From owner-freebsd-hackers  Thu Apr 11 17:09:33 1996
Return-Path: owner-hackers
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.3/8.7.3) id RAA25889
          for hackers-outgoing; Thu, 11 Apr 1996 17:09:33 -0700 (PDT)
Received: from mail.think.com (Mail1.Think.COM [131.239.33.245])
          by freefall.freebsd.org (8.7.3/8.7.3) with SMTP id RAA25884
          for <hackers@FreeBSD.org>; Thu, 11 Apr 1996 17:09:29 -0700 (PDT)
Received: from Early-Bird-1.Think.COM by mail.think.com; Thu, 11 Apr 96 20:08:52 -0400
Received: from compound.think.com ([206.10.99.158]) by Early-Bird.Think.COM; Thu, 11 Apr 96 20:08:49 EDT
Received: (from alk@localhost) by compound.think.com (8.7.5/8.6.112) id TAA03198; Thu, 11 Apr 1996 19:08:50 -0500 (CDT)
Date: Thu, 11 Apr 1996 19:08:50 -0500 (CDT)
Message-Id: <199604120008.TAA03198@compound.think.com>
From: Tony Kimball <alk@Think.COM>
To: langfod@dihelix.com
Cc: hackers@FreeBSD.org
In-Reply-To: <199604112329.NAA19322@caliban.dihelix.com> (langfod@dihelix.com)
Subject: Re: pgcc kernel compiles automated
Sender: owner-hackers@FreeBSD.org
X-Loop: FreeBSD.org
Precedence: bulk

   Date: Thu, 11 Apr 1996 13:29:13 -1000 (HST)
   From: "David Langford" <langfod@dihelix.com>

   Are you actually using "-mpentium"?  IF not I am not sure that using pgcc
   actually has an advantage :)

; gcc -mpentium -O6 -fomit-frame-pointer t.c -o pentium -lm
; gcc -O6 -fomit-frame-pointer t.c -o generic -lm
; kcc -O6 -fomit-frame-pointer t.c -o shipped -lm
; pentium 10000
1.655895 user
; generic 10000
1.648273 user
; shipped 10000
2.133230 user

As far as I can tell, -mpentium doesn't do anything statistically
significant, in pgcc 2.7.2.p9 anyhow.  I haven't controlled for the
gcc version, though, so for all the evidence to date it could just
as well be attributable to the antiquity of the shipped compiler.
(t.c is a mix of floating point and pointer-chasing.)

//alk