X-Authentication-Warning: delorie.com: mailnull set sender to djgpp-bounces using -f
From: Kelsey Bjarnason <kelseyb AT xxspamxx DOT telus DOT net>
Subject: Re: gcc 2.953 output - is this a bug?
Newsgroups: comp.os.msdos.djgpp,comp.lang.c
Message-ID: <pan.2002.02.12.01.01.51.570132.27481@xxspamxx.telus.net>
References: <3C61E6BB DOT 7F9AD526 AT yahoo DOT com> <87d6zim2vy DOT fsf AT pfaff DOT stanford DOT edu> <3C620456 DOT 967C433F AT yahoo DOT com>
User-Agent: Pan/0.11.1.91 (Unix)
Mime-Version: 1.0
Content-Type: text/plain; charset=iso-8859-1
Content-Transfer-Encoding: 8bit
X-Comment-To: "CBFalconer" <cbfalconer AT yahoo DOT com>
Lines: 17
Date: Tue, 12 Feb 2002 09:05:01 GMT
NNTP-Posting-Host: 216.232.62.170
X-Trace: news0.telusplanet.net 1013504701 216.232.62.170 (Tue, 12 Feb 2002 02:05:01 MST)
NNTP-Posting-Date: Tue, 12 Feb 2002 02:05:01 MST
To: djgpp AT delorie DOT com
DJ-Gateway: from newsgroup comp.os.msdos.djgpp
Reply-To: djgpp AT delorie DOT com

[snips]

On Wed, 06 Feb 2002 20:51:47 -0800, CBFalconer wrote:

> Thanks.  If I had a handy x86 manual I *might* have eventually figured
> it out. Incidentally, I found that gcc generated better code for:
> 
>               h = h * 31UL + *string++;
> than for:
>               h = (h << 5) - h;
>               h += *string++;
> 
> the latter used an additional temporary.  Which proves something or
> other.
 
That once again, "Optimize your algorithms, not your source code" is
wisdom to live by? :)