Re: c/8506: 3.3 (1104) disallowing omission of parameters in fn def

Re: c/8506: 3.3 (1104) disallowing omission of parameters in fn def

[Christian Ehrhardt]

The following reply was made to PR c/8506; it has been noted by GNATS.

From: "Christian Ehrhardt" <ehrhardt@mathematik.uni-ulm.de>
To: jdonner0@earthlink.net
Cc: gcc-gnats@gcc.gnu.org
Subject: Re: c/8506: 3.3 (1104) disallowing omission of parameters in fn def
Date: Fri, 8 Nov 2002 19:35:28 +0100

 On Fri, Nov 08, 2002 at 05:52:24PM -0000, jdonner0@earthlink.net wrote:
 > gcc now makes it an error to omit parameter names, eg
 > 
 > int foo(int /* note, omitted parameter name*/){}
 > now gives an error.
 > 
 > $ gcc foo.c
 > foo.c: In function `foo':
 > foo.c:1: error: parameter name omitted
 
 This behaviour seems to be mandated by the standard. 6.9.1[#5] states:
    If the declarator includes a parameter type list, the declaration of
    each parameter shall include an identifier, except for the special case
    of a parameter list consisting of a single parameter of type void, in
    which case there shall not be an identifier. No declaration list shall
    follow.
 
    regards   Christian
 
 -- 
 THAT'S ALL FOLKS!

Re: c/8506: 3.3 (1104) disallowing omission of parameters in fn def

[pme]

Synopsis: 3.3 (1104) disallowing omission of parameters in fn def

State-Changed-From-To: open->closed
State-Changed-By: pme
State-Changed-When: Fri Nov  8 11:59:56 2002
State-Changed-Why:
    User error, communicated privately.

http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&database=gcc&pr=8506

Re: c/8506: 3.3 (1104) disallowing omission of parameters in fn def

[Richard Earnshaw]

The following reply was made to PR c/8506; it has been noted by GNATS.

From: Richard Earnshaw <rearnsha@arm.com>
To: jdonner0@earthlink.net
Cc: gcc-gnats@gcc.gnu.org, Richard.Earnshaw@arm.com
Subject: Re: c/8506: 3.3 (1104) disallowing omission of parameters in fn 
 def
Date: Fri, 08 Nov 2002 18:13:13 +0000

 This is a C++ism.  It's not part of C, and I don't believe GCC has ever 
 allowed this.
 
 R.
 

Re: c/8506: 3.3 (1104) disallowing omission of parameters in fn def

[Phil Edwards]

The following reply was made to PR c/8506; it has been noted by GNATS.

From: Phil Edwards <phil@jaj.com>
To: jdonner0@earthlink.net
Cc: gcc-gnats@gcc.gnu.org
Subject: Re: c/8506: 3.3 (1104) disallowing omission of parameters in fn def
Date: Fri, 8 Nov 2002 13:02:33 -0500

 On Fri, Nov 08, 2002 at 05:52:24PM -0000, jdonner0@earthlink.net wrote:
 > >Description:
 > gcc now makes it an error to omit parameter names, eg
 > 
 > int foo(int /* note, omitted parameter name*/){}
 > now gives an error.
 > 
 > $ gcc foo.c
 > foo.c: In function `foo':
 > foo.c:1: error: parameter name omitted
 
 It's correct C++, but I don't believe it's correct C.  Try renaming the
 file and running "gcc -c foo.cc" (or .cpp or whatever) instead.
 
 You can always omit the names in declarations, of course, but not
 definitions.
 
 
 Phil
 
 -- 
 I would therefore like to posit that computing's central challenge, viz. "How
 not to make a mess of it," has /not/ been met.
                                                  - Edsger Dijkstra, 1930-2002

c/8506: 3.3 (1104) disallowing omission of parameters in fn def

[jdonner0]

>Number:         8506
>Category:       c
>Synopsis:       3.3 (1104) disallowing omission of parameters in fn def
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    unassigned
>State:          open
>Class:          rejects-legal
>Submitter-Id:   net
>Arrival-Date:   Fri Nov 08 09:56:04 PST 2002
>Closed-Date:
>Last-Modified:
>Originator:     Jeff Donner
>Release:        gcc version 3.3 20021104 (experimental)
>Organization:
>Environment:
cygwin, Win2K
>Description:
gcc now makes it an error to omit parameter names, eg

int foo(int /* note, omitted parameter name*/){}
now gives an error.

$ gcc foo.c
foo.c: In function `foo':
foo.c:1: error: parameter name omitted

I'm not a language lawyer but I know there's scads of code 
that uses this as a clean way to show a fn doesn't 
actually use its parameters.  And I believe it's correct 
usage, many other compilers (VC, Borland at least) allow it. 
This would break a lot of code.  This shouldn't even be 
a warning unless it is specifically asked for. 
>How-To-Repeat:
gcc foo.c
>Fix:

>Release-Note:
>Audit-Trail:
>Unformatted:
----gnatsweb-attachment----
Content-Type: application/octet-stream; name="foo.c"
Content-Transfer-Encoding: base64
Content-Disposition: attachment; filename="foo.c"

aW50IGZvbyhpbnQpe30NCg==