From owner-freebsd-current  Wed Mar 27 13:41:00 1996
Return-Path: owner-current
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.3/8.7.3) id NAA00713
          for current-outgoing; Wed, 27 Mar 1996 13:41:00 -0800 (PST)
Received: from phaeton.artisoft.com (phaeton.Artisoft.COM [198.17.250.211])
          by freefall.freebsd.org (8.7.3/8.7.3) with SMTP id NAA00694
          for <freebsd-current@FreeBSD.org>; Wed, 27 Mar 1996 13:40:56 -0800 (PST)
Received: (from terry@localhost) by phaeton.artisoft.com (8.6.11/8.6.9) id OAA02042; Wed, 27 Mar 1996 14:38:42 -0700
From: Terry Lambert <terry@lambert.org>
Message-Id: <199603272138.OAA02042@phaeton.artisoft.com>
Subject: Re: lint
To: bde@zeta.org.au (Bruce Evans)
Date: Wed, 27 Mar 1996 14:38:41 -0700 (MST)
Cc: bde@zeta.org.au, terry@lambert.org, freebsd-current@FreeBSD.org,
        j@uriah.heep.sax.de
In-Reply-To: <199603272038.HAA20036@godzilla.zeta.org.au> from "Bruce Evans" at Mar 28, 96 07:38:51 am
X-Mailer: ELM [version 2.4 PL24]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Sender: owner-current@FreeBSD.org
X-Loop: FreeBSD.org
Precedence: bulk

> >> It doesn't understand them even with -g.  I think the author plans to
> >> fix this.  There has to be a way to tell lint that certain functions
> >> don't return, and __attribute__((__noreturn__)) is a good way.
> 
> >Putting "/* NOTREACHED*/" after the function call is the traditional
> >lint method... assuming this is supposed to be an implementation of
> >a traditional lint, instead of some new thing.
> 
> Not according to lint man pages:
> 
> "BUGS
> The routines exit(2) [sic] ... and other functions that do not return are
> not understood; this causes various incorrect diagnostics".
> 
> /*NOTREACHED*/ only works for the function call, not for the function
> declaration.  You have to put the comment after thousands of calls to
> exit() instead of after one declaration of exit.  If /*NOTREACHED*/
> worked to qualify function declarations, then you might be able to
> implement it using __dead, __dead2 or __attribute__((__noreturn__)).
> E.g.,
> 
> #define __dead		/*NOTREACHED*/	/* probably misplaced */
> #define __dead2		/*NOTREACHED*/	/* more likely to work */

Hmmmmmmm.... that's interesting, actually.

But if that worked, then you could make "exit" intrinsically work
without the NOTREACHED (or the __dead or __dead2).

I may have to take a look at the lexer to see if it's possible to
add state for exit... if someone gets to it before I do (I expect
it'll be about 1998 before I do anything on it 8-)), then so much
the better.

Could you:

#define	exit( x)	lint_exit(x);		\
			/* NOTREACHED*/

And make sure the lint lib defined lint_exit, and not exit?

This wouldn't fix "return" from main, unfortunately...


					Terry Lambert
					terry@lambert.org
---
Any opinions in this posting are my own and not those of my present
or previous employers.