From owner-freebsd-commit  Wed Nov 29 10:54:30 1995
Return-Path: owner-commit
Received: (from root@localhost)
          by freefall.freebsd.org (8.6.12/8.6.6) id KAA10083
          for freebsd-commit-outgoing; Wed, 29 Nov 1995 10:54:30 -0800
Received: (from root@localhost)
          by freefall.freebsd.org (8.6.12/8.6.6) id KAA10057
          for cvs-all-outgoing; Wed, 29 Nov 1995 10:53:01 -0800
Received: (from root@localhost)
          by freefall.freebsd.org (8.6.12/8.6.6) id KAA10043
          for cvs-usrsbin-outgoing; Wed, 29 Nov 1995 10:52:59 -0800
Received: from Root.COM (implode.Root.COM [198.145.90.17])
          by freefall.freebsd.org (8.6.12/8.6.6) with ESMTP id KAA10030
          ; Wed, 29 Nov 1995 10:52:49 -0800
Received: from corbin.Root.COM (corbin [198.145.90.50]) by Root.COM (8.6.12/8.6.5) with ESMTP id KAA01399; Wed, 29 Nov 1995 10:52:42 -0800
Received: from localhost (localhost [127.0.0.1]) by corbin.Root.COM (8.6.12/8.6.5) with SMTP id KAA00310; Wed, 29 Nov 1995 10:51:03 -0800
Message-Id: <199511291851.KAA00310@corbin.Root.COM>
To: "Justin T. Gibbs" <gibbs@freefall.freebsd.org>
cc: CVS-committers@freefall.freebsd.org, cvs-usrsbin@freefall.freebsd.org
Subject: Re: cvs commit: src/usr.sbin/config Makefile main.c 
In-reply-to: Your message of "Wed, 29 Nov 95 10:46:28 PST."
             <199511291846.KAA00293@corbin.Root.COM> 
From: David Greenman <davidg@Root.COM>
Reply-To: davidg@Root.COM
Date: Wed, 29 Nov 1995 10:51:03 -0800
Sender: owner-commit@FreeBSD.ORG
Precedence: bulk

>>>That's impossible to do given the current scheme of 'everything is an
>>>option'.  However, I think a much better solution to the problem is to
>>>have config kick out a 'remember to do a make clean' if the directory
>>>exists.  If the user chooses not to do that, then it's their problem not
>>>ours.
>>
>>Why not make option lines add an entry to a common header file.  Only
>>modules that export an option need to include the "option" header.
>>We could have a separate keyword "define" that gives the current behavior
>>so you can still do it the old quick way for debugging.  I think that
>>a make depend would have much more value if this was done.
>
>   The other idea is to add an option file for each supported option. Some
>people don't like this because it makes adding a new option a little more
>difficult (all the options would have to be listed in an options file, and
                                          ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

   I just realized that this might be confusingly worded... I mean you would
have a options master file /sys/conf/options and /sys/i386/conf/options.i386
that would list all of the options. Config would use this in conjunction with
your kernel config file to produce individual options_foobar.h files (where
foobar is an option) that would/wouldn't have a #define FOOBAR.

>you need to add a #include plus a #ifdef to each file that uses the option).
>The idea is that the option file would have a name like "option_foobar.h" or
>something deterministic which would be autogenerated by config. The file
>would either have the #define or it would be null (consuming no disk space
>and would be relative fast access). This approach allows for fine control
>over the kernel options dependency.
>
>-DG