#47 no-modules and gfortran

[Anonymous]

Originally created by: anders.j...@gmail.com
Originally owned by: anders.j...@gmail.com

If I run the code with zero gas variables (no-modules or only particles),
with gfortran start.x dies with the error:

Fortran runtime error: In line 1 of namelist init_pars:
Array reference out of bounds
LPROCZ_SLOWEST = T, LCOPYSNAPSHOTS_EXP = F, BCX = B

The problem seems to be that bcx is of size zero.

[Anonymous]

Originally posted by: anders.j...@gmail.com

I have similar experiences.

[Anonymous]

Originally posted by: wdobler@gmail.com

Which version(s) of gfortran?

Owner: ajo...@strw.leidenuniv.nl

[Anonymous]

Originally posted by: anders.j...@gmail.com

First of all let me congratulate myself on Comment 1 above where I reply to my own bug report (I thought it was Wolfgang's).

Secondly I have not seen the error message for some time now, which means that gfortran probably solved the problem in the mean time.

[Anonymous]

Originally posted by: anders.j...@gmail.com

Now this problem has returned, in the form of the variable 'timestep_scaling' which is of size mvar. If mvar is zero, as in pure particle runs, then param2.nml contains just 'timestep_scaling=', which gives problems with IDL when I read in the data. My compiler is

pc392:~> mpif90 --version
GNU Fortran (Ubuntu 4.4.3-4ubuntu5) 4.4.3
Copyright (C) 2010 Free Software Foundation, Inc.

Is there a way to get nl2idl to read lines like 'XXX=' and assume the value to be zero?

Cc: wdobler