<posted & mailed> Johannes Wiedersich wrote:
> Hello list, > > I have a rather funny and annoying problem with a third-party fortran > program. It is a non-free scientific program for some physical > calculations. I get different output values, if I compile and run the > same fortran code on amd64 or i386. > > In fact, depending on the input file it exits with some error message in > i386, when the same file runs fine on amd64. This is even true of the > 'test' file provided with the program. Therefore it seems there is > something wrong with my configuration or the compiler on i386. > > I can run the binary compiled on i386 on amd64, but the binary compiled > on amd64 is different and yields different results. The i386-binary runs > without errors on amd64, however, when it exits on error on i386. > > I have no idea, where to start looking for what is wrong here. > Is your algorithm stable? One of the characteristics of unstable algorithms is that they give totally different results when you change precision. try running the program with single/double precisions and see if there are significant changes in the results. also try disabling the optimizations, compile with -Wall, -fbounds-check options and see if there is any change in the final results. hth raju -- Kamaraju S Kusumanchi http://malayamaarutham.blogspot.com/ -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
