[RFC] Generalizing the library to arbitrary floating-point modes

[François-Xavier Coudert]

> Do you plan to implement mathematical functions where the library doesn't
> provide them (say, sqrt(1._10) on an architecture where long double corresponds
> to real(16))?

That's a good question. Well, it could be done, but I guess it's not
high priority.

> writing all specific versions into a single file.

This is what I had in mind. [looking at the mail you answered... ok, I
didn't say it explicitely, but that's what i was thinking of].

FX