Have to use an unreasonably big QM box in a QM/MM calculation

[ghislain.m... at gmail.com]

Hi all, 

I'm running a QM/MM calculation of ions in water and don't understand the 
following.

When I have a 20 by 20 by 20 A QM box sitting in a 40 by 40 by 40 all is 
fine and I can run it as long as I want. However, my QM region is small 
enough to fit in a 10 by 10 by 10 QM box (which would cut the cost A LOT). 
I checked the coordinates and everything and it does fit in the box. Now my 
simulation will crash with a seg fault as follows:

Program received signal SIGSEGV: Segmentation fault - invalid memory 
reference.

Backtrace for this error:
#0  0x416933D in _gfortrani_backtrace
#1  0x4144940 in _gfortrani_backtrace_handler
#2  0x7F1EDE27FD3F
#3  0x7F1EDF0130DA
#4  0x7F1ED8A6F2EA
#5  0x7F1ED8AF37E8
#6  0x7F1ED8AF228C
#7  0x7F1EDF0258DB
#8  0x7F1EE029C714
#9  0x7F1EE02A0ECC
#10  0x7F1EE0FF3B18
#11  0x10EEAA8 in 
__qs_integrate_potential_product_MOD_integrate_v_rspace._omp_fn.0 at 
qs_integrate_potential_product.F:?
#12  0x7F1EDEA4B349
#13  0x7F1EDE617181
#14  0x7F1EDE34400C
#15  0xFFFFFFFFFFFFFFFF
srun: error: n06: task 0: Segmentation fault

I really don't understand what can be the problem. Any suggestions?

Thanks in advance
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20160107/ac963b9e/attachment.htm>