[Pw_forum] error in routine for cdiagh2 in matdyn.x

Andrew Supka supka1ar at cmich.edu
Mon Jun 13 11:35:30 CEST 2016


Hello all,

I am having this trouble when running matdyn.x with fd=.true. and
na_ifc=.true. when doing the phDOS with QE 5.4.0.

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

     Error in routine cdiagh2 (404):
     info =/= 0
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

     stopping ...

It will work without error for some q point grid densities: 5 5 5 or 7 7 7

but not others: 6 6 6 or 2 2 2 or 30 30 30

The q point grids that work seem to vary depending on structure and
chemistry. I think the problem may be occurring during the call to ZHEEV in
lapack (I'm using intel MKL) but I am not sure.

The FD phonon calcs seem to be fine and matdyn.x for phonon dispersion
doesn't seem to have this problem. I appreciate any thoughts on this.

inputs/outputs as well as details in included libraries are attached as a
tar to conserve space. (the ifc file for matdyn.x is over 200kb alone)



Thanks,

Andrew Supka

Graduate Research Asst.

Central Michigan University
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20160613/efd3b36f/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: out_tar.tar.gz
Type: application/x-gzip
Size: 22223 bytes
Desc: not available
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20160613/efd3b36f/attachment.bin>


More information about the users mailing list