V3 PATCH: numeric_limits<> support, fix PR/3865

[Richard Henderson]

On Tue, Sep 03, 2002 at 11:57:46AM -0600, Roger Sayle wrote:
> bool has_snan = !(__builtin_snan("") == 0.0);

If -fsignaling-nan, this ought to result in sigfpe.
Thus you can't do that.

> Otherwise I agree that we might also need __builtin_has_snan{,f,l} to
> determine whether we support signaling NaNs.

There's two questions of "support".  (1) Can we represent the value,
and (2) do we conserve signals.  I suspect that std::numeric_limits
only cares about question 1, and not about question 2.


r~