[gmx-users] Fatal error: ci = -2147483648

[Pandini Alessandro - Dottorati di Ricerca]

Dear All,

I got this error:
Fatal error: ci = -2147483648 should be in 0 .. 1520 [FILE nsgrid.c, 
LINE 214]

Here's my mdp parameter (important ones):
integrator               = md
tinit                    = 0
dt                       = 0.001

......

nstlist                  = 10
ns_type                  = grid
rlist                    = 0.9

coulombtype              = PME
rcoulomb                 = 0.9
rvdw                     = 0.9
fourierspacing           = 0.12
pme_order                = 4
ewald_rtol               = 1e-05

tcoupl                   = Berendsen
tc-grps                  = Protein	Non-Protein
tau_t                    = 0.1  0.1
ref_t                    = 300  300
Pcoupl                   = no

gen_vel                  = yes
gen_temp                 = 300
gen_seed                 = 173529

constraints              = all-bonds
constraint-algorithm     = Lincs
lincs-order              = 4
lincs-warnangle          = 30

I tried three times the MD with Gromacs 3.0.5 and 3.1.3, checking the 
starting structure.
No strange behaviour in temp or energy before crashes.
Every time errors are on 1520 (water).

The crashes are all in the range 750-950 ps

My machine is:
Server SGI1450 (4-cpu Xeon)
Linux RH 7.2 (kernel 2.4.7-10smp)
and I'm running parallel md with lammpi.

I had a look on the mailing list archive and found no help (a lot of 
people having similar problems but no solutions).

Any idea?
Am I doing something wrong?
Shall I check deeper the protein structure or solvent?

Thanks a lot

Bye
Alex