Re: [9fans] C compiler and junk mail

Re: [9fans] C compiler and junk mail

[Russ Cox]

If you add the following to /sys/src/cmd/cc/lex.c,
it will pass -+ through to the invoked cpp:

			av[0] = CPP;
			i = 1;
>>>			if(debug['+']) {
>>>				sprint(opt, "-+");
>>>				av[i++] = strdup(opt);
>>>			}
			for(c = 0; c < ndef; c++) {
				sprint(opt, "-D%s", defs[c]);
				av[i++] = strdup(opt);
			}

Then you can use 8c -p+, etc.  I added this back in
January to solve exactly the problem you describe.
It will make it out for real when we next do a release.

Russ

Re: [9fans] C compiler and junk mail

[dmr]

> If you really care, use 8c -p to invoke a strict ANSI preprocessor.

But this won't deal with // either, except with the -+ option
(which -p doesn't generate).  You need to run cpp in
a pipeline.

	Dennis

Re: [9fans] C compiler and junk mail

[rob pike]

	1)  I sent a mild complaint earlier about how the 2nd edition C
	compiler fails to remove // type comments from #define statements.
	This caused me some grief when importing code into plan 9.
	Is this an over sight in the implementation or leaving these comments
	in place part of the language specifications?

Well, first, those are not ANSI C comments. They may be in C99.
But they were added to the Plan 9 compilers because some people
prefer them.

The real problem here is that the Plan 9 compiler remove comments
at a slightly different point in the processing than the ANSI standard
suggests, so comments on #include and #define lines can cause trouble.
If you really care, use 8c -p to invoke a strict ANSI preprocessor.

-rob

Re: [9fans] C compiler and junk mail

[Scott Schwartz]

> I thought this was a moderated list. Am I wrong?

The usenet->mail direction is moderated, but the other direction isn't.

[9fans] C compiler and junk mail

[ed]

1)  I sent a mild complaint earlier about how the 2nd edition C
compiler fails to remove // type comments from #define statements.
This caused me some grief when importing code into plan 9.
Is this an over sight in the implementation or leaving these comments
in place part of the language specifications?

2)  Occasionally I get junk mail from 9fans, e.g.

	From 9fans-admin@cse.psu.edu  Wed Feb 14 02:20:17 2001
Return-Path: <9fans-admin@cse.psu.edu>
Received: from mail.cse.psu.edu (postfix@psuvax1.cse.psu.edu [130.203.4.6])
        by aspen.cs.unr.edu (8.9.2/8.9.2) with ESMTP id CAA18146
        for <ed@cs.unr.edu>; Wed, 14 Feb 2001 02:20:16 -0800 (PST)
Received: from psuvax1.cse.psu.edu (psuvax1.cse.psu.edu [130.203.4.6])
        by mail.cse.psu.edu (CSE Mail Server) with ESMTP
        id B0E04199D5; Wed, 14 Feb 2001 05:20:11 -0500 (EST)
Received: from g04.syd.iprimus.net.au (g04.syd.iprimus.net.au [203.134.65.6])
        by mail.cse.psu.edu (CSE Mail Server) with SMTP id 68AC8199D5
        for <9fans@cse.psu.edu>; Wed, 14 Feb 2001 05:19:50 -0500 (EST)
Received: (qmail 16404 invoked from network); 14 Feb 2001 10:20:14 -0000
Received: from unknown (HELO kdhbooks.com) (203.134.133.16)
  by g04.syd.iprimus.net.au with SMTP; 14 Feb 2001 10:20:14 -0000
From: "e-Publisher's Weekly" <admin@kdhbooks.com>

...

I thought this was a moderated list. Am I wrong?

ed wishart