Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 21 Jul 2004 10:18:53 -0700
From:      Ted Faber <faber@ISI.EDU>
To:        Nate Lawson <nate@root.org>
Cc:        freebsd-acpi@freebsd.org
Subject:    Re: Dell Latitude 610 CPU clock rate degrading (debug info included)
Message-ID:  <20040721171853.GB45921@pun.isi.edu>
In-Reply-To: <40FD6BEB.4060102@root.org>
References:  <20040720165016.GA35412@pun.isi.edu> <40FD6BEB.4060102@root.org>

next in thread | previous in thread | raw e-mail | index | archive | help

--2B/JsCI69OhZNC5r
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline

On Tue, Jul 20, 2004 at 12:00:59PM -0700, Nate Lawson wrote:
> Ted Faber wrote:
> >My Dell Latatude 610 is having its clock rate go from 1000 Mhz to 733
> >Mhz under -CURRENT that was cvsupped and installed yesterday.  I'm
> >honestly not sure how long this has been happening, since I only got the
> >clock speed monitor (the one that runs under gkrellm) hooked up a couple
> >weeks ago.  I hooked it up because the machine seemed slow.  It had been
> >happening for a couple weeks at least.
> >
> >The rate seems to change after using the CPU hard for a while - a big
> >compile, like make buildworld seems to do it, though even a few minutes
> >of compilation seems sufficient.
> >
> >When the CPU speed changes I get a message in the logs saying 
> >
> >cpu0: Performance states changed
> >
> >The state never seems to change back to a high performance one.  I've
> >configured the BIOS not to do power management with the AC connected,
> >and this happens with the AC connected.
> 
> This transition is being done by the BIOS for passive cooling.  We can't 
> control it until the cpufreq driver is committed.  When the system gets 
> hot, the BIOS steps the processor down to 733 mhz but the timecounters have 
> no way of knowing it happened.

Thanks for the info.  If you've got a list of people to ping when the
cpufreq driver appears, put me on it.  I'll be watching for the HEADSUP
in any case.

> 
> >I tried booting without ACPI, but my time of day clock ran consistenty
> >slow - which might mean that the same degredation was occurring without
> >the OS readjusting for it.  In any case it wasn't clear that no ACPI
> >helped.
> 
> Same problem.  What timecounter are you using?  Switching away from TSC 
> would help this.

With ACPI there's no clock skew problem, so that's how I've been
running.  Just for my information, how might I switch timecounters?  (It
sounds like it won't solve the problem to switch away from ACPI, but I'm
curious).

Thanks again for your quick response.

-- 
Ted Faber
http://www.isi.edu/~faber           PGP: http://www.isi.edu/~faber/pubkeys.asc
Unexpected attachment on this mail? See http://www.isi.edu/~faber/FAQ.html#SIG 

--2B/JsCI69OhZNC5r
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (FreeBSD)

iD8DBQFA/qV9aUz3f+Zf+XsRAqumAKD1beqQU66u4HOQ1bOZny0AqbHe2QCgmfY4
rY56EkotXoIqWoLwDrIgRC8=
=isyl
-----END PGP SIGNATURE-----

--2B/JsCI69OhZNC5r--



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20040721171853.GB45921>