This is the mail archive of the
mailing list for the GCC project.
Re: Floating point trouble with x86's extended precision
- From: dewar at gnat dot com (Robert Dewar)
- To: reichelt at igpm dot rwth-aachen dot de, wilson at tuliptree dot org
- Cc: gcc at gcc dot gnu dot org, lucier at math dot purdue dot edu
- Date: Fri, 22 Aug 2003 07:35:17 -0400 (EDT)
- Subject: Re: Floating point trouble with x86's extended precision
> I've found FP bugs in all of the x86 compilers that I've ever used. I
> am not sure if there are any that get it right, so I am skeptical that
> gcc will ever get it right. I haven't tried any of the compilers from
> companies that specialize in FP though, maybe some of them get it right.
One way to get this right is to use the stack model of evaluation, and
then write a handler that extends the stack on overflow transparently.
I did this for the Alsys Ada compiler and it works fine. Of course this
is very different from the gcc model :-(
(this is just to respond to the claim that no compilers get this right!)