From owner-freebsd-alpha  Mon Feb 18  1:30:52 2002
Delivered-To: freebsd-alpha@freebsd.org
Received: from iona.dcs.gla.ac.uk (iona.dcs.gla.ac.uk [130.209.240.35])
	by hub.freebsd.org (Postfix) with ESMTP id 909B137B402
	for <freebsd-alpha@freebsd.org>; Mon, 18 Feb 2002 01:30:49 -0800 (PST)
Received: from therese.dcs.gla.ac.uk ([130.209.241.134] helo=therese.dcs.gla.ac.uk.dcs.gla.ac.uk)
	by iona.dcs.gla.ac.uk with esmtp (Exim 3.13 #1)
	id 16ck8G-0006Y5-00; Mon, 18 Feb 2002 09:30:48 +0000
Received: by therese.dcs.gla.ac.uk.dcs.gla.ac.uk (8.11.6/Dumb)
	id g1I9Uk626041; Mon, 18 Feb 2002 09:30:46 GMT
To: Andrew Gallatin <gallatin@cs.duke.edu>
Cc: freebsd-alpha@freebsd.org
Subject: Re: cvs commit: src/sys/alpha/include cpufunc.h
References: <200202120120.g1C1Ko136446@freefall.freebsd.org>
	<20020211204721.A49383@grasshopper.cs.duke.edu>
	<ysqr8nkj6is.fsf@therese.dcs.gla.ac.uk>
	<15472.2563.368794.540600@grasshopper.cs.duke.edu>
From: Rolf Neugebauer <neugebar@dcs.gla.ac.uk>
Date: 18 Feb 2002 09:30:45 +0000
In-Reply-To: <15472.2563.368794.540600@grasshopper.cs.duke.edu>
Message-ID: <ysqn0y7jfwa.fsf@therese.dcs.gla.ac.uk>
Lines: 27
User-Agent: Gnus/5.0808 (Gnus v5.8.8) Emacs/20.7
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: owner-freebsd-alpha@FreeBSD.ORG
Precedence: bulk
List-ID: <freebsd-alpha.FreeBSD.ORG>
List-Archive: <http://docs.freebsd.org/mail/> (Web Archive)
List-Help: <mailto:majordomo@FreeBSD.ORG?subject=help> (List Instructions)
List-Subscribe: <mailto:majordomo@FreeBSD.ORG?subject=subscribe%20freebsd-alpha>
List-Unsubscribe: <mailto:majordomo@FreeBSD.ORG?subject=unsubscribe%20freebsd-alpha>
X-Loop: FreeBSD.org

Andrew Gallatin <gallatin@cs.duke.edu> writes:

> Rolf Neugebauer writes:
>  > I wonder if these are related (disabling device polling "solved" the
>  > problem).
> 
> Actually, I do have a clue.. Was the machine under load?  I never
> implemented the hooks to poll from trap() -- this means that the only
> time the polling code will get called is if the context switch code
> does not find a process ready to run.  If the machine is under load,
> it will never get called.

It wasn't under really heavy load but maybe enough to not idle for
over a second or so.

> If I'm reading the x86 trap() code correctly, I don't much care for
> where Luigi put the hook on x86. I think it belongs in userret().
> 
> Would you like to try this totally untested patch?


Will do over the next couple of days (I am quite busy at the moment)
and get back to you and/or -net.

Thanks
Rolf


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