[comp.std.c] Must main return a value?

walterm@hpwrce.HP.COM (Walter Murray) (06/27/91)

Is the following program strictly conforming?

   int main(void) { return; }

Many people have written that the main function must return a value.
I can't find such a requirement in the Standard.  I have read 2.1.2.2.3
and 3.6.6.4.  Am I missing something?

Walter Murray

michi@ptcburp.ptcbu.oz.au (Michael Henning) (06/28/91)

walterm@hpwrce.HP.COM (Walter Murray) writes:

>Is the following program strictly conforming?

>   int main(void) { return; }

>Many people have written that the main function must return a value.
>I can't find such a requirement in the Standard.  I have read 2.1.2.2.3
>and 3.6.6.4.  Am I missing something?

Section 2.1.2.2.3 (Program termination) is quite explicit:

"A return from the initial call to the main function is equivalent to
calling the exit function with the value returned by the main function as
its argument. If the main function executes a return that specifies no
value, the termination status returned to the host environment is undefined."

Section 3.6.6.4 (The return Statement) says:

"If a return statement without an expression is executed, and the value of
the function call is used by the caller, the behavior is undefined."

Note that for section 3.6.6.4 to apply, there must be a caller, but the
caller in this case is the environment, so 2.1.2.2.3 applies instead.

The upshot is that the program is strictly conforming, but the exit
status the OS gets to see is undefined.

To write truly portable code, it pays to explicitely call exit() since
some (non-conforming) implementations to not return a defined value
to the OS when a return from main is executed (even if that return
specifies a value). So it pays to write

int main(void) { exit(0); }

to be sure. Unfortunately, many versions of lint will complain about
an undefined return value from main in this case, because they do not
know about exit(). If you want to be paranoid, you can write

int main(void)
{
	exit(0);
	return 0;	/* Keep lint happy */
}

						Michi.
-- 
      -m------- Michael Henning			+61 75 950255
    ---mmm----- Pyramid Technology		+61 75 522475 FAX
  -----mmmmm--- Research Park, Bond University	michi@ptcburp.ptcbu.oz.au
-------mmmmmmm- Gold Coast, Q 4229, AUSTRALIA	uunet!munnari!ptcburp.oz!michi

diamond@jit533.swstokyo.dec.com (Norman Diamond) (06/28/91)

In article <7830001@hpwrce.HP.COM> walterm@hpwrce.HP.COM (Walter Murray) writes:
>Is the following program strictly conforming?
>   int main(void) { return; }
>Many people have written that the main function must return a value.
>I can't find such a requirement in the Standard.  I have read 2.1.2.2.3
>and 3.6.6.4.  Am I missing something?

For the effect of the program to be defined, the main function must return
a value.  It does seem strange that the compiler is required to issue a
diagnostic for a return statement with a value in a void function, but not
for the opposite mistake.

OK, if you don't want your program to send a small nuclear device at your car,
the main function must return a value.  (2.1.2.2 neither allows nor disallows
this extreme an effect; 3.6.6.4 allows it.)
--
Norman Diamond       diamond@tkov50.enet.dec.com
If this were the company's opinion, I wouldn't be allowed to post it.
Permission is granted to feel this signature, but not to look at it.

hooverm@sysjj.mdcbbs.com (SQUID 6 on the DARK side) (06/28/91)

In article <1991Jun28.043342.27540@ptcburp.ptcbu.oz.au>, michi@ptcburp.ptcbu.oz.au (Michael Henning) writes:
> walterm@hpwrce.HP.COM (Walter Murray) writes:
> 
>>Is the following program strictly conforming?
> 
>>   int main(void) { return; }
> 
>>Many people have written that the main function must return a value.
>>I can't find such a requirement in the Standard.  I have read 2.1.2.2.3
>>and 3.6.6.4.  Am I missing something?
>
>...stuff deleted...
>
> some (non-conforming) implementations to not return a defined value
> to the OS when a return from main is executed (even if that return
> specifies a value). So it pays to write
> 
> int main(void) { exit(0); }
> 
> to be sure. Unfortunately, many versions of lint will complain about
> an undefined return value from main in this case, because they do not
> know about exit(). If you want to be paranoid, you can write
> 
> int main(void)
> {
> 	exit(0);
> 	return 0;	/* Keep lint happy */
> }

