[LLVMdev] [RFC] Add a simple soft-float class

[Robinson, Paul]

> > Because it seems like you (and Andy) are supporting this as a
> suboptimal but convenient replacement for a proper integer-based
> representation.
> 
> In a way, but I think of it as a suboptimal but safe replacement for
> hardware floats.  We currently use hardware floats as a suboptimal but
> convenient replacement for proper integer-based representations -- I'd
> like
> to swap in a soft-float instead.
> 
> For some problems, proper integer-based representations are really hard,
> and can resemble a poor man's soft-float even when done right.

In the fixed-point package I used to own, doing anything complicated
meant converting to soft-float and back.  ("Complicated" includes
division.)  Soft-float behaves more intuitively and doesn't require
the client to manage scale, which is a big plus in my book.
--paulr