From: Jim Phillips (jim_at_ks.uiuc.edu)
Date: Tue Sep 16 2003 - 21:19:03 CDT
This error happens on every test simulation, and they used to work? Hmmm.
Your compiler and processor are both newer that what I've gotten to test
at NCSA, and the "unique coordinates" warning really means that extra
exclusions were calculated, which is very unusual. You might want to back
off on the optimization of ComputeNonbonded..., maybe start by commenting
out CXXNOALIASOPTS from arch/Linux-ia64-ecc.arch. If that eliminates the
warning then we're at least on the right track (but Itanium is pretty
pointless without those optimization flags).
-Jim
On Tue, 16 Sep 2003, Jerry Ebalunode wrote:
> Hi all,
> I just compiled namd2.5b2 on an itanium2 cluster that uses SCALI's
> implementation of the mpi. When I run the binary, I always get this warning
> "Warning: Not all atoms have unique coordinates." in the output log. Does
> anyone know how to correct this problem? I initially thought it had to do
> with the tcl or fftw libraries I was using, so used the precompiled
> libraries from http://www.ks.uiuc.edu/Research/namd/libraries/ meant for my
> architecture. However, after recompiling and running namd2 I still had the
> same problems.
>
>
> Compiler information
> intel C/C++ compiler for itanium ecc version 7.1
>
> --
> Cheers,
>
> Jerry Ebalunode
> Graduate Research Assistant
> RM 402F Houston Science Center
> Phone: 713-743-8367
> Dept. of Biology and Biochemistry
> University of Houston.
> 4800 Calhoun Road
> Houston, TX 77204
>
This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:37:02 CST