denis,
please always copy the list on your replies.
Dear Axel,
I have just find out that it does not work because of SELF,
but it does work with direct name of the file.
it works for me with SELF, too.
what platform are you running on?
That's probably because inside those loops I have
include filename1
which might affects SELF somehow.
no. i use include files, too.
in fact, i have loops *outside*
and *inside* the include files
both using jumps and labels.
you may want to check, whether
the jump and label statements
are well balanced, though.
axel.
Dear Axel,
denis,
please always copy the list on your replies.
Sorry, I forgot adding it.
Dear Axel,
I have just find out that it does not work because of SELF,
but it does work with direct name of the file.
it works for me with SELF, too.
what platform are you running on?
I run on Ubuntu 11.
Sources are compiled with open-mpi, I run on 1 processor.
That's probably because inside those loops I have
include filename1
which might affects SELF somehow.
no. i use include files, too.
in fact, i have loops *outside*
and *inside* the include files
both using jumps and labels.
you may want to check, whether
the jump and label statements
are well balanced, though.
What do you mean by "well balanced"?
My input script is in a sense similar in.elastic (for elastic constants),
so that I first input a file with box and atoms definitions, then another one with potentials,
minimize and write restart file. And inside loop I include another file which does:
clear..
read_restart and input potentials
displace_atoms...
denis,
if you don't send a *complete* script
that can be run right away by somebody,
there are little chances that somebody
will be debugging it. strip it down to the
smallest possible input that will reproduce
the problem and then post it.
check out for example the regression
test library for restart2data i've started
to work on:
http://github.com/akohlmey/lammps-omp/tree/openmp-master/regtest/restart
axel.
Dear Axel,
I tried running a serial version of LAMMPS, that went fine, no problems.
But the parallel version (open mpi) does gives an error
ERROR: Label wasn't found in input script (input.cpp:160)
Below is a simple script that reproduces the error with
mpirun -np 1 lmp_openmpi < in.myInput
----------- in.myInput -------------
variable j loop 2
label loop_j
print "-------------------------------- j = $j ---------------------"
variable n equal 3
variable i loop -$n $n
label loop_i
print "-------------------------------- i = $i ---------------------"
next i
jump SELF loop_i
variable h delete
next j
jump SELF loop_j