From owner-cvs-include  Tue Apr  1 02:36:06 1997
Return-Path: <owner-cvs-include>
Received: (from root@localhost)
          by freefall.freebsd.org (8.8.5/8.8.5) id CAA23469
          for cvs-include-outgoing; Tue, 1 Apr 1997 02:36:06 -0800 (PST)
Received: from werple.net.au (melb.werple.net.au [203.9.190.18])
          by freefall.freebsd.org (8.8.5/8.8.5) with SMTP id CAA23438
          for <cvs-include@freefall.freebsd.org>; Tue, 1 Apr 1997 02:35:38 -0800 (PST)
Received: (qmail 22673 invoked by uid 5); 1 Apr 1997 10:34:59 -0000
Received: (from jb@localhost) by freebsd1.cimlogic.com.au (8.7.5/8.7.3) id RAA10278; Tue, 1 Apr 1997 17:59:38 +1000 (EST)
From: John Birrell <jb@cimlogic.com.au>
Message-Id: <199704010759.RAA10278@freebsd1.cimlogic.com.au>
Subject: Re: cvs commit: src/include unistd.h
To: bde@zeta.org.au (Bruce Evans)
Date: Tue, 1 Apr 1997 17:59:37 +1000 (EST)
Cc: bde@zeta.org.au, jb@cimlogic.com.au, cvs-all@freefall.freebsd.org,
        CVS-committers@freefall.freebsd.org, cvs-include@freefall.freebsd.org,
        peter@freefall.freebsd.org, peter@spinner.dialix.com
In-Reply-To: <199704010359.NAA03631@godzilla.zeta.org.au> from Bruce Evans at "Apr 1, 97 01:59:22 pm"
X-Mailer: ELM [version 2.4ME+ PL22 (25)]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Sender: owner-cvs-include@FreeBSD.ORG
X-Loop: FreeBSD.org
Precedence: bulk

Bruce Evans wrote:
> >This doesn't comply with POSIX 1003.1b (and I assume .c). For POSIX 1003.1b,
> >the _POSIX_C_SOURCE feature test is supposed to be compared against 199309L.
> >And for 1003.1c, I guess that it is supposed to be 1996XXL where XX is the
> >month that the update was (finally) published.
> 
> glibc attempts to support POSIX.1b 1993 stuff, but it (glibc-2.0 at
> least) seems to assume that POSIX.1b contains POSIX.2.  Is this wrong?

As far as POSIX.1b is concerned, yes I think it is wrong. 8-)

>From my reading of POSIX 1003.1b, I'd expect to see things like:

#if (!defined(_ANSI_SOURCE) && !defined(_POSIX_SOURCE) &&
    !defined(_POSIX_C_SOURCE)) || (_POSIX_C_SOURCE >= 199309L)
int nanosleep __P((const struct timespec *, struct timespec *));
#endif

and for functions like ttyname_r that first appeared in 1003.1c, the
header needs:

#if (!defined(_ANSI_SOURCE) && !defined(_POSIX_SOURCE) &&
    !defined(_POSIX_C_SOURCE)) || (_POSIX_C_SOURCE >= 1996xxL)
int ttyname_r __P((int, char *, size_t));
#endif

> A POSIX.2 draft rationale says that "each new value [of _POSIX_C_SOURCE
> is expected to reserve] the name space for that new standard or revision,
> plus all earlier POSIX standards".

The rationale in POSIX 1003.1b says:

"Since _POSIX_SOURCE as specified by IEEE Std 1003.1-1990 did not
have a value associated with it, the _POSIX_C_SOURCE macro 
replaces it, allowing an application to inform the system of the
version of the standard to which it conforms."

I thought all the C API stuff was in 1003.1, so where does the
glibc-2.0 get the _POSIX_C_SOURCE >= 2 from?

Maybe I'm confused 'cause I've never seen POSIX.2. 8-)

> 
> Bruce
> 

Regards,

-- 
John Birrell - jb@cimlogic.com.au; jb@netbsd.org; jb@freebsd.org
CIMlogic Pty Ltd, 119 Cecil Street, South Melbourne Vic 3205, Australia
Tel +61 3 9690 6900 Fax +61 3 9690 6650 Mob +61 418 353 137