I am getting "forrtl: error (65): floating invalid" with the release version, but the program runs OK with thedebug version and the same input. It is not easy to send the program to Intel or isolate the code, so I am looking for general guidance. I have removed optimization from the release version, but it still gives me the error. I also changed to the safer-sounding floating-point options, but the error remains.
I am trying to compare the release and debugcompiler options by looking at the vprojf file. One difference I find is that the release version hasInitLocalVarToNAN="true", but not the debug version. How do I set this to true for the debug version (using VS2010)? What I am trying to do is reproduce the problem in the debug version, so I can understand its cause, but I have not been able to do it.
BTW, I am using the latest Intel Fortran Composer.