Thermal trip_points, raise them?

Thermal trip_points, raise them?

[Albert Schueller]

Tue Jan 20 00:23:44 PST 2004

Hi,

After a brief search of the archives on this subject, I have the
following questions:

(I'm running ACPI that's in the stock 2.6.1 kernel, this is a debian
woody (mostly stable) machine, with, of course, the 2.6.1 kernel. ) 

	- Ever since I started running acpi on my Dell SmartStep 200N
	  I've been getting thermal shutdowns.  I suppose having them is
	  better than not having them, but even in a well-ventilated,
	  cool room, the machine will shutdown if I compile a kernel.
	  This is a 2GHz P4 (mobile).  It's clean inside, no dust.
	  According to the specs at Intel the chip is safe upto 100°C
	  (though maybe not the stuff around it, I don't know).

	  In
	  
	  /proc/acpi/thermal_zone/THR1/trip_points = 72 C
	  /proc/acpi/thermal_zone/THR2/trip_points = 78 C

	  well shy of the 100°C on the Intel spec.

	  I was planning to hack the acpi source and raise the trip
	  points to something like 80 and 86, but I can't find where
	  they're being set.  This makes me think that they're being
	  obtained from the BIOS somehow.

	  Is it ok to just:

	  cat "80 C" > /proc/acpi/thermal_zone/THR1/trip_points
	  cat "86 C" > /proc/acpi/thermal_zone/THR2/trip_points

	  I'll take responsibility for any thermal damage, I'm just
	  wondering if this will raise the trip_points for me.

	  Alternatively, would removing the thermal module before a big
	  job, like a kernel compile, and then reinserting it be ok?
	  Or, would it break the ACPI that's currently running?
	  (again I take responsibility for thermal damage).

	- Just out of curiosity, what is the difference between the ACPI
	  code in the stock 2.6.1 kernel and, say, the 2.6.1 patch
	  provided on the acpi.sf.net page?  (I know, I know, look at
	  the diffs.)  Are there different development groups for ACPI?
	  One that's getting their stuff into the current kernels, and
	  this group that provides a patch replacement?

Sorry if these are FAQ's.

A

-- 
Albert Schueller                                 Department of Mathematics
Office Phone:  509-527-5140                      Whitman College
Public Key:  http://carrot.whitman.edu/gpg.key   Walla Walla, WA USA 99362


-------------------------------------------------------
The SF.Net email is sponsored by EclipseCon 2004
Premiere Conference on Open Tools Development and Integration
See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
http://www.eclipsecon.org/osdn

Re: Thermal trip_points, raise them?

[Pavel Machek]

Hi!

> After a brief search of the archives on this subject, I have the
> following questions:
> 
> (I'm running ACPI that's in the stock 2.6.1 kernel, this is a debian
> woody (mostly stable) machine, with, of course, the 2.6.1 kernel. ) 
> 
> 	- Ever since I started running acpi on my Dell SmartStep 200N
> 	  I've been getting thermal shutdowns.  I suppose having them is
> 	  better than not having them, but even in a well-ventilated,
> 	  cool room, the machine will shutdown if I compile a kernel.
> 	  This is a 2GHz P4 (mobile).  It's clean inside, no dust.
> 	  According to the specs at Intel the chip is safe upto 100°C
> 	  (though maybe not the stuff around it, I don't know).
> 
> 	  In
> 	  
> 	  /proc/acpi/thermal_zone/THR1/trip_points = 72 C
> 	  /proc/acpi/thermal_zone/THR2/trip_points = 78 C
> 
> 	  well shy of the 100°C on the Intel spec.

Heh, on my athlon omnibook xe3 they have critical at ~100C, but
hardware kills it at 83C. They obviously never tested it :-(

> 	  I was planning to hack the acpi source and raise the trip
> 	  points to something like 80 and 86, but I can't find where
> 	  they're being set.  This makes me think that they're being
> 	  obtained from the BIOS somehow.
> 
> 	  Is it ok to just:
> 
> 	  cat "80 C" > /proc/acpi/thermal_zone/THR1/trip_points
> 	  cat "86 C" > /proc/acpi/thermal_zone/THR2/trip_points

You should be able to

  echo "critical:passive:active1:active2" > .../trip_points 

. [Ouch, someone should write readme about this one. Do you
volunteer]?

> 	  I'll take responsibility for any thermal damage, I'm just
> 	  wondering if this will raise the trip_points for me.

With passive cooling, you should be able to keep it below trip
points...

> 	  Alternatively, would removing the thermal module before a big
> 	  job, like a kernel compile, and then reinserting it be ok?
> 	  Or, would it break the ACPI that's currently running?
> 	  (again I take responsibility for thermal damage).

You should not be able to cause thermal damage. If you hit it *real*
hard, hardware should hard-off it. [And it works on most machines].

								Pavel
-- 
When do you have a heart between your knees?
[Johanka's followup: and *two* hearts?]


-------------------------------------------------------
The SF.Net email is sponsored by EclipseCon 2004
Premiere Conference on Open Tools Development and Integration
See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
http://www.eclipsecon.org/osdn