[Wien] system- or usererror?

[Peter Blaha]

Most likely a user error.

TiC should converge within 6-7 iterations.

Was this the first  "run_lapw", or some continuation step ?

Do not forget to "save_lapw" (or at least  rm *.broyd*) after an scf
calculation (and befor you start a new one with modified input
parameters).
The *.broyd* files contain the charge history of an scf cycle, and with
different input (RKMAX, lattice parameters,...) one must remove this
"history", otherwise the scf cycle diverges.

> I think I got WIEN compiled successfully now, but still run into problems. I'm
> not sure whether WIEN is still not compiled error-free or whether it's an
> user error by me.
> I try to calculate the sample-TiC-calculation as instructed in the manual.
> Everything looks good, but the scf-run fails.
> When doing the run interactively, I get:
> in cycle 12    ETEST: .0359855000000000   CTEST: .0027511
> FORTRAN STOP  LAPW0 END
> FORTRAN STOP SECLR4 - Error
>
> and the file "lapw1.error" reads:
>  Cholesky INFO =           72
>  'SECLR4' - POTRF (Scalapack/LAPACK) failed.


                                      P.Blaha
--------------------------------------------------------------------------
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-15671             FAX: +43-1-58801-15698
Email: blaha at theochem.tuwien.ac.at    WWW: http://info.tuwien.ac.at/theochem/
--------------------------------------------------------------------------