Dear Lammps Users
I am trying out restart command to see how it works and I try to generate one restart file after energy minimization which occurs fine and a file is written after that but with the message
:glibc detected double free or corruption(out):0x08744fc0***
Aborted
But still a file is generated which I try to use in the start of my subsequent simulation with the command read_restart restart.min and I get the same message again after it reads unit style ,box size and processor grid :
:glibc detected double free or corruption(out):0x098c6efc0***
Aborted
and it just ends
I am new to Unix as well as lammps and I don’t know C++ much
Can anybody tell me what’s wrong?
I get this glibc detected error in the end of other successful simulations as well but sicne I was getting good results out of it, I didn’t much bother about it
Thanks in advance
chetan
Dear Lammps Users
dear chetan,
I am trying out restart command to see how it works and I try to generate
one restart file after energy minimization which occurs fine and a file is
written after that but with the message
:***glibc detected*** double free or corruption(out):0x08744fc0***
Aborted
this most likely is caused by one of three things:
a) you have something in your input that creates
memory corruption or out-of-bounds access.
b) you have a buggy compiler/library
c) you have found a bug in some part of lammps
in all three cases, you should try to produce the
smallest possible test input that would reproduce this
and make it available so somebody else here can
cross-check and figure out which of the three applies.
in any case i would be a bit worried about the
correctness of the results. you could have gotten
lucky, but with MD you can sometimes get something
that looks ok, but still is wrong.
cheers,
axel.
Hi
Thanks for the reply
I am attaching herewith in.min, which needed backbonedata_hf4.txt as input
and created restart. min file(with the error mentioned below) which when fed
into in.hf6, gives another error mentioned below.
I am using a modified multi harmonic dihedral potential, which works fine
since I checked through my input C code, initial dihedral energy of the
system, which is exactly same as when asked to lammps to output out at the
step 0.
I am attaching .cpp and .h files for modified potential as well.
If anyone could try this out, so that it can be compared to my case, it will
be great I think. I can't attach my input backbonedata_hf4.txt file and
restart file here, as they exceed maximum permissible size limit. If anyone
is ready to try this out, I will be more than happy to send that person, my
input data file.
I appreciate the help.
Thanks
Chetan
in.min (1.01 KB)
log.min (2.63 KB)
in.hf6 (1.31 KB)
dihedral_multi_harmonic_seven.cpp (11 KB)
dihedral_multi_harmonic_seven.h (1.09 KB)
style_user.h (1.69 KB)
Please post a (simple as possible) input file and data file
that crashes and I'll take a look. It should be with the
unmodified current version of LAMMPS, not one with
your potential, since that could be causing the crash.
Steve