[fortran,patch] Higher-level access to floating-point builtins

[FX]

> IIRC, it can be either.  It depends on the target.  If
> REAL(10) matches long double, then that's what is used.  If
> REAL(16) matches long double, then that's what is used.  Until
> FX finishes his patch, REAL(10) and REAL(16) cannot co-exist
> in gfortran.

You're right. In this comment, however, Tobias is right: it is real(kind=16) I'm talking about.

> This does bring to mind the question of C interop.  Does
> gfortran need to do anything special with mapping a software
> floating point type to something in C?

The only real floating-point types we need to take care for C interop are float, double and long double. In the current setup, these are never emulated with floating-point as far as I know. Also, this will not change with the introduction of __float128 support.

FX