[gmx-users] Segmentation fault

[Justin Lemkul]

On 10/29/14 7:54 AM, Vo Cam Quy wrote:
> Dear All,
> I am running my energy minimization using gromacs 4.6.5 and have got this
> problem
> Steepest Descents:
>     Tolerance (Fmax)   =  1.00000e+02
>     Number of steps    =       100000
> [uc1n996:06400] *** Process received signal ***
> [uc1n996:06400] Signal: Segmentation fault (11)
> [uc1n996:06400] Signal code: Address not mapped (1)
> [uc1n996:06400] Failing at address: 0x2ae4edece048
> [uc1n996:06400] [ 0] /lib64/libpthread.so.0() [0x315fe0f710]
> [uc1n996:06400] [ 1]
> /opt/bwhpc/common/chem/gromacs/4.6.5/lib/libgmx_mpi.so.8(do_nonbonded_listed+0x71d)
> [0x2afce6f43b7d]
> [uc1n996:06400] [ 2]
> /opt/bwhpc/common/chem/gromacs/4.6.5/lib/libgmx_mpi.so.8(+0x10ddc4)
> [0x2afce6efcdc4]
> [uc1n996:06400] [ 3]
> /opt/bwhpc/common/chem/gromacs/4.6.5/lib/libgmx_mpi.so.8(calc_bonds+0xb92)
> [0x2afce6efbc12]
> [uc1n996:06400] [ 4]
> /opt/bwhpc/common/compiler/intel/compxe.2013.5.192/lib/intel64/libiomp5.so(__kmp_invoke_microtask+0x93)
> [0x2afce8bed293]
> [uc1n996:06400] *** End of error message ***
> Segmentation fault
> I was trying on different machine and folder and checked the available
> space but have the same error. My em.mdp is
> ;define         = -DFLEXIBLE
> constraints     = none
> integrator      = steep
> nsteps          = 100000
> nstlist         = 10
> ns_type         = grid
> rlist           = 1.3
> coulombtype     = PME
> rcoulomb        = 1.3
> vdwtype         = cut-off
> rvdw            = 1.4
> fourierspacing  = 0.12
> fourier_nx      = 0
> fourier_ny      = 0
> fourier_nz      = 0
> pme_order       = 4
> ewald_rtol      = 1e-5
> optimize_fft    = yes
> emtol           = 100
> emstep          = 0.01
> with 8 proteins in water. Could anybody please help me explain why it
> happened and how to fix this? I am really new to Gromacs.

Something is catastrophically wrong with the system.  Either the coordinates 
overlap, the topology is broken, or something else.  The cutoffs look very 
randomly chosen, which is probably not the leading cause of your issue, but can 
contribute to broken physics.

-Justin

-- 
==================================================

Justin A. Lemkul, Ph.D.
Ruth L. Kirschstein NRSA Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 629
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

jalemkul at outerbanks.umaryland.edu | (410) 706-7441
http://mackerell.umaryland.edu/~jalemkul

==================================================