Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 10 Oct 2007 14:59:05 +0200
From:      Mel <fbsd.mobile@rachie.is-a-geek.net>
To:        freebsd-mobile@freebsd.org
Cc:        Lars Engels <lme@freebsd.org>, Norberto Meijome <freebsd@meijome.net>
Subject:   Re: Management of Thermal
Message-ID:  <200710101459.05867.fbsd.mobile@rachie.is-a-geek.net>
In-Reply-To: <20071008111601.kfrb6qt8isw4owc0@0x20.net>
References:  <20071008172756.2aed69e7@meijome.net> <20071008173604.1e449ca2@meijome.net> <20071008111601.kfrb6qt8isw4owc0@0x20.net>

next in thread | previous in thread | raw e-mail | index | archive | help
On Monday 08 October 2007 11:16:01 Lars Engels wrote:
> Quoting Norberto Meijome <freebsd@meijome.net>:
> > On Mon, 8 Oct 2007 17:27:56 +1000
> > Norberto Meijome <freebsd@meijome.net> wrote:
> >
> > Powerd is enabled.
> >
> > I am currently doing not much at all (listening to music stream,
> > load of  0.24, 0.45, 0.69) and tz0  =3D 78C. it drops down a lot more
> > when running on batteries only
>
> Hi Norberto,
>
> your temperature values are a little bit too high. On my Notebook
> (model name      : Intel(R) Pentium(R) M processor 1.73GHz) the CPU is
> at between 50=B0C - 60=B0C when I use normal desktop applications and up
> to 80=B0C when I compile something.
> You could open your case and look if there's dust in the airflow so
> the fans cannot blow out all the hot air.
> I only use the powerd with "-a adaptive -b adaptive" flags. There
> should be no need for the cpufreq module.

Experience with a HP laptop shows this to be not true. It currently is work=
ing=20
perfectly with a laptop cooling pad and not going anywhere near critical=20
values, but before that, we had to use the following in /etc/sysctl.conf to=
=20
prevent the computer from locking up (I suspect that the Intel CPU shuts=20
itself down before burning out):
hw.acpi.thermal.user_override=3D1
hw.acpi.thermal.tz0._PSV=3D"75C"
hw.acpi.thermal.tz0._CRT=3D"80C"

These values are 5C lower then detected values. I started with 10C lower an=
d=20
reached critical too early. 5C works perfectly. Logging the temp[1] each=20
minute I could see the effect of the cooling fans once it reached 75C.

[1]
# cat /usr/local/sbin/tempd
#!/bin/sh

=46ACILITY=3Dkern.notice
SLEEP=3D60

while true; do
        MSG=3D$(sysctl -n hw.acpi.thermal.tz0.temperature)
        /usr/bin/logger -p ${FACILITY} -i -t tempd "CPU temp: ${MSG}"
        sleep ${SLEEP}
done

# cat /usr/local/etc/rc.d/tempd
#!/bin/sh
# vim: ts=3D4 sw=3D4 syn=3Dsh noet nobackup
# REQUIRE: syslog
# PROVIDE: tempd
# KEYWORD: FreeBSD

=2E /etc/rc.subr

name=3Dtempd
rcvar=3D`set_rcvar`
command=3D"/usr/local/sbin/${name}"
pidfile=3D"/var/run/${name}.pid"
start_cmd=3D${name}_start

tempd_start()
{
        /usr/sbin/daemon -f -p ${pidfile} -c ${command}
}
load_rc_config $name
run_rc_command "$1"

=2D-=20
Mel



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200710101459.05867.fbsd.mobile>