[Patch, fortran] PR24398 - broken module files

[FX Coudert]

Hi Paul,

> This patch deals with a problem in which something other than a  
> gfortran module file was given to read_module; the resulting error  
> was more than a little incomprehensible.  The fix is to verify that  
> the first two words of the first line read "GFORTRAN module", using  
> module.c's parser.  I do not see any point in a testcase for this.

Here's something I've wanted to ask your for some time, and now is  
the perfect occasion: do you think it would be good for us to include  
a "version number" in the module files generated by gfortran, so that  
when incompatible changes are made to the module file format, this  
number can be bumped and gfortran can error out and ask the user to  
regenerate the module file?

Thanks, and sorry if that's a stupid question,
FX