Hello all,
I tried to look for a place to send information about bugs found within
LAMMPS' website but I did not find. Is there such a place for that?
I have found LAMMPS segfoult when I write:
.
.
.
variable prod_cycles equal 100
fix temp_field_fix flow_g ave/chunk 10 \(v\_prod\_cycles/10\)
{prod_cycles}binning_cmp temp file temp_profile.dat
.
.
.
I tracked down the problem to be "\{prod\_cycles\}binning\_cmp" so this
line has to be "{prod_cycles} binning_cmp" (with a space in between).
So it seems to be a case not handled when the parameters of this fix are
parsed. Although the syntax is wrong a segfoult is not desirable.
The output (printed to screen) is attached. I am using the May2015
version with opt, user-misc, kspace and molecule packages compiled. I do
not include the log because due to the crash it gives no relevant
information. If you need the input files I can give them to you (there
are a few).
I just wanted to help, so tell me if you need more info.
Regards,
Eduardo
output.txt (12.8 KB)
Hello all,
I tried to look for a place to send information about bugs found within
LAMMPS' website but I did not find. Is there such a place for that?
instructions for reporting (potential) bugs are given in the manual
(as one should expect).
http://lammps.sandia.gov/doc/Section_errors.html#err_2
axel.
Thanks to point me to the right page. Apparently I did it the right way!
Regards,
Eduardo
Thanks to point me to the right page. Apparently I did it the right way!
not entirely.
there are many ways how somebody can screw up with variables and you
don't provide a minimal input deck to reproduce it.
chances are, that the "bug" is somewhere else in your input, and this
line in your output is a very strong hint to confirm it:
ERROR: Invalid thermo keyword in variable formula (../variable.cpp:1716)
axel.
Well, if I separate the parameter names (as the command expect) the
simulation runs fine. So it is likely that there is no problem
elsewhere. It is just a undesired case not taken into account while
checking parameters I guess. , when I have a bit of time I will make a
simple test case so if that makes things easier for you.
Regards,
Eduardo
When I add these 3 lines (without the space) to bench/in.lj
variable prod_cycles equal 100
compute binning_cmp all chunk/atom bin/1d x lower 1.0
fix temp_field_fix all ave/chunk 10 (v_prod_cycles/10) &
{prod_cycles}binning_cmp temp file temp_profile.dat
I get this error:
ERROR: Expected integer parameter in input script or data file (…/fix_ave_chunk.cpp:51)
When I add the space, it runs fine.
This is with the current 6 Jul 2015 version.
Steve