[gmx-users] Segmentation fault with gromacs 3.3.3

Cesar Avila cavila at fbqf.unt.edu.ar
Mon Jun 30 18:06:22 CEST 2008


Once again, in reply to myself I could state that gromacs-3.3.3 won't
build correctly on debian systems with gcc-4.3 neither. gmxtest simple
dumps the following information.

*** glibc detected *** mdrun: realloc(): invalid next size: 0x082a20d0 ***
======= Backtrace: =========
/lib/i686/cmov/libc.so.6[0xb7c0a673]
/lib/i686/cmov/libc.so.6(realloc+0x10b)[0xb7c0c5cb]
mdrun[0x8146fdd]
======= Memory map: ========
08048000-08283000 r-xp 00000000 08:05 1182361    /usr/local/gromacs/bin/mdrun
08283000-0828a000 rwxp 0023b000 08:05 1182361    /usr/local/gromacs/bin/mdrun
0828a000-082b0000 rwxp 0828a000 00:00 0          [heap]
b7a00000-b7a21000 rwxp b7a00000 00:00 0
b7a21000-b7b00000 ---p b7a21000 00:00 0
b7b5c000-b7b5e000 rwxp b7b5c000 00:00 0
b7b5e000-b7b62000 r-xp 00000000 08:05 654102     /usr/lib/libXdmcp.so.6.0.0
b7b62000-b7b63000 rwxp 00003000 08:05 654102     /usr/lib/libXdmcp.so.6.0.0
b7b63000-b7b65000 r-xp 00000000 08:05 656196     /usr/lib/libXau.so.6.0.0
b7b65000-b7b66000 rwxp 00001000 08:05 656196     /usr/lib/libXau.so.6.0.0
b7b66000-b7b68000 r-xp 00000000 08:05 963745     /lib/i686/cmov/libdl-2.7.so
b7b68000-b7b6a000 rwxp 00001000 08:05 963745     /lib/i686/cmov/libdl-2.7.so
b7b6a000-b7b81000 r-xp 00000000 08:05 653184     /usr/lib/libxcb.so.1.0.0
b7b81000-b7b82000 rwxp 00016000 08:05 653184     /usr/lib/libxcb.so.1.0.0
b7b82000-b7b83000 r-xp 00000000 08:05 653441    
/usr/lib/libxcb-xlib.so.0.0.0
b7b83000-b7b84000 rwxp 00000000 08:05 653441    
/usr/lib/libxcb-xlib.so.0.0.0
b7b84000-b7b85000 rwxp b7b84000 00:00 0
b7b85000-b7b99000 r-xp 00000000 08:05 963756    
/lib/i686/cmov/libpthread-2.7.so
b7b99000-b7b9b000 rwxp 00013000 08:05 963756    
/lib/i686/cmov/libpthread-2.7.so
b7b9b000-b7b9d000 rwxp b7b9b000 00:00 0
b7b9d000-b7ce5000 r-xp 00000000 08:05 963742     /lib/i686/cmov/libc-2.7.so
b7ce5000-b7ce6000 r-xp 00148000 08:05 963742     /lib/i686/cmov/libc-2.7.so
b7ce6000-b7ce8000 rwxp 00149000 08:05 963742     /lib/i686/cmov/libc-2.7.so
b7ce8000-b7ceb000 rwxp b7ce8000 00:00 0
b7ceb000-b7dd6000 r-xp 00000000 08:05 219962     /usr/lib/libX11.so.6.2.0
b7dd6000-b7dd9000 rwxp 000eb000 08:05 219962     /usr/lib/libX11.so.6.2.0
b7dd9000-b7dda000 rwxp b7dd9000 00:00 0
b7dda000-b7dee000 r-xp 00000000 08:05 654098     /usr/lib/libICE.so.6.3.0
b7dee000-b7def000 rwxp 00014000 08:05 654098     /usr/lib/libICE.so.6.3.0
b7def000-b7df1000 rwxp b7def000 00:00 0
b7df1000-b7df8000 r-xp 00000000 08:05 656152     /usr/lib/libSM.so.6.0.0
b7df8000-b7df9000 rwxp 00006000 08:05 656152     /usr/lib/libSM.so.6.0.0
b7df9000-b7e1c000 r-xp 00000000 08:05 963746     /lib/i686/cmov/libm-2.7.so
b7e1c000-b7e1e000 rwxp 00023000 08:05 963746     /lib/i686/cmov/libm-2.7.so
b7e1e000-b7e1f000 rwxp b7e1e000 00:00 0
b7e1f000-b7efc000 r-xp 00000000 08:05 652016     /usr/lib/libfftw3f.so.3.1.2
b7efc000-b7f02000 rwxp 000dd000 08:05 652016     /usr/lib/libfftw3f.so.3.1.2
b7f02000-b7f16000 r-xp 00000000 08:05 963748     /lib/i686/cmov/libnsl-2.7.so
b7f16000-b7f18000 rwxp 00013000 08:05 963748     /lib/i686/cmov/libnsl-2.7.so
b7f18000-b7f1a000 rwxp b7f18000 00:00 0
b7f1d000-b7f29000 r-xp 00000000 08:05 676535     /lib/libgcc_s.so.1
b7f29000-b7f2a000 rwxp 0000b000 08:05 676535     /lib/libgcc_s.so.1
b7f2a000-b7f2d000 rwxp b7f2a000 00:00 0
b7f2d000-b7f47000 r-xp 00000000 08:05 677881     /lib/ld-2.7.so
b7f47000-b7f49000 rwxp 00019000 08:05 677881     /lib/ld-2.7.so
bf985000-bf99c000 rwxp bffe9000 00:00 0          [stack]
ffffe000-fffff000 r-xp 00000000 00:00 0          [vdso]
sh: line 1:  6267 Abortado                mdrun > mdrun.out 2>&1
FAILED. Check files in rb1
1 out of 16 simple tests FAILED

