AMBER Archive (2005)

Subject: Re: AMBER: seg fault error in resp

From: FyD (fyd_at_u-picardie.fr)
Date: Thu Oct 27 2005 - 09:39:44 CDT

Quoting Kateryna Miroshnychenko <kateryna_mirosh_at_ire.kharkov.ua>:

> >> Check carefully, the number of sections in your input,...
> >
> > If you have any programming skills, you could also try compiling resp
> > with the debug flag (-g) set and find where it is coredumping and fix
> > the bug.. input may be bad, but ideally the program should tell what
> > the problem is rather than trip over itself.
>
> Thanks all for trying help me. Input was OK. I just forgot to increase
> also maxmol (besides maxq) in resp.f. The default maxmol was 30 and it was
> not enough
> in my case. After I've increased it to 100, all works fine :))
> May be it is worth to add something like following in SUBROUTINE mult_mol:
>
> c trap for too many molecules
> if( nmol .gt. maxmol) then
> write(6,'(t2,a,i3,/,t2,a,i3)')
> . ' ERROR: too many molecules', nmol,
> . ' Maximum allowed: ', maxmol
> call mexit(6, 1)
> endif
>

Here is what I changed in resp.f:

       parameter (maxq = 2000)
       parameter (maxlgr = 200)
       parameter (maxmol = 100)

regards, Francois

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu