Mailing-List:	contact cygwin-help AT sourceware DOT cygnus DOT com; run by ezmlm
List-Unsubscribe:	<mailto:cygwin-unsubscribe-archive-cygwin=delorie DOT com AT sourceware DOT cygnus DOT com>
List-Subscribe:	<mailto:cygwin-subscribe AT sourceware DOT cygnus DOT com>
List-Archive:	<http://sourceware.cygnus.com/ml/cygwin/>
List-Post:	<mailto:cygwin AT sourceware DOT cygnus DOT com>
List-Help:	<mailto:cygwin-help AT sourceware DOT cygnus DOT com>, <http://sourceware.cygnus.com/ml/#faqs>
Sender:	cygwin-owner AT sourceware DOT cygnus DOT com
Delivered-To:	mailing list cygwin AT sourceware DOT cygnus DOT com
From:	N8TM AT aol DOT com
Message-ID:	<0.cf8a4e43.25406ea9@aol.com>
Date:	Thu, 21 Oct 1999 09:27:05 EDT
Subject:	libm bugs
To:	cygwin AT sourceware DOT cygnus DOT com
MIME-Version:	1.0
X-Mailer:	AOL 4.0 for Windows sub 15

I've been testing the built-in math library in cygwin-b20.1 along with my 
port of the mathinline.h from glibc-2.1. 
(ftp://members.aol.com/n8tm/llibm.tgz).  Aside from slowness, the standard 
library gives fairly good results.  I mention a few exceptions:

asin[fd] produce old-fashioned 0. results rather than NaN when argument is 
out of range
asinf() does not carry full precision in range (.75,1)
expf() overflows to Inf for arguments <= log(FLT_MAX)
log[fd](-2) = -Inf (should be NaN)

and, of course, long doubles outside the range of doubles are formatted as 
Inf by the sprintf() functions.

Tim
tprince AT computer DOT org

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe AT sourceware DOT cygnus DOT com

- Raw text -