newbie willing to contribute...

[Steve Kargl]

On Mon, Nov 07, 2005 at 12:05:35PM -0600, Erik Schnetter wrote:
> On Nov 7, 2005, at 12:01:09, Steve Kargl wrote:
> >http://gcc.gnu.org/onlinedocs/gcc-3.4.4/g77/Complex- 
> >Intrinsic.html#index-Complex-intrinsic-846
> 
> What about introducing an intrinsic gfortran_good_real_kind(a,b) that  
> looks as the types and kinds of a and b (which can be integer, real,  
> or complex) and finds a "good" real kind that corresponds to a and b,  
> according to the rules laid out in the web page above?  Than you can say
> 
> complex(a,b) = cmplx(a,b,gfortran_good_real_kind(a,b))
> 
> and have a much easier implementation.  And since the routine is  
> codified in the compiler, the conversion rules are also clear.
> 

This is essentially what gfc_resolve_complex().

-- 
Steve