From owner-freebsd-hackers  Fri Aug 23 09:54:34 1996
Return-Path: owner-hackers
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.5/8.7.3) id JAA10290
          for hackers-outgoing; Fri, 23 Aug 1996 09:54:34 -0700 (PDT)
Received: from rocky.mt.sri.com (rocky.mt.sri.com [206.127.76.100])
          by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id JAA10281;
          Fri, 23 Aug 1996 09:54:27 -0700 (PDT)
Received: (from nate@localhost) by rocky.mt.sri.com (8.7.5/8.7.3) id KAA06985; Fri, 23 Aug 1996 10:15:01 -0600 (MDT)
Date: Fri, 23 Aug 1996 10:15:01 -0600 (MDT)
Message-Id: <199608231615.KAA06985@rocky.mt.sri.com>
From: Nate Williams <nate@mt.sri.com>
To: "Jordan K. Hubbard" <jkh@time.cdrom.com>
Cc: sos@freebsd.org, hackers@freebsd.org
Subject: Re: Am I wrong or is this just stupid? 
In-Reply-To: <20809.840815838@time.cdrom.com>
References: <199608230836.KAA17678@ra.dkuug.dk>
	<20809.840815838@time.cdrom.com>
Sender: owner-hackers@freebsd.org
X-Loop: FreeBSD.org
Precedence: bulk

Jordan K. Hubbard writes:
> > Hmm, I guess its because we wan't to build our tools etc, with the
> > newest version.. This would resolve some (but not all) bootstrap
> > problems in the tools...
> 
> But you would anyway.  Once you've done the bootstrap, the new tools
> are installed and leaving the binaries in the tree won't change a
> thing - everything subsequent will still be linked with the new copies.

But they won't be re-built with the new 'build-tools', which may have
bugs in them such as the -O2 -fno-strength-reduce bugs fixed a while
back.  So, the lib-tools will be buggy if the build-tools are buggy.

> Like I said, I think there's no logical reason at all for this.

There *is* a logical reason, but it's not always necessary when the
build-tools don't change.  But, generally speaking they do change
between releases, although they haven't since 2.0.5, but will change in
2.2 I suspect, so it will be relevant there.

> Furthermore, here's my make world time for the standard tree:
> 
>      6366.76 real      4213.03 user       711.55 sys
> 
> And for a tree which has make depend and make all combined:
> 
>      6189.28 real      4215.11 user       706.74 sys
> 

< 3% isn't much for the possibility of getting buggy lib-tools, which
could theoretically affect things adversely.


Nate