[molpro-user] Nga_create_irreg_array being called incorrectly in molpro

Manhui Wang wangm9 at Cardiff.ac.uk
Mon Jul 28 09:15:51 BST 2008


Dear Kurt,
      This has been fixed in development version, and it should work in 
new version which will be released very soon.

Best wishes,
Manhui

Glaesemann, Kurt R (Kort) wrote:
> Greetings fellow users,
> 
> I have found that as I increased from 8 to 16 cpus some of the examples
> failed to work anymore.  For example h2o_eom did not work.  It is a cluster
> of 4 cores per node.  Version 2006.1  Patch level:      154
> 
> The problem is that nga_create_irreg_array is being called with a length of
> zero for one element of the length array.
> This is because ipair(4,*) maps 1 to 1, 2 to 3, 3 to 4, etc. and thus 2 is
> left zero.
> 
> src/ccsd/reserv_ccsd.f
>       do ip=1,npp
>         i=ipair(1,ip)
>         j=ipair(2,ip)
>         isy=ipair(4,ip)   <--- This line
>  
> Is there a patch for this, or is this just a problem with trying to over
> scale a problem?
> 
> Kurt

-- 
-----------
Manhui  Wang
School of Chemistry, Cardiff University,
Main Building, Park Place,
Cardiff CF10 3AT, UK
Telephone: +44 (0)29208 76637
E-mail: wangm9 at cardiff.ac.uk



More information about the Molpro-user mailing list