[Q-e-developers] Possible bug in vloc_psi.f90

[Martin Schlipf]

This happens in the LR code that I develop. I tried to generate an example
with QE and failed and now I realized what happened. According to the
recent change I replaced all h_psiq with h_psi and s_psi calls. Because I
wrote a wrapper, I accidently called it with h_psi(lda, *lda*, m, psi,
hpsi). Because the second dimension was to large, it looped over components
where igk_k is 0, so it is only a bug in my code and not in QE.

Might I suggest to add a check to h_psi/s_psi that igk_k(n) is nonzero to
avoid someone else running into the same problem?

Martin


On Fri, Jul 1, 2016 at 5:37 PM, Paolo Giannozzi <paolo.giannozzi at uniud.it>
wrote:

> Please provide a test where this happens
>
> Paolo
>
> On Fri, Jul 1, 2016 at 6:14 PM, Martin Schlipf <martin.schlipf at gmail.com>
> wrote:
> > I recently ran the code with array boundary checking and it crashed in
> > vloc_psi.f90. In particular the array igk_k has components that are 0 and
> > then accessing nls of these elements is not within the bounds of nls
> (which
> > is 1:ngms).
> >
> > Martin
> >
> > Martin Schlipf
> > Department of Materials
> > University of Oxford
> > Parks Road
> > Oxford OX1 3PH, UK
> >
> > _______________________________________________
> > Q-e-developers mailing list
> > Q-e-developers at qe-forge.org
> > http://qe-forge.org/mailman/listinfo/q-e-developers
> >
>
>
>
> --
> Paolo Giannozzi, Dip. Scienze Matematiche Informatiche e Fisiche,
> Univ. Udine, via delle Scienze 208, 33100 Udine, Italy
> Phone +39-0432-558216, fax +39-0432-558222
> _______________________________________________
> Q-e-developers mailing list
> Q-e-developers at qe-forge.org
> http://qe-forge.org/mailman/listinfo/q-e-developers
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/developers/attachments/20160701/8b4ccfca/attachment.html>