Hi Ross,
On Fri, 11 Aug 2006, Ross Walker wrote:
> Hi Ray,
>
> > compilation aborted for _amoeba_adjust.f (code 3)
> >
> > When I lowered the optimization from O3 to O2, it passed. Is this a
> > known problem?
>
> This is a bug in the compiler itself and nothing to do with Amber other than
> the fact that that specific file is obviously laid out in some way that
> makes the compiler try to do a specific type of optimization during which
> the compiler segfaults. See if there is a newer version of the Intel
> compiler available on premier.intel.com. The one they give you for download
> on the regular site is often not the latest version. If you have the very
> latest version you may want to consider submitting a bug report to Intel.
>
> /\
> \/
> |\oss Walker
>
> | HPC Consultant and Staff Scientist |
> | San Diego Supercomputer Center |
They pay you to recommend "a newer version of the Intel compiler" !
;-)
Here's my recommendation: use an older version of the Intel compiler.
I'm still using:
Intel(R) Fortran Itanium(R) Compiler for Itanium(R)-based applications
Version 8.0 Build 20040716 Package ID: l_fc_pc_8.0.046_pl050.1
Scott Brozell
Senior Systems Developer/Engineer
Ohio Supercomputer Center
Received on Sun Aug 13 2006 - 06:07:18 PDT
