[gmx-users] need help with mdrun

[Erik Lindahl]

Hi Dodi,

I'll try to look into it, but I'm travelling in Europe right now.

Since the error seems to be repeatable across different platforms it 
would probably help if you could send me the files.

Cheers,

Erik

PS: Remind me if I don't get back to you - I've got quite a lot of mail 
piled up the last two weeks :-)






Dodi Heryadi wrote:

> Hello,
>
> One of our Gromacs users at NCSA reported that his Gromacs job was 
> aborted after a few seconds of
> running mdrun. The Gromacs he used was compiled with MPI on SGI IRIX 
> 6.5.14 using MIPSPro 7.3.1.3m
> compiler. Similar behavior was also observed when I ran the same tpr 
> file on MPI Gromacs compiled on
> our IA32 Linux Cluster using Intel Compiler.
>
> I'm new to Gromacs. Nevertheless, using dbx on the SGI Origin I tried 
> to find out what might cause the error.
> At the end of this  note, I  am including the last few lines from 
> running mdrun and the first few lines from running dbx.
> We'd really  appreciate it if someone could look at this problem and 
> shed us some light. I'll be happy to send you the input files.
>
> thanks,
> --Dodi
> -------------------------------------------------------------------------
> Dodi Heryadi, PhD
> NCSA @ University of Illinois at Urbana-Champaign
> Phone:(217)-333-6078 Fax: (217)-244-2909
> -------------------------------------------------------------------------- 
>
>
> ================ start of running mdrun ==============================
>
> Back Off! I just backed up md.log to ./#md.log.2#
> Reading file dppc128c.tpr, VERSION 3.1.4 (single precision)
> Reading file dppc128c.tpr, VERSION 3.1.4 (single precision)
>
> Back Off! I just backed up ener.edr to ./#ener.edr.2#
> starting mdrun 'dppc in water'
> 250000 steps, 500.0 ps.
>
> MPI: MPI_COMM_WORLD rank 0 has terminated without calling MPI_Finalize()
> MPI: aborting job
>
> ================== end of running mdrun =============================
>
>
> =================== start of running dbx 
> =========================================
> modi4 167% dbx 
> /usr/apps/chemistry/gromacs-3.1.4/mips-sgi-irix6.5/r10000/bin/mdrun 
> core.43780720
> dbx version 7.3.1 68542_Oct26 MR Oct 26 2000 17:50:34
> Core from signal SIGBUS: Bus error
> (dbx) where
> > 0 inl3100(0x107fd804, 0x11181178, 0x111cc1a8, 0x1149b2c0, 
> 0x111b3198, 0x102262a0, 0x102263f0, 0x1119a188) 
> ["/usr/apps/chemistry/gromacs-3.1.4/gromacs-3.1.4/src/gmxlib/innerf.f":16460, 
> 0x1011e470]
>  1 do_fnbf(log = 0xdb33e50, cr = 0x10227148, fr = 0x107fd710, x = 
> 0x107977f8,
> f = 0x10b489f0, mdatoms = 0x10226fa0, egnb = 0x10223748, egcoul = 
> 0x102236d0, box_size = 0xffffff2668, nrnb = 0xffffff2980, lambda = 
> 0.0, dvdlambda = 0x10223850, bLR = 0, eNL = -1) 
> ["/usr/apps/chemistry/gromacs-3.1.4/gromacs-3.1.4/src/gmxlib/fnbf.c":729, 
> 0x100cacd8]
>  2 force(fp = 0xdb33e50, step = 0, fr = 0x107fd710, ir = 0x10224da0, 
> idef = 0x1078b3a0, nsb = 0x10223d78, cr = 0x10227148, mcr = (nil), 
> nrnb = 0xffffff2980, grps = 0x10222290, md = 0x10226fa0, ngener = 2, 
> opts = 0x10224f20, x = 0x107977f8, f = 0x10b489f0, epot = 0x102237a8, 
> fcd = 0x10226e98, bVerbose = 0, box = 0x10224ff0, lambda = 0.0, graph 
> = 0x10222248, excl = 0x107966c8, bNBFonly = 0, lr_vir = 0xffffff2dd0, 
> mu_tot =
> 0xffffff2e20, qsum = 1.1444091796875e-05, bGatherOnly
> = 0) 
> ["/usr/apps/chemistry/gromacs-3.1.4/gromacs-3.1.4/src/mdlib/force.c":880, 
> 0x1005df7c]
>  3 do_force(log = 0xdb33e50, cr = 0x10227148, mcr = (nil), parm = 
> 0x10224da0,
> nsb = 0x10223d78, vir_part = 0xffffff2da8, pme_vir = 0xffffff2dd0, 
> step = 0, nrnb = 0xffffff2980, top = 0x1078b398, grps = 0x10222290, x 
> = 0x107977f8, v = 0x107ca600, f = 0x10b489f0, buf = 0x10b15be8, 
> mdatoms = 0x10226fa0, ener = 0x102237a8, fcd = 0x10226e98, bVerbose = 
> 0, lambda = 0.0, graph = 0x10222248, bNS = 1, bNBFonly = 0, fr = 
> 0x107fd710, mu_tot = 0xffffff2e20, bGatherOnly = 0) 
> ["/usr/apps/chemistry/gromacs-3.1.4/gromacs-3.1.4/src/mdlib/sim_util.c":282, 
> 0x10054d00]
>
> ======================= end of running dbx 
> ================================================
>
> _______________________________________________
> gmx-users mailing list
> gmx-users at gromacs.org
> http://www.gromacs.org/mailman/listinfo/gmx-users
> Please don't post (un)subscribe requests to the list. Use the www 
> interface or send it to gmx-users-request at gromacs.org.