Some errors after running a test?

Submitted by Donald Trump on Sun, 03/08/2020 - 09:33

Hey, can someone help me solve this problems? I did some modification on the code, and the compiling was successful, however, when I actually run a test case, it gave me the following errors on the screen:
[VB2:27524] *** Process received signal ***
[VB2:27524] Signal: Segmentation fault (11)
[VB2:27524] Signal code: Address not mapped (1)
[VB2:27524] Failing at address: 0xfffffffffffffff8
[VB2:27524] [ 0] /lib/x86_64-linux-gnu/libpthread.so.0(+0x11390)[0x7fda5c2a2390]
[VB2:27524] [ 1] ./lmp_auto[0xa8bac7]
[VB2:27524] [ 2] ./lmp_auto[0x530b3a]
[VB2:27524] [ 3] ./lmp_auto[0x491b52]
[VB2:27524] [ 4] ./lmp_auto[0x492127]
[VB2:27524] [ 5] ./lmp_auto[0x42266b]
[VB2:27524] [ 6] /lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xf0)[0x7fda5bee7830]
[VB2:27524] [ 7] ./lmp_auto[0x423529]
[VB2:27524] *** End of error message ***
Segmentation fault (core dumped)

Does anyone know what kind of these problems are? Which part of my program goes wrong here? I will appreciate if you give me some advice.

Thank you!

Donald Trump | Wed, 05/06/2020 - 15:32

I am changing some parts in fix_template_sphere.cpp to try to apply the coarse grain factor. I find that
else if (strcmp(arg[iarg],"radius") == 0)
{
hasargs = true;
if(strcmp(this->style,"particletemplate/sphere"))
error->fix_error(FLERR,this,"keyword radius only valid for particletemplate/sphere");
if (iarg+3 > narg)
error->all(FLERR,"Illegal fix particletemplate/sphere command, not enough arguments");
pdf_radius = new PDF(error);
if (strcmp(arg[iarg+1],"constant") == 0)
{
double value = atof(arg[iarg+2])*force->cg(atom_type);
if( value <= 0.)
error->all(FLERR,"Illegal fix particletemplate/sphere command, radius must be >= 0");
pdf_radius->set_params(value);
iarg += 3;
}
else
error->fix_error(FLERR,this,"invalid radius random style");
}

The radius was passed to the value from the input parameters. However, I can not find how the value of that radius being stored, and how the radius value was stored, used and passed by in building the neighbor list, computing overlaps and forces among particles? Because I find all the other radius variables were defined in their own classes.
When I want to change the forces with adding the coarse grain factor, I tried to modify the variables in contact_interface.h. But I am not sure if that will change the force calculation in those force models. I also checked those radi, radj, radsum, rsq, deltan and r, I am not sure how they were passed by the value from those classes like fix_template_sphere.cpp etc..
Can you give me some clues about this?
Thank you!

maruusa | Wed, 01/03/2024 - 03:49

I believe that the information slither io you share, will bring a lot of value to the readers and I hope that they will absorb the good and useful things.