topic Re: thermal management in Operating System - Linux

thermal management

gert ormel — Mon, 12 Dec 2005 18:39:40 GMT

My son is running SuSE 10.0 on his HP nc6220 laptop. He thinks the tempature regulation on his machine is unreliable, with the fans often starting too late or not all.

I found several ACPI warnings in the dmesg output, one them is:
'ACPI: Looking for DSTD in inird...not found! not found!'

Has anybody any idea what this could mean and what to do about it?

Suggestions mostly welcome.

Gert,

Re: thermal management

Bill Thorsteinson — Tue, 13 Dec 2005 17:16:11 GMT

Linux tends to be picky about ACPI tables
following the standard. Many don't.

Newer Debian kernels allow you to load
a corrected ACPI table at boot time.
SuSE may do so as well.
The DSTD message may actuually be from this
option.

There is a site with corrected ACPI tables,
as well as tools to extract and correct them.
I don't have the URL handy.

Re: thermal management

gert ormel — Tue, 13 Dec 2005 19:14:26 GMT

Thanks for the reply!

I already extracted and re-compiled the DSDT table, which gave me 2 warnings only; no errors.
So, I don't believe it's a faulty DSTD table, although it apparently has been compiled with a Microsoft compiler.
I think I'll have to submit a bugzilla report,
although I'm inclined to think it's an OS related issue.

Best regards,

Re: thermal management

dirk dierickx — Wed, 14 Dec 2005 03:23:07 GMT

so perhaps acpi is not working correctly, what about lm-sensors? have you tried that to get fan/temp data?

Re: thermal management

gert ormel — Wed, 14 Dec 2005 19:12:49 GMT

Hello Dirk,

Thanks for the reply.

I've noticed that the temp. at which the fan(s) start working varies a lot.

When the machine boots up, a fan is always spinning but stops after a few minutes. Then after some time, sometimes fans start working at around 50 degrees, sometimes they haven't started yet at 60 or higher and I've to reboot to cool the machine down. This happens on battery as well as on AC power.

In the dmesg output I can't find error messages related to ACPI; there're quite a few warning messages, as :
CPI: Subsystem revision 20050408
ACPI-0309: *** Warning: Unknown GPE method type:
C274 (name not of form _Lxx or _Exx)

ACPI-0212: *** Warning: Device is not power
manageable

ACPI-0309: *** Warning: Unknown GPE method type:
ACPI-1048: *** Warning: Failed to acquire
semaphore[dffe19a0|1|0], AE_TIME

I've no idea what the following message means, but noticed it's repeated over and over int the /var/log:

shpchp: acpi_shpchprm:get_device PCI ROOT HID fail=0x5
shpchp: acpi_shpchprm:\_SB_.C002 evaluate _BBN
fail=0x5
shpchp: acpi_shpchprm:get_device PCI ROOT HID fail=0x5

The temp in the 3 thermal zones at bootup
are registered as:

ACPI: Thermal Zone [TZ1] (55 C)
ACPI: Thermal Zone [TZ2] (48 C)
ACPI: Thermal Zone [TZ3] (27 C)

The output of the powersave -T is:

powersave -T
Thermal Device no. 0:

Temperature: 23
Critical: 85
Passive: 50
Thermal Device no. 1:
Temperature: 45
Critical: 83
Thermal Device no. 2:
Temperature: 48
State: ACTIVE
Critical: 105
Passive: 100
Active 0: 80
Active 1: 70
Active 2: 58
Active 3: 35

Does it make sense to you? It could well be an ACPI error, or the temp. in the various zones isn't correctly registered by the OS.

Best regards,

Gert