From owner-freebsd-smp  Fri Dec 15 14:51:38 2000
From owner-freebsd-smp@FreeBSD.ORG  Fri Dec 15 14:51:34 2000
Return-Path: <owner-freebsd-smp@FreeBSD.ORG>
Delivered-To: freebsd-smp@freebsd.org
Received: from mail.interware.hu (mail.interware.hu [195.70.32.130])
	by hub.freebsd.org (Postfix) with ESMTP
	id 9577237B699; Fri, 15 Dec 2000 14:51:33 -0800 (PST)
Received: from luanda-49.budapest.interware.hu ([195.70.51.49] helo=elischer.org)
	by mail.interware.hu with esmtp (Exim 3.16 #1 (Debian))
	id 1473hL-00064E-00; Fri, 15 Dec 2000 23:51:32 +0100
Sender: julian@FreeBSD.ORG
Message-ID: <3A3AA04B.68EEB813@elischer.org>
Date: Fri, 15 Dec 2000 14:50:51 -0800
From: Julian Elischer <julian@elischer.org>
X-Mailer: Mozilla 4.7 [en] (X11; U; FreeBSD 5.0-CURRENT i386)
X-Accept-Language: en, hu
MIME-Version: 1.0
To: John Baldwin <jhb@FreeBSD.org>
Cc: Alfred Perlstein <bright@wintelcom.net>, smp@FreeBSD.org
Subject: Re: atomic increment?
References: <XFMail.001215110205.jhb@FreeBSD.org>
Content-Type: text/plain; charset=iso-8859-15
Content-Transfer-Encoding: 7bit
Sender: owner-freebsd-smp@FreeBSD.ORG
Precedence: bulk
X-Loop: FreeBSD.org

John Baldwin wrote:
> 
> On 15-Dec-00 Alfred Perlstein wrote:
> > * John Baldwin <jhb@FreeBSD.ORG> [001215 10:51] wrote:
> >>
> >> On 15-Dec-00 Julian Elischer wrote:
> >> > CAn we have an atomic increment and decrement primative?
> >> >
> >> > presently we get:
> >> >
> >> > .L565:
> >> >         movl $1,%eax
> >> >#APP
> >> >         lock
> >> >         addl %eax,4(%ebx)
> >> >
> >> >
> >> > the movl is totally useless and it would be
> >> > an absolutly trivial addition..
> >> > the question is;
> >> > is there a religious reason we don't already have it?
> >>
> >> man atomic
> >
> > I think he's looking for a useable and intuative interface.
> >
> > no offence I hope. :)
> 
> I didn't come up with it. :)  atomic_add_int(&my_int_var, 1) doesn't appear too
> weird to me.  Of course, the movl being complained about is actually rather
> rediculous.  It's in the instruction cache, so the '1' is a cache hit, and this
> instruction is probably all of 1 clock.  Compared to the overhead of the bus
> lock, that is completely lost in the noise.  Also, staring at the microcosm of
> the asm on one machine is not necessarily going to apply to other machines at
> all.

the mov uses up a register which in one example I was using here, forced 
a much needed number to be written to memory and later restored.
That caused

1/ Extra delay as some percentage of the read-ahead cache was used
by the move. I'm sure it was not a lot but it probably pused something else
back a bit maing it slightly less likely to be in by the time it was needed.

2/ Cluttering of the registers.. which forced:

3/ extra bus cycles as the processor wrote values to memory (for no real reason)

> 
> --
> 
> John Baldwin <jhb@FreeBSD.org> -- http://www.FreeBSD.org/~jhb/
> PGP Key: http://www.baldwin.cx/~john/pgpkey.asc
> "Power Users Use the Power to Serve!"  -  http://www.FreeBSD.org/
> 
> To Unsubscribe: send mail to majordomo@FreeBSD.org
> with "unsubscribe freebsd-smp" in the body of the message

-- 
      __--_|\  Julian Elischer
     /       \ julian@elischer.org
    (   OZ    ) World tour 2000
---> X_.---._/  presently in:  Budapest
            v


To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-smp" in the body of the message