This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
Re: Possible variadic macro bug
- To: Zack Weinberg <zack at codesourcery dot com>
- Subject: Re: Possible variadic macro bug
- From: seebs at plethora dot net (Peter Seebach)
- Date: Fri, 09 Nov 2001 18:02:21 -0600
- cc: Neil Booth <neil at daikokuya dot demon dot co dot uk>, gcc-bugs at gcc dot gnu dot org
- Reply-To: seebs at plethora dot net (Peter Seebach)
In message <20011109155933.B15127@codesourcery.com>, Zack Weinberg writes:
>On Fri, Nov 09, 2001 at 05:07:06PM -0600, Peter Seebach wrote:
>> >It's not correct C99 - C99 requires that arguments be comma separated,
>> >including the "...".
>> Oops! Rereading this, I see what you're saying. I noticed the existance of
>> the (...) form, and didn't think this through correctly. Sorry! Hmm. I'm
>> still convinced we *meant* to make this possible, though. Ugh.
>The syntax you thought worked conflicts with GCC's extension for giving
>the variable arguments a friendly name, which is widely used, so I hope
>the committee didn't mean to do that. :-)
We didn't have any GNU C people available.
>The revised ",##__VA_ARGS__" extension in gcc 3.0 was meant to be suitable
>for standardization (unlike the older version, which was awful and I'm
>embarrassed it survived as long as it did). In fact, if I knew how, I
>would have written up a formal proposal already. Perhaps you can shed
>some light on the procedure?
You contact your national body (ANSI, in the U.S., or the British standards
organization, in the UK) and ask them to submit a Defect Report. Actually,
this looks like a friendly enough extension to me that I might be able to
submit a DR on it and get it accepted as correcting an unfortunate limitation.
I just wish someone had caught it *before* the standard went out. Hmm.
-s