On the other hand, mpi version of gromacs built on the same system,
succesfully passes all simple test.

Cesar Avila escribió:
> It seems to be a problem related to the debian package. I have compiled
> gromacs 3.3.3 from source with gcc-4.3 and apparently the problem has
> dissapeared. Perhaps, the binaries on the debian package were compiled
> with gcc-4.1.
>
>
> Cesar Avila escribió:
>> Dear all,
>> I am new to the gromacs simulation package. Thus far I have been using
>> Charmm and NAMD, but now I am interested on testing Marrink's
>> Coarse-Grained Martini FF. I managed to setup an initial configuration for
>> a protein in a water box. After a few minimization steps, I ran the
>> simulation for about 10 ns until it crashed with a Segmentation fault
>> message.
>> Using tpbconv, I resumed the simulation from 10200 ps step. Nevertheless I
>> systematically found a segmentation fault  error on step 6530. Using mdrun
>> -nocompact gives no additional information about the crash. When I use
>> mdrun -debug the simulation is able to run without any problem (until I
>> run out of disk space).
>> I found the same error to occur on both a Pentium IV (debian etch) and an
>> Intel Core 2 Duo (Debian lenny). Both systems
>> are running gromacs 3.3.3 from Debian testing packages. If anyone is
>> interested on reproducing the error on their own computer, I can send them
>> the tpr file (476K compressed).
>> Best regards
>> Cesar Avila
>> Universidad Nacional de Tucuman.
>>
>>
>> _______________________________________________
>> gmx-users mailing list    gmx-users at gromacs.org
>> http://www.gromacs.org/mailman/listinfo/gmx-users
>> Please search the archive at http://www.gromacs.org/search before posting!
>> Please don't post (un)subscribe requests to the list. Use the
>> www interface or send it to gmx-users-request at gromacs.org.
>> Can't post? Read http://www.gromacs.org/mailing_lists/users.php
>>
>
>
>
> _______________________________________________
> gmx-users mailing list    gmx-users at gromacs.org
> http://www.gromacs.org/mailman/listinfo/gmx-users
> Please search the archive at http://www.gromacs.org/search before posting!
> Please don't post (un)subscribe requests to the list. Use the
> www interface or send it to gmx-users-request at gromacs.org.
> Can't post? Read http://www.gromacs.org/mailing_lists/users.php
>








More information about the gromacs.org_gmx-users mailing list