules with no pbc, no cutoff). I use orca for QM part amd gromacs for MM part. But when I run mdrun -nt 1 -s sp.tpr -c sp.gro -g sp.log, the ORCA run quickly and terminate naormally, but after that mdrun run stops with a segmentation fault. Here is the error information: 
<b>Back Off! I just backed up sp.log to ./#sp.log.1#</b>
<b>Reading file sp.tpr, VERSION 4.5.5 (single precision)</b>
<b>QM/MM calculation requested.</b>
<b>there we go!</b>
<b>Layer 0</b>
<b>nr of QM atoms 9</b>
<b>QMlevel: MP2/3-21G</b>

<b>/home/user/orca_x86_64_exe_r2360/home/user/programs/orca_x86_64_exe_r2360...</b>
<b>orca initialised...</b>

<b>Back Off! I just backed up traj.trr to ./#traj.trr.1#</b>

<b>Back Off! I just backed up ener.edr to ./#ener.edr.1#</b>
<b>starting mdrun &#39;Protein&#39;</b>
<b>0 steps,      0.0 ps.</b>
<b>Calling &#39;/home/li/programs/orca_x86_64_exe_r2360/orca /home/li/Test/test_gmx/qmmm/test/sp.inp &gt;&gt; /home/li/Test/test_gmx/qmmm/test/sp.out&#39;</b>
<b>Segmentation fault</b>

Here is my sp.mdp
title                    = test
cpp                      =/lib/cpp
integrator               = md
nsteps                   = 0
dt                       = 0.001
;constraints             = none
emtol                    = 10.0
emstep                   = 0.01
nstcomm                  = 1
ns-type                  = simple
nstlist                  = 0
rlist                    = 0
rcoulomb                 = 0
rvdw                     = 0
tcoupl                   = no
pcoupl                   = no
gen-vel                  = no
nstxout                  = 1
nstlog                   = 1
nstenergy                = 1
nstvout                  = 1
nstfout                  = 1
pbc                      = no
comm_mode                = angular
; non-equilibrium md
freezegrps               = System
freezedim                = y y y
xtc_grps                 = System
energygrps               = System
;qmmm calculation
QMMM                     = yes
QMMM-grps                = QMatoms
QMMMscheme               = normal
QMmethod                 = MP2
QMbasis                  = 3-21g
QMcharge                 = 0
QMmult                   = 1
The *.ORCAINFO is:
 <b>! RKS b3lyp/g sv(p) tightscf</b>
And I also tried QMMMsheme=ONIOM, but I get the segmentation fault too. Am I missing something in all of this?  Any help would be highly appreciated.</pre>