Date:	Tue, 22 Apr 2003 17:51:07 +0200
From:	"Eli Zaretskii" <eliz AT elta DOT co DOT il>
Sender:	halo1 AT zahav DOT net DOT il
To:	djgpp-workers AT delorie DOT com
Message-Id:	<3395-Tue22Apr2003175107+0300-eliz@elta.co.il>
X-Mailer:	emacs 21.3.50 (via feedmail 8 I) and Blat ver 1.8.9
In-reply-to:	<200304221232.OAA06267@lws256.lu.erisoft.se> (message from Martin
	Stromberg on Tue, 22 Apr 2003 14:32:06 +0200 (MET DST))
Subject:	Re: Yet another try on nan in strto{f,d,ld}
References:	<200304221232 DOT OAA06267 AT lws256 DOT lu DOT erisoft DOT se>
Reply-To:	djgpp-workers AT delorie DOT com
Errors-To:	nobody AT delorie DOT com
X-Mailing-List:	djgpp-workers AT delorie DOT com
X-Unsubscribes-To:	listserv AT delorie DOT com

> From: Martin Stromberg <Martin DOT Stromberg AT epl DOT ericsson DOT se>
> Date: Tue, 22 Apr 2003 14:32:06 +0200 (MET DST)
> 
> Ok. But do you see that it's two different paragraphs describing
> different cases of input? It's not redundant, IMHO.

It was sufficiently redundant to catch my eye.  But I won't argue if
you feel that it's needed.

> > That is, won't that line by itself produce a SIGFPE in that case, and
> > if it does, is that okay, as far as C9x and our common sense are
> > concerned?
> 
> NAN is a QNaN (of type float), so it shouldn't. I think that it
> remains that even when we assign it to double and long double
> variables.

I'd still like to hear someone who could actually run a test program
that enabled FP exceptions and then assigned a NaN to a variable.

- Raw text -