Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!uunet!tdatirv!sarima
From: sarima@tdatirv.UUCP (Stanley Friesen)
Newsgroups: comp.lang.c
Subject: Re: Here's a challenge for floating point lovers.
Message-ID: <101@tdatirv.UUCP>
Date: 26 Jan 91 00:06:50 GMT
References: <1991Jan22.180927.29232@zoo.toronto.edu> <14938@smoke.brl.mil> <16035@crdgw1.crd.ge.com> <14964@smoke.brl.mil>
Reply-To: sarima@tdatirv.UUCP (Stanley Friesen)
Organization: Teradata Corp., Irvine
Lines: 17

In article <14964@smoke.brl.mil> gwyn@smoke.brl.mil (Doug Gwyn) writes:
>Oops, I missed that; thanks for the correction.  FLT_RADIX then is okay
>for use in #if expressions, but the other <float.h> macros aren't, not
>even FLT_ROUNDS.

To emphasize this point, I had occasion to look at float.h for SysVr4
a bit ago and it defined FLT_ROUNDS as follows:

#define FLT_ROUNDS (__flt_rounds)

(At least if __STDC__ is defined as 1)

So, there exists an ANSI compliant C compiler that in fact
makes use of this permission for a non-constant float expression.
-- 
---------------
uunet!tdatirv!sarima				(Stanley Friesen)