Segmentation fault on cluster

Hi,

i’m trying to run Lammps 1-feb-2014 on 16 processors on my university’s cluster with the reax/c potential. A few steps after the simulation starts it stops with a segmentation fault error message, it also generates a core.17445 file of about 180 Mb:
core.17445: ELF 64-bit LSB core file x86-64, version 1 (SYSV), SVR4-style, from ‘/share/apps/lammps-1Feb14/lmp_hal’

The cluster admin says he made some tests and he can run lammps in 1, 2 or 4 processors but not more, but he needs more information.

I’ve found this other thread about a similar error but it didn’t provide a solution:
http://sourceforge.net/p/lammps/mailman/message/29987953/

I’d appreciate any help in solving this issue

Here are the error message and input file:

[compute-0-17:12575] *** Process received signal ***
[compute-0-17:12575] Signal: Segmentation fault (11)
[compute-0-17:12575] Signal code: Address not mapped (1)
[compute-0-17:12575] Failing at address: 0x2aaab0f1d000
[compute-0-17:12575] [ 0] /lib64/libpthread.so.0 [0x2ba5851864c0]
[compute-0-17:12575] [ 1] /share/apps/lammps-1Feb14/lmp_hal(_ZN9LAMMPS_NS9PairReaxC7computeEii+0xc4a) [0x879f5a]
[compute-0-17:12575] [ 2] /share/apps/lammps-1Feb14/lmp_hal(_ZN9LAMMPS_NS6Verlet3runEi+0x77f) [0x9b96ff]
[compute-0-17:12575] [ 3] /share/apps/lammps-1Feb14/lmp_hal(_ZN9LAMMPS_NS3Run7commandEiPPc+0xc78) [0x983a08]
[compute-0-17:12575] [ 4] /share/apps/lammps-1Feb14/lmp_hal(_ZN9LAMMPS_NS5Input15command_creatorINS_3RunEEEvPNS_6LAMMPSEiPPc+0x29) [0x6acb49]
[compute-0-17:12575] [ 5] /share/apps/lammps-1Feb14/lmp_hal(_ZN9LAMMPS_NS5Input15execute_commandEv+0x2323) [0x6b12f3]
[compute-0-17:12575] [ 6] /share/apps/lammps-1Feb14/lmp_hal(_ZN9LAMMPS_NS5Input4fileEv+0x15e) [0x6ae8ae]
[compute-0-17:12575] [ 7] /share/apps/lammps-1Feb14/lmp_hal(main+0xad) [0x6c8ccd]
[compute-0-17:12575] [ 8] /lib64/libc.so.6(__libc_start_main+0xf4) [0x39f841d974]
[compute-0-17:12575] [ 9] /share/apps/lammps-1Feb14/lmp_hal [0x48ee19]
[compute-0-17:12575] *** End of error message ***
[compute-0-64:17445] *** Process received signal ***
[compute-0-64:17445] Signal: Segmentation fault (11)
[compute-0-64:17445] Signal code: Address not mapped (1)
[compute-0-64:17445] Failing at address: 0x2aaab0f2a000
[compute-0-64:17445] [ 0] /lib64/libpthread.so.0 [0x2b1f404c54c0]
[compute-0-64:17445] [ 1] /share/apps/lammps-1Feb14/lmp_hal(_ZN9LAMMPS_NS9PairReaxC7computeEii+0xc4a) [0x879f5a]
[compute-0-64:17445] [ 2] /share/apps/lammps-1Feb14/lmp_hal(_ZN9LAMMPS_NS6Verlet3runEi+0x77f) [0x9b96ff]
[compute-0-64:17445] [ 3] /share/apps/lammps-1Feb14/lmp_hal(_ZN9LAMMPS_NS3Run7commandEiPPc+0xc78) [0x983a08]
[compute-0-64:17445] [ 4] /share/apps/lammps-1Feb14/lmp_hal(_ZN9LAMMPS_NS5Input15command_creatorINS_3RunEEEvPNS_6LAMMPSEiPPc+0x29) [0x6acb49]
[compute-0-64:17445] [ 5] /share/apps/lammps-1Feb14/lmp_hal(_ZN9LAMMPS_NS5Input15execute_commandEv+0x2323) [0x6b12f3]
[compute-0-64:17445] [ 6] /share/apps/lammps-1Feb14/lmp_hal(_ZN9LAMMPS_NS5Input4fileEv+0x15e) [0x6ae8ae]
[compute-0-64:17445] [ 7] /share/apps/lammps-1Feb14/lmp_hal(main+0xad) [0x6c8ccd]
[compute-0-64:17445] [ 8] /lib64/libc.so.6(__libc_start_main+0xf4) [0x3c4ea1d974]
[compute-0-64:17445] [ 9] /share/apps/lammps-1Feb14/lmp_hal [0x48ee19]
[compute-0-64:17445] *** End of error message ***

Hi,
i'm trying to run Lammps 1-feb-2014 on 16 processors on my university's
cluster with the reax/c potential. A few steps after the simulation starts
it stops with a segmentation fault error message, it also generates a
core.17445 file of about 180 Mb:
core.17445: ELF 64-bit LSB core file x86-64, version 1 (SYSV), SVR4-style,
from '/share/apps/lammps-1Feb14/lmp_hal'

The cluster admin says he made some tests and he can run lammps in 1, 2 or
4 processors but not more, but he needs more information.
I've found this other thread about a similar error but it didn't provide a
solution:
http://sourceforge.net/p/lammps/mailman/message/29987953/

​actually, that mail thread *does* provide a (possible) solution: try your
input with the latest LAMMPS version!​

to be a bit more specific: segmentation faults are very nondescript error
messages and can have many reasons. most often they are the result of bad
input parameters or a bad starting configuration. occasionally, they
indicate a bug, but in that case, you have to provide a (small) test case
and first verify that the segmentation fault still happens with the latest
version of the code.

axel.