From owner-cvs-sys  Mon May 27 01:38:50 1996
Return-Path: owner-cvs-sys
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.5/8.7.3) id BAA24794
          for cvs-sys-outgoing; Mon, 27 May 1996 01:38:50 -0700 (PDT)
Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19])
          by freefall.freebsd.org (8.7.5/8.7.3) with SMTP id BAA24788;
          Mon, 27 May 1996 01:38:43 -0700 (PDT)
Received: (from bde@localhost) by godzilla.zeta.org.au (8.6.12/8.6.9) id SAA28879; Mon, 27 May 1996 18:33:38 +1000
Date: Mon, 27 May 1996 18:33:38 +1000
From: Bruce Evans <bde@zeta.org.au>
Message-Id: <199605270833.SAA28879@godzilla.zeta.org.au>
To: bde@zeta.org.au, coredump@nervosa.com
Subject: Re: cvs commit: src/sys/i386/isa syscons.c
Cc: CVS-committers@freefall.freebsd.org, cvs-all@freefall.freebsd.org,
        cvs-sys@freefall.freebsd.org, peter@freefall.freebsd.org
Sender: owner-cvs-sys@FreeBSD.ORG
X-Loop: FreeBSD.org
Precedence: bulk

>> How much larger and slower does the extra inlining make syscons? ;-)
>> Bruce

>I agree that inlining would definetly make it larger; but slower?

It reduces locality of reference and may displace more important code
from the cache.  A single cache miss costs about the same time as a
function call.

Inlining may also interfere with optimization and result in worse code
generation for more important surrounding code.  This problem is
particularly noticeable on the i386 because there aren't enough
registers.

Bruce