Fan only gets the supposed velocity when executing acpi -t.

English Hardware
#1

Hi guys,

This is one more openSUSE 11.1 x_64 bug that is driving me crazy.
I realized much time ago that sometimes my laptop start up with the fan
noisier then the usual. Sometimes, the fan noise just reduces by it self.
The problem is that, sometimes the noise stays the same for too
long. When I try to see the CPU temperature using “acpi -t”, the noise
just reduces when I execute this command.

The same happens also when the CPU temperature is high. Another day,
running a numerical computation in my laptop, I wanted to see what was
the CPU temperature. The fan noise was high. I executed “acpi -t” and the
fan noise got yet higher just as the command was executed. So, if I do not
execute “acpi -t” when hard computations are running, my laptop CPU
overheats.

Sorry, I do not know where get more information about this problem.

So, any idea about solving this issue?
Have any one experimented the same problem?

Bye.

#2

What is the brand/model? Could be DSDT related. Look here: http://forums.opensuse.org/how-faq-read-only/unreviewed-how-faq/386054-how-fix-your-buggy-dsdt.html

or here: HOWTO Fix A Buggy DSDT File - Ubuntu Forums

This is very problematic for HP/Compaq laptops.

#3

Hi guy,

Thanks for the reply. I could not find any errors or warnings with the process in the forum you suggested. Well, after following some steps, I get


linux-8z8e:/home/ednilton # iasl -tc dsdt.dsl

Intel ACPI Component Architecture
ASL Optimizing Compiler version 20081031 [Dec  3 2008]
Copyright (C) 2000 - 2008 Intel Corporation
Supports ACPI Specification Revision 3.0a

ASL Input:  dsdt.dsl - 6153 lines, 204593 bytes, 2501 keywords
AML Output: dsdt.aml - 21745 bytes, 652 named objects, 1849 executable opcodes

Compilation complete. 0 Errors, 0 Warnings, 0 Remarks, 42 Optimizations
linux-8z8e:/home/ednilton #

But my dmesg show some acpi erros:


'linux-8z8e:/home/ednilton # dmesg | grep -2i acpi | grep Error
ACPI Error (psargs-0358): [CDW1] Namespace lookup failure, AE_NOT_FOUND
ACPI Error (psparse-0530): Method parse/execution failed \_SB_.PCI0._OSC] (Node ffff88013fb067b0), AE_NOT_FOUND
ACPI Error (psargs-0358): [CDW1] Namespace lookup failure, AE_NOT_FOUND
ACPI Error (psparse-0530): Method parse/execution failed \_SB_.PCI0._OSC] (Node ffff88013fb067b0), AE_NOT_FOUND
ACPI Error (psargs-0358): [BRTW] Namespace lookup failure, AE_NOT_FOUND
ACPI Error (psparse-0530): Method parse/execution failed \_SB_.PCI0.GFX0.DD03._BCM] (Node ffff88013fb1d710), AE_NOT_FOUND
ACPI Error (psargs-0358): [BRTW] Namespace lookup failure, AE_NOT_FOUND
ACPI Error (psparse-0530): Method parse/execution failed \_SB_.PCI0.GFX0.DD03._BCM] (Node ffff88013fb1d710), AE_NOT_FOUND
ACPI Error (psargs-0358): [BRTW] Namespace lookup failure, AE_NOT_