Message-Id: <200003161216.HAA02023@delorie.com>
From: "Dieter Buerssner" <buers AT gmx DOT de>
To: Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
Date: Thu, 16 Mar 2000 13:16:03 +0100
MIME-Version: 1.0
Content-type: text/plain; charset=US-ASCII
Content-transfer-encoding: 7BIT
Subject: Re: Unnormals???
CC: djgpp-workers AT delorie DOT com
References: <Pine DOT LNX DOT 4 DOT 10 DOT 10003151904260 DOT 21378-100000 AT acp3bf>
In-reply-to: <Pine.SUN.3.91.1000316114140.3117C-100000@is>
X-mailer: Pegasus Mail for Win32 (v3.12b)
Reply-To: djgpp-workers AT delorie DOT com

On 16 Mar 00, Eli Zaretskii wrote:

> You seem to assume that the FPU has some way of dealing with these bit
> patterns in a reasonable way.  This isn't true: the FPU treats them as
> if they were NaNs; no useful FP result can ever be generated out of
> their use.

One small addition: fldt fstpt leave the bit pattern untouched (here 
at least).

> So there's no ``punishment'' here.  Quite to the contrary, this is
> IMHO a valuable debugging aid, exactly like "(null)".

And like aborting would be as well.

After searching through the Standard for normalize and after looking
at fpclassify(), it seems to me, that unnormalized numbers are not
allowed. fpclassify() cannot produce a result for them. There is
only one mention of non-normalized (in Annex J), but I think 
subnormal (denormal) is meant there.

Perhaps somebody should ask in comp.std.c.

Regards,
Dieter