From owner-freebsd-current  Sun Sep 10 12:03:15 1995
Return-Path: current-owner
Received: (from majordom@localhost)
          by freefall.freebsd.org (8.6.12/8.6.6) id MAA16676
          for current-outgoing; Sun, 10 Sep 1995 12:03:15 -0700
Received: from phaeton.artisoft.com (phaeton.Artisoft.COM [198.17.250.211])
          by freefall.freebsd.org (8.6.12/8.6.6) with ESMTP id MAA16670
          for <current@freebsd.org>; Sun, 10 Sep 1995 12:03:14 -0700
Received: (from terry@localhost) by phaeton.artisoft.com (8.6.11/8.6.9) id LAA15026; Sun, 10 Sep 1995 11:53:31 -0700
From: Terry Lambert <terry@lambert.org>
Message-Id: <199509101853.LAA15026@phaeton.artisoft.com>
Subject: Re: 20-40% faster wc(1)
To: bde@zeta.org.au (Bruce Evans)
Date: Sun, 10 Sep 1995 11:53:31 -0700 (MST)
Cc: pst@shockwave.com, wosch@freebsd.first.gmd.de, current@freebsd.org
In-Reply-To: <199509100758.RAA27567@godzilla.zeta.org.au> from "Bruce Evans" at Sep 10, 95 05:58:09 pm
X-Mailer: ELM [version 2.4 PL24]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 580       
Sender: current-owner@freebsd.org
Precedence: bulk

> 
> >I don't understand... the original isspace() was just a lookup in a table
> >followed by an 'and'.  I assume this changed because of the rune code?
> 
> The rune code has 2 or 3 more tests.
> 
> >The question you should be asking yourself is WHY is isspace(3) slow?
> 
> Because gcc does a poor job of optimizing away the extra tests.

[ ... GCC code generation analysis ... ]

Bruce, you should send this to the GCC people, *really*.


					Terry Lambert
					terry@lambert.org
---
Any opinions in this posting are my own and not those of my present
or previous employers.