AMBER Archive (2004)

Subject: Re: AMBER: Program error in testing of Amber8

From: Scott Brozell (
Date: Sun May 16 2004 - 18:17:44 CDT


On Sun, 16 May 2004, Do Anh Tuan wrote:
> I have compiled Amber8 with ifort on Redhat AS 3.0. Everything may be OK.
> But the testing of Amber8 has an error.
> ==============================================================
> cd qmmm/divcon; ./Run.crambin
> Segmentation fault
> ./Run.crambin: Program error
> make: *** [test.sander.QMMM] Error 1
> Can you tell me what is wrong in this case ?

Not in any detail; the tail of the output file might have been useful.
A goggle search of Run.crambin would have found some clues.
I repost a couple of points to underscore their importance in Amber8.

On Wed, 7 Apr 2004, [iso-8859-1] Alessio Comisso wrote:

> I'm tryng to install Amber 8 on a p4 with fedora core 1
> and Intel Ifort 8 and Mkl 6.1. Installation is
> ./Run.crambin: Program error
> make: *** [test.sander.QMMM] Error 1

On the other hand sander.QMMM has a big memory footprint;
for example, it won't execute on an old SGI octane with 128M:
cd qmmm/divcon; ./Run.crambin
  ./Run.crambin: Program error

It may be that your machine doesn't have enough free memory or
that your resource limits are too restrictive.

3. General notice: Beware of resource limits with Amber8
The conversion of some executables to Fortran 90 in Amber 8
has been known to cause strange failures which we believe are due
to automatic array usage.
Apparently, some compilers do not create code to detect stack overruns,
and a stack overrun may cause a segmentation violation, etc.

One can try to increase ones limits with unlimit or limit:

 Sets your user limits to the maximums.

limit datasize unlimited
limit stacksize unlimited
limit memoryuse unlimited
 Increases the individual limits.

Scott Brozell
The AMBER Mail Reflector
To post, send mail to
To unsubscribe, send "unsubscribe amber" to