ichss0: transition to 1 failed

Kevin Oberman oberman at es.net
Thu Jun 9 18:08:54 GMT 2005

> Date: Thu, 9 Jun 2005 02:07:29 +0200
> From: Marcin Jessa <lists at yazzy.org>
> Sender: owner-freebsd-mobile at freebsd.org
> After I recompiled my kernel with device cpufreq , dmesg spits out tons of messages like:
> ichss0: transition to 1 failed
> What does that mean?
> sysctl -a |grep ich shows:
> dev.ichss.0.%desc: SpeedStep ICH
> dev.ichss.0.%driver: ichss
> dev.ichss.0.%parent: cpu0
> dev.ichss.0.freq_settings: 1397/-1 -1/-1
> sysctl dev.cpu|grep freq shows
> dev.cpu.0.freq: 1397
> dev.cpu.0.freq_levels: 1397/-1 1222/-1 1047/-1 873/-1 698/-1 523/-1 349/-1 174/-1
> I dont use est.
> My CPU: (Thinkpad R50e):
> CPU: Intel(R) Celeron(R) M processor         1400MHz (1395.48-MHz 686-class CPU)
>   Origin = "GenuineIntel"  Id = 0x695  Stepping = 5

OK. This one looks really weird to me.
dev.ichss.0.freq_settings: 1397/-1 -1/-1
This says that you have two CPU speeds, 1.397 GHz and -1. ????

Worse, this system is a Celeron-M. I am not terribly familiar with the
Celerons, but I thought the Celeron-M supported EST, not ICHSS. Why do
you not use est? It's much more powerful than any other power management

Can you post a full DMESG? I'm curious what the device probe is picking
up. Something looks bogus. I'm just not sure what.

Is someone else running a Celeron-M system on current? What do you see
for 'sysctl dev.ichss' and 'sysctl dev.cpu.0'? I have a hunch that the
system is not probing correctly and really does not do ICHSS.
R. Kevin Oberman, Network Engineer
Energy Sciences Network (ESnet)
Ernest O. Lawrence Berkeley National Laboratory (Berkeley Lab)
E-mail: oberman at es.net			Phone: +1 510 486-8634

More information about the freebsd-mobile mailing list