From owner-freebsd-hackers  Tue Jan 14 13:35:48 1997
Return-Path: <owner-hackers>
Received: (from root@localhost)
          by freefall.freebsd.org (8.8.4/8.8.4) id NAA09723
          for hackers-outgoing; Tue, 14 Jan 1997 13:35:48 -0800 (PST)
Received: from who.cdrom.com (who.cdrom.com [204.216.27.3])
          by freefall.freebsd.org (8.8.4/8.8.4) with ESMTP id NAA09718
          for <hackers@freebsd.org>; Tue, 14 Jan 1997 13:35:47 -0800 (PST)
Received: from basil.cdt.luth.se (root@basil.cdt.luth.se [130.240.64.67])
          by who.cdrom.com (8.7.5/8.6.11) with ESMTP id NAA26882
          for <hackers@freebsd.org>; Tue, 14 Jan 1997 13:35:33 -0800 (PST)
Received: from garlic.cdt.luth.se (root@garlic.cdt.luth.se [130.240.64.62]) by basil.cdt.luth.se (8.7.5/8.7.3) with ESMTP id WAA00915; Tue, 14 Jan 1997 22:28:31 +0100 (MET)
Received: from garlic.cdt.luth.se (dot@localhost [127.0.0.1]) by garlic.cdt.luth.se (8.6.12/8.6.12) with ESMTP id WAA14508; Tue, 14 Jan 1997 22:20:48 +0100
Message-Id: <199701142120.WAA14508@garlic.cdt.luth.se>
X-Mailer: exmh version 2.0beta 12/23/96
To: Joao Carlos Mendes Luis <jonny@mailhost.coppe.ufrj.br>
cc: andrew@ugh.net.au, hackers@freebsd.org, dot@cdt.luth.se
Subject: Re: Compiling kernel with optimisation 
In-reply-to: Your message of "Sun, 12 Jan 1997 20:31:17 -0200."
             <199701122231.UAA22599@gaia.coppe.ufrj.br> 
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Date: Tue, 14 Jan 1997 22:20:47 +0100
From: Magnus Enbom <dot@cdt.luth.se>
Sender: owner-hackers@freebsd.org
X-Loop: FreeBSD.org
Precedence: bulk

> #define quoting(David Greenman)
> // >I just noticed that when compiling a kernel it is done with the -O flag.
> // >Would there be much speed improvement in the sytem if it was done with
> // >-O3? Would this break the kernel or is the added time it takes to compile
> // >not worth the benfits?
> // 
> //    It has very little effect on performance and optimizations levels > "-O"
> // have traditionally been broken in gcc.
> 
> Well, the NetBSD team has managed to compile their kernel with -O6
> and -Wall, but they had to change lots of things..
> 
> I don't know what do you call "little effect on performance", but
> 5% gain would be enough to make me think about.  Unfortunately,
> I don't have any measurements (and even don't know how to do them).
> 
> I tried -O6 once, and 2 files were broken in compile time !!! (Wow)
> I did dare once more, fixing those files and running that kernel.
> Panic in 5 seconds.  :(

I've built kernels with -O2 -m486 and used them without problems.
If you're using anything higher you get the -finline-functions option added to 
gcc, that will break a few things. 

/.