Re: [PATCH] Don't constant fold 1.0/0.0 at compile-time

[Geoff Keating <geoffk at geoffk dot org>]

Richard Henderson <rth@redhat.com> writes:

> On Tue, Jul 01, 2003 at 02:12:22AM -0600, Roger Sayle wrote:
> > ... hence code such as the initialized global variable in fp-cmp-1.c now
> > fail to compile:
> > 
> > double dnan = 1.0/0.0 - 1.0/0.0;
> 
> Ug.
> 
> > Is there some mechanism that fold
> > can use to determine whether we're folding an initializer?
> 
> No, but I think we can add one.  Probably not to fold
> directly, as that would touch too much code, but to a
> new sister function.
> 
> > Could one of the language lawyers post their opinions on "x = 1.0/0.0"?
> 
> Not supported by a standard, per-se, but well entrenched as an idiom.

I believe that
- an expressions in a static initializer that causes an FP exception
  triggers undefined behaviour
- in other initializers (of automatic variables) FP exceptions happen
  as you'd expect from the abstract machine.

But, this is from memory, so I may be wrong.

-- 
- Geoffrey Keating <geoffk@geoffk.org>