LinuxQuestions.org - why message: Unable to turn cooling device "off"?

- Linux - General (https://www.linuxquestions.org/questions/linux-general-1/)

- - why message: Unable to turn cooling device "off"? (https://www.linuxquestions.org/questions/linux-general-1/why-message-unable-to-turn-cooling-device-off-592324/)

why message: Unable to turn cooling device "off"?

I'm running Debian Etch on an eMachines 1.8GHz. About every 10 minutes, I get this message:

Code:

  ACPI: Unable to turn cooling device [c1474784] 'off'

ACPI: Transitioning device [FAN1] to D3

ACPI: Transitioning device [FAN1] to D3

These messages cause a momentary lockup of my system and make my log files huge. After some searching, I think I found the problem: the thermal cutoffs for the cooling fans are in the -250 C range. I found this by running this command:

Code:

server:/var/log# cat /proc/acpi/thermal_zone/THRM/trip_points

critical (S5):          -264 C

passive:                -273 C: tc1=0 tc2=0 tsp=600 devices=0xc14742c0

active[0]:              -267 C: devices=0xc1474784

I even found a way to set these trip points as follows, and this appears to fix the problem:

Code:

server:/var/log# echo -n "65:60:50:55:50:45" > /proc/acpi/thermal_zone/THRM/trip_points

server:/var/log# cat /proc/acpi/thermal_zone/THRM/trip_points

critical (S5):          65 C

passive:                50 C: tc1=0 tc2=0 tsp=600 devices=0xc14742c0

active[0]:              55 C: devices=0xc1474784

but, the old trip points of -264 C are restored upon a reboot. So, I guess I need to know how to fix it so the ideal trip points are maintained upon reboot, or another solution if anyone knows of one. Thanks!

Well, some people thrive on over-clocking things - buy cheap, then push it further, no harm in that. But they then have to install other things to make sure that they are not overheating the hardware... Which may lead to inappropriate "warnings" when things are not configured just exactly-perfectly. I quit mandriva / suse because of this, among other things.

Frankly, I can no longer be bothered with this level of hardware-monitoring: I prefer to let the kernel use the hardware as is it is supposed to be used, otherwise I end up in a configuration nightmare because the OS assumes that I am over clocking my hardware, and sends me (inappropriate) messages about temperatures ( Because the sensor daemons, need configuring with adjustments for particular sensors according to your motherboard's hardware [Voltage, clock rate, offsets, multiplication factors etc] )

That said, if you have a command string that sorts things out for you after every boot then you can make it run automatically after the boot process has completed by appending that command to the file /etc/rc.local

Try it.

Otherwise, you could do as I have done - disabled the monitoring

If my PC has meltdown one day (it will do), I have backups.

Hi tredegar,

I'm not sure about overclocking, this system isn't overclocked, so I don't really think that is the problem. I'm not excited about disabling the monitoring and would only considering it as a last resort.

Anyway - I did as you advised and it appears to have worked. No messages in the last 12 hrs or so. Actually, I knew it would, I just needed a way to start it at boot. Thanks for your help.

al