Hi Adam, I have a system with gfortran 4.1.1.20070105 (Red Hat 4.1.1-51), so very close to yours, and see the same problem (even the same line number for feff84.f). I think the main issue is not the arithmetic ifs (though they generate piles of warnings). I think the real issue is that there is uninitialized data for the 'kappa' array when writing pot.bin. I'm not sure what the best solution is... Feff is full of uninitialized data and undocumented arrays and code. Doing any real work on the single-source-code-file is basically impossible, and I've long ago given up trying to actually understand Feff8. Strangely, compiling feff84.f with g77 from gcc 3.2.3 on the same system seems to work fine. Hope that helps, --Matt On Tue, Mar 4, 2008 at 2:07 AM, Adam Webb wrote:

Hi all,

I have having difficulty compiling feff 8.4. I am using gcc version 4.1.1 20070105 (Red Hat 4.1.1-52) on Scientific Linux 5.0. It compiles but gives many warnings like: Warning: Obsolete: arithmetic IF statement at (1) In file FEFF810.F:4208

if (i-1) 85,85,83

The compile finished however and I tried to run it. The input file was a BN example file that came with Feff 8.1. It runs for a little while and stops with: Done with module 1: potentials. Calculating LDOS ... At line 47994 of file feff84.f Fortran runtime error: Bad value during integer read

I compiled using: gfortran -O2 -o feff84 feff84.f

Has anyone seen this sort of thing? Or compiled Feff 8.4 with gfortran? I did the same with feff 8.1 and it compiled and ran correctly.

Cheers, Adam

-- ---------------------------------------------------------- Dr. M. Adam Webb HASYLAB at DESY Notkestrasse 85 D-22607 Hamburg Germany

Phone: +49 40 8998-1994, Fax: +49 40 8998-2787 E-mail: adam.webb@desy.de ----------------------------------------------------------- _______________________________________________ Ifeffit mailing list Ifeffit@millenia.cars.aps.anl.gov http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit