[Pw_forum] lapack error in efg calculation

Lilong Li lilong at unc.edu
Tue Mar 15 19:58:57 CET 2005

Dear pwscf users,

I got such an error when running efg.x

  On entry to DGETRF parameter number 4 had an illegal value

>From what I searched on the web and in the source file, it came from
DGETRF subroutine which calls the XERBLA error handler (in lapack.f).

I ran example24 without problem. So  the error might be due to my
system or the pseudo-potential, but with so many parameters, for me it's
hard to see what went wrong. I am ignorant to programming, so digging
into the source code is not a good cnoice. I ask for your hints and
suggestions. here are some further information and questions.

1) the error appears after printing out the ionic contribution. So it's
somewhere in calculating efg_corr.

2) Could it be something related to metallic phase? I used band

3) Does the recon file have more restriction than being norm-conserving?
I noticed that in atomic_doc/pseudo-gen directory, two files were used
to generate Si pseudopotential and reconstruction file. In si_nc_paw.in,
there's a line file_pseudopw='SiPBE_nc.bad'. Why does one need separate
files for generating pseudopoetial and reconstruction file? 

4) Is it related to symmetry? I am calculating a bulk system with no
symmtry except 3 translational vectors. I used ibrav=14.

I appreciate your help and suggestions!


More information about the users mailing list