Re: [AMBER] Amber 11 and AmberTools 1.5 on Debian 6

From: Jason Swails <jason.swails.gmail.com>
Date: Sun, 12 Jun 2011 10:48:46 -0600

On Sun, Jun 12, 2011 at 8:59 AM, Jiri Wiesner <wiesner.chemi.muni.cz> wrote:

> Dear Siavoush,
> Debian 6 is a very good choice, but what is more important is the compiler.
> You might use the GNU gcc compiler suite which works well, but the intel
> compilers are the best for this kind of job. The version 11.1.072 works for
> me. Interestingly, version 11.1.073 fails with such a nice error during
> linking:
>

Ubuntu, a debian variant, is a very popular choice for personal desktops,
and there is a page on the amber website that describes how to build Amber
on Ubuntu:

http://ambermd.org/ubuntu.html

As long as your debian system has aptitude/apt-get, you should be able to
follow those instructions to build amber.


>
> /usr/local/intel/Compiler/11.1/073/lib/intel64/libirc.a(cpu_disp.o) is
> referenced by DSO ld: final link failed: Nonrepresentable section on
> output make[1]: *** [pbsa] Error 1 make[1]: Leaving directory
> `/sl4/wiesner/compiling/amber11/AmberTools/src/pbsa' make: *** [serial]
> Error 2
>

Enter why I hate the Intel compilers sometimes (though they *do* make fast
binaries).


>
> So, if you get to see something like this, do not lose hope and try a
> different compiler version ;-).
>

Another thing you can try is to remove -ipo, or change it to -ip, to remove
inter-file optimizations. Sometimes that works for me...

All the best,
Jason

Cheers,
> Jiri Wiesner
>
>
> On 12/06/11 13:10, Siavoush Dastmalchi wrote:
> > Dear List,
> >
> > I want to install Amber 11 and AmberTools 1.5 on a Linux cluster.
> Currently OS is Fedora 5 which is too old and I am thinking to upgrade my
> system. I was wondering if anyone can let me know if it is a good choice to
> upgrading the OS of my cluster to Debian 6. Has anyone got any experience
> with installation of Amber11 on Debian-based cluster?
> >
> > Cheers,
> > Siavoush
> > _______________________________________________
> > AMBER mailing list
> > AMBER.ambermd.org
> > http://lists.ambermd.org/mailman/listinfo/amber
> >
> >
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>



-- 
Jason M. Swails
Quantum Theory Project,
University of Florida
Ph.D. Candidate
352-392-4032
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Sun Jun 12 2011 - 10:00:03 PDT
Custom Search