From owner-freebsd-commit  Sat Jan  6 17:06:15 1996
Return-Path: owner-commit
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.3/8.7.3) id RAA03778
          for freebsd-commit-outgoing; Sat, 6 Jan 1996 17:06:15 -0800 (PST)
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.3/8.7.3) id RAA03767
          for cvs-all-outgoing; Sat, 6 Jan 1996 17:06:10 -0800 (PST)
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.3/8.7.3) id RAA03756
          for cvs-user-outgoing; Sat, 6 Jan 1996 17:06:08 -0800 (PST)
Received: from jhome.DIALix.COM (root@jhome.DIALix.COM [192.203.228.69])
          by freefall.freebsd.org (8.7.3/8.7.3) with ESMTP id RAA03751
          Sat, 6 Jan 1996 17:06:01 -0800 (PST)
Received: (from peter@localhost) by jhome.DIALix.COM (8.7.3/8.7.3) id JAA05651; Sun, 7 Jan 1996 09:05:41 +0800 (WST)
Date: Sun, 7 Jan 1996 09:05:40 +0800 (WST)
From: Peter Wemm <peter@jhome.DIALix.COM>
To: =?KOI8-R?Q?=E1=CE=C4=D2=C5=CA_=FE=C5=D2=CE=CF=D7?= <ache@astral.msk.su>
cc: CVS-committers@freefall.freebsd.org, cvs-user@freefall.freebsd.org
Subject: Re: cvs commit: src/lkm Makefile
In-Reply-To: <hCifmxmOH1@ache.dialup.ru>
Message-ID: <Pine.BSF.3.91.960107085707.771G-100000@jhome.DIALix.COM>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Sender: owner-commit@FreeBSD.ORG
Precedence: bulk

On Sun, 7 Jan 1996, =?KOI8-R?Q?=E1=CE=C4=D2=C5=CA_=FE=C5=D2=CE=CF=D7?= wrote:
> >  Detach fpu and gnufpu from Makefile..  These lkm's build (or used to)
> >  but are not quite useable.  (fpu is used by init, and process switch
> >  apparently..)
> 
> When they are enabled, people are sure that they at least compiles
> when yet one kernel change occurse. If you disable them, chance
> to detect compile failures becomes very little...
> What wrong happens if they are enabled/installed even not functional?

In a nutshell:
- if you needed them, you can't boot to load them. (can't switch processes)
- the gnufpu module isn't exactly quick to compile.
- the code compilability is already tested as part of the LINT kernel.
- we need some pretty major work before we even have a chance that it'll 
work, and by then they'll probably need look quite different to what they 
are now. 

Besides, I hate waiting for something compile that can't yet be used.. :-)

Cheers,
-Peter