From mail-miner-10529@localhost Wed Dec 18 12:07:55 2002
Received: from mailman.opengroup.org ([192.153.166.9])
by deep-dark-truthful-mirror.pad with smtp (Exim 3.36 #1 (Debian))
id 18Buh5-0006Zr-00
for <posix@simon-cozens.org>; Wed, 13 Nov 2002 10:24:23 +0000
Received: (qmail 1679 invoked by uid 503); 13 Nov 2002 10:10:49 -0000
Resent-Date: 13 Nov 2002 10:10:49 -0000
Date: Wed, 13 Nov 2002 10:06:51 GMT
From: Andrew Josey <ajosey@rdg.opengroup.org>
Message-Id: <1021113100650.ZM12997@skye.rdg.opengroup.org>
In-Reply-To: Joanna Farley's message as of Nov 13, 9:56am.
References:
<200211120937.JAA28130@xoneweb.opengroup.org>
<1021112125524.ZM7503@skye.rdg.opengroup.org>
<3DD221BB.13116D47@sun.com>
X-Mailer: Z-Mail (5.0.0 30July97)
To: austin-group-l@opengroup.org
Subject: Re: Defect in XBD lround
MIME-Version: 1.0
Resent-Message-ID: <gZGK1B.A.uY.iUi09@mailman>
Resent-To: austin-group-l@opengroup.org
Resent-From: austin-group-l@opengroup.org
X-Mailing-List: austin-group-l:archive/latest/4823
X-Loop: austin-group-l@opengroup.org
Precedence: list
Resent-Sender: austin-group-l-request@opengroup.org
Content-Type: text/plain; charset=us-ascii
Joanna, All
Thanks. I got the following response from Fred Tydeman.
C99 Defect Report (DR) 240 covers this. The main body of C99
(7.12.9.7) says range error, while Annex F (F.9.6.7 and F.9.6.5)
says "invalid" (domain error). The result was to change 7.12.9.7
to allow for either range or domain error. The preferred error
is domain error (so as match Annex F). So, no need to change XBD.
regards
Andrew
On Nov 13, 9:56am in "Re: Defect in XBD lr", Joanna Farley wrote:
> Sun's expert in this area after some discussions with a colleague
> outside of Sun concluded that for lround, to align with both C99 and SUS
> changes of the following form were necessary:
>
> + If x is +/-Inf/NaN, a domain error occurs, and
> + errno is set to EDOM in MATH_ERRNO mode;
> + the invalid exception is raised in MATH_ERREXCEPT mode.
> [to align with C99 Annex F.4]
>
> + If x is too large to be represented as a long, a *range* error
> may occur, and
> + errno *may be* set to ERANGE in MATH_ERRNO mode;
> [to align with C99 7.12.9.7]
> + the invalid exception *is* raised in MATH_ERREXCEPT mode.
> [to align with C99 Annex F.4]
>
> They believe it is a bit awkward to have errno set to ERANGE in
> MATH_ERRNO mode yet the invalid exception raised in MAH_ERREXCEPT mode,
> but that just reflects an imperfect mapping of the C notion of errno to
> the IEEE 754 notion of data conversion.
>
> I'll work with our expert to draft text refecting the above to suggest
> replacement text for lines 23678-23684 on lround page 721 of XSH6.
>
> Thanks
>
> Joanna
>
>
> Andrew Josey wrote:
> >
> > The text referred to is MX shaded and part of the ISO 60559 floating
> > point option. I do not think changing the Domain Error to a Range Error
> > is the fix or at least not the fix for the NaN and +-Inf cases. ISO C
> > 99 describes the range error case if the magnitude of x is too large as a
> > may fail. I'll ask Fred T for his thoughts on this one...
> > regards
> > Andrew
> >
> > On Nov 12, 9:37am in "Defect in XBD lround", Erwin.Unruh@fujitsu-siemens.com wrote:
> > > Defect report from : Erwin Unruh , Fujitsu Siemens Computers
> > >
> > > (Please direct followup comments direct to austin-group-l@opengroup.org)
> > >
> > > @ page 0 line 0 section lround objection {0}
> > >
> > > Problem:
> > >
> > > Defect code : 1. Error
> > >
> > > The function lround is described in http://www.opengroup.org/onlinepubs/007904975/functions/lround.html
> > > On Error it is specified that errno has to be set to EDOM. However, the C99 standard ISO/IEC 9899:1999 (E) specifies this as a range error, which would result in a value of ERANGE. So an implementation could not be conformant to both these standards.
> > >
> > > Action:
> > >
> > > Change the value of errno to ERANGE, if the result is not represantable. More specific: In the description of the function, replace all occurences of "domain error" with "range error" and replace "EDOM" with "ERANGE"
> > >-- End of excerpt from Erwin.Unruh@fujitsu-siemens.com
> >
-----
Andrew Josey The Open Group
Austin Group Chair Apex Plaza,Forbury Road,
Email: a.josey@opengroup.org Reading,Berks.RG1 1AX,England
Tel: +44 118 9508311 ext 2250 Fax: +44 118 9500110