From: leathm AT solwarra DOT gbrmpa DOT gov DOT au (Leath Muller)
Message-Id: <199703030319.NAA09782@solwarra.gbrmpa.gov.au>
Subject: Re: Netlib code [was Re: flops...]
To: billc AT blackmagic DOT tait DOT co DOT nz
Date: Mon, 3 Mar 1997 13:19:24 +1000 (EST)
Cc: djgpp AT delorie DOT com
In-Reply-To: <331A4EF0.7CF0@blackmagic.tait.co.nz> from "Bill Currie" at Mar 3, 97 04:09:20 pm
Content-Type: text

> have you tried -ffast-math ? I don't know if this will help, as I don't
> do alot of fp code, but it might help as the gcc docs say the compiler
> will violate a lot of ieee and ansi rules/assumptions (which could
> include type conversion).

I did try this switch, but it seems that DJGPP assumes that the FPU is
in full precision mode, when its really in single precision. Of course
it has no way of knowing this, which is why I wondered about a -ffpu-single
switch or something similar...

Leathal.