Date: Tue, 7 Dec 1999 12:03:41 +0200 (IST)
From: Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
X-Sender: eliz AT is
To: Moshe Greenberg <grini AT rafael DOT co DOT il>
cc: djgpp AT delorie DOT com
Subject: Re: fftw give nonsense with float ok with double
In-Reply-To: <384bf935.1468622@news.netvision.net.il>
Message-ID: <Pine.SUN.3.91.991207120322.29709E-100000@is>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Reply-To: djgpp AT delorie DOT com
X-Mailing-List: djgpp AT delorie DOT com
X-Unsubscribes-To: listserv AT delorie DOT com
Precedence: bulk


On Mon, 6 Dec 1999, Moshe Greenberg wrote:

> I compile the fftw by the djgpp gcc 2.952 optimization -o3  to make
> lib.
> The lib work fine on double and give NAN on float.

Questions/suggestions:

 1. What is fftw?
 2. Why do you use -O3?  If you don't have any special reason, try -O2
    instead, it usually yields better results anyway.
 3. Did you link with -lm?  If not, try it.
 4. Did you enable the -fstrict-aliasing option?  If so, disable it.