From owner-cvs-all  Wed Sep  9 11:27:08 1998
Return-Path: <owner-cvs-all>
Received: (from daemon@localhost)
          by hub.freebsd.org (8.8.8/8.8.8) id LAA23749
          for cvs-all-outgoing; Wed, 9 Sep 1998 11:27:08 -0700 (PDT)
          (envelope-from owner-cvs-all)
Received: from rover.village.org (rover.village.org [204.144.255.49])
          by hub.freebsd.org (8.8.8/8.8.8) with SMTP id LAA23726;
          Wed, 9 Sep 1998 11:27:04 -0700 (PDT)
          (envelope-from imp@village.org)
Received: from harmony [10.0.0.6] 
	by rover.village.org with esmtp (Exim 1.71 #1)
	id 0zGox7-0005KF-00; Wed, 9 Sep 1998 12:26:49 -0600
Received: from harmony.village.org (localhost.village.org [127.0.0.1]) by harmony.village.org (8.9.1/8.8.3) with ESMTP id MAA04743; Wed, 9 Sep 1998 12:27:44 -0600 (MDT)
Message-Id: <199809091827.MAA04743@harmony.village.org>
To: Bruce Evans <bde@zeta.org.au>
Subject: Re: cvs commit: src/sys/conf files 
Cc: cvs-all@freebsd.org, cvs-committers@freebsd.org
In-reply-to: Your message of "Thu, 10 Sep 1998 04:15:40 +1000."
		<199809091815.EAA10016@godzilla.zeta.org.au> 
References: <199809091815.EAA10016@godzilla.zeta.org.au>  
Date: Wed, 09 Sep 1998 12:27:44 -0600
From: Warner Losh <imp@village.org>
Sender: owner-cvs-all@freebsd.org
X-Loop: FreeBSD.org
Precedence: bulk

In message <199809091815.EAA10016@godzilla.zeta.org.au> Bruce Evans writes:
: This is fairly bogus and doesn't actually work for `make -jN'.
: First, subr_bus.c doesn't depend on *_if.h.  subr_bus.o depends
: on them.  config(8) converts the dependency.  Second, several
: other .o files depend on *_if.h.  This is handled for the usual
: case by totally disordering the files list to put subr_bus.c
: before the sources for the other object files.  This kludge fails
: completely for `make -jN'.  The problem is handled for application
: makefiles by making all .o files depend on all generated .h files
: when .depend doesn't exist.  See bsd.prog.mk and about 100 log
: messages for application makefiles.

Hmmm, the make -j 5 works for me after a make clean.  I wonder why
that would work if it were completely bogus :-)

I agree that there are other files that depend on these files.  I also
agree that subr_bus.o is the file that actually depends on this file.
Config generates the proper dependency on the .o file.  There are at
least two files that depend on bus_if.h and device_if.h.

If this isn't the place to do this for the kernel, what is the right
thing to do?

Warner