[Wien] Patch to fix ROTDEF error in lapw0 for Fd3m et al.

[Martin Kroeker]

Thank you very much for the confirmation. I did see that the additional
symmetrization attempts appear to play an important role, but assumed
they were added for a reason (and I certainly did not want to bring back
any old problems that may have been successfully worked around by it).

So far I have not had time to check much beyond the SCF cycle - lapw0,
lapw1, lapw2 and possibly mixer are affected. With the mixer, the
actual issue may have been only a side effect from combining your
Intel-compiled binaries with my gfortran. (As a side note, it appears
that the line in the "x_lapw" script that adds a non-conformant declaration
for UNIT 15 to lapw2.def - a problem noted in june 2019 for version
19.1, is still present in the current distribution. Cf
https://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/msg18771.html

I can provide patches for the rotdef.f in lapw1/lapw2/mixer - which each
have a slightly different version of it due to added variables - if desired.
At a glance, the copies used in telnes, aim and arrows could be affected
as well.
-- 
Dr. Martin Kroeker            martin at ruby.chemie.uni-freiburg.de
c/o Prof.Dr. Caroline Roehr
Institut fuer Anorganische und Analytische Chemie der Universitaet Freiburg