	I've run into this problem running code on 3 platforms (VMS,SUN & PC)

I use:
	void main()
	{
	insert magic code here
	}

This seems to work well on all three platforms.


			Mark  <o===6

kremer@cs.odu.edu (Lloyd Kremer) (06/28/91)

In article <1991Jun28.043342.27540@ptcburp.ptcbu.oz.au> michi@ptcburp.ptcbu.oz.au (Michael Henning) writes:
>If you want to be paranoid, you can write
>
>int main(void)
>{
>	exit(0);
>	return 0;	/* Keep lint happy */
>}

int main(void)
{
	exit(0);
	/*NOTREACHED*/	/* Keeps most lints happy without extra code */
}

					Lloyd Kremer
					Hilton Systems, Inc.
					kremer@cs.odu.edu

exspes@gdr.bath.ac.uk (P E Smee) (06/28/91)

In article <7830001@hpwrce.HP.COM> walterm@hpwrce.HP.COM (Walter Murray) writes:
>Is the following program strictly conforming?
>
>   int main(void) { return; }
>
>Many people have written that the main function must return a value.
>I can't find such a requirement in the Standard.  I have read 2.1.2.2.3
>and 3.6.6.4.  Am I missing something?

Yep.  The construct you used, 'int main (...' declares main to be a
function which returns an int.  You have declared it as returning a
value, and then not returned a value of the type specified.  Speaking
pedantically, that's a prima facie violation.

In practice, since return from main() generally throws you out of your
C-program environment, and back into your native operating system
environment, whether or not it makes any difference depends on your
OS.  Many OSes expect progs to return status codes -- especially progs
which have been declared to return status codes.  If you don't return
one, the results will be OS-dependent but can range the full gamut from
unpleasantly surprising to unpleasantly mysterious.

If you're really unlucky, your failure to return the value you promised
won't make any difference at all -- until next year, when you've got to
move the prog to another OS platform and have forgotten what you did.

-- 
Paul Smee, Computing Service, University of Bristol, Bristol BS8 1UD, UK
 P.Smee@bristol.ac.uk - ..!uunet!ukc!bsmail!p.smee - Tel +44 272 303132

steve@taumet.com (Stephen Clamage) (06/28/91)

michi@ptcburp.ptcbu.oz.au (Michael Henning) writes:

>The upshot is that the program is strictly conforming, but the exit
>status the OS gets to see is undefined.

No.  A strictly-conforming program "shall not produce output dependent
on any unspecified, undefined, or implementation-defined behavior"
(section 1.7).

>So it pays to write
>int main(void) { exit(0); }

Except that EXIT_SUCCESS need not have the value 0.  This might be
wrong in an ANSI environment.
-- 

Steve Clamage, TauMetric Corp, steve@taumet.com

eggert@twinsun.com (Paul Eggert) (06/29/91)

steve@taumet.com (Stephen Clamage) writes:

>michi@ptcburp.ptcbu.oz.au (Michael Henning) writes:
>>So it pays to write
>>int main(void) { exit(0); }

>Except that EXIT_SUCCESS need not have the value 0.  This might be
>wrong in an ANSI environment.

But section 4.10.4.3 says that exit(0) and exit(EXIT_SUCCESS) have the
same effect even if EXIT_SUCCESS!=0.

Anyway, Posix 1003.1-1990 (section 8.1) says that EXIT_SUCCESS==0,
so hosts where EXIT_SUCCESS!=0 are an endangered species.