RECURSIVE vs PURE and ELEMENTAL

[Jerry DeLisle]

Walter Spector wrote:
> Steve Kargl wrote:
>> ...
>>     r = factorial_inner (val * next, next - 1)
>>        1
>> Error: Function 'factorial_inner' at (1) cannot call itself, as it is not
>> RECURSIVE
> 
> If you are getting the above, then it seems the problem must have been
> fixed between 4.1.0 and 4.2.0.
> 
>>> 2.) I tried upgrading my gfortran to the latest cygwin binary.
>>> Unfortunately, it is now broken ("nothing happens" when I issue
>>> the gfortran command.  The driver just returns a exit code of 1.)
>> I don't know cygwin.  I wonder if your picking up the old 4.1.0
>> driver instead of a newer version.
> 
> The 4.1.0 version was on a SGI Altix machine - which is IA-64.
> No confusion there.
> 
> It is possible that on my home system, I have been confusing a
> non-working cygwin version of gfortran with binaries generated
> by g95.  G95 definately has problems with RECURSIVE vs PURE/ELEMENTAL.
> And I am now seeing some sort of problems with gfortran and cygwin on
> my machine...
> 
> Unfortunately I don't have time to investigate it right now.
> Perhaps tonight.
> 
> Appreciating your interest!
> 
> W.
> 

Steve, yes I will try this on Cygwin.  Walt, the Cygwin build is i686.  The 
mingw builds at the wiki will also run under cygwin, so you can try that one.  I 
think it may be 386, but I am not sure.

Also if you are really ambitious, you can build your own.  I can guide you on 
that if you need.

In the meantime I will try to do some testing.

Jerry