Minimum floating-point requirements

[Joseph S. Myers]

On Mon, 17 Feb 2014, Rich Felker wrote:

> Anyone doing interval arithmetic wants to change rounding modes.
> However, the need for correct rounding semantics is independent of
> rounding mode. You want it even in default rounding mode.

To be clear: I'm not proposing to make this arithmetic correctly rounding 
or support interval arithmetic, simply (given -mieee) make it produce 
results in all rounding modes that are actually valid long double values 
at all (and in the case of overflow or underflow, are consistent with the 
rounding mode).

-- 
Joseph S. Myers
joseph@codesourcery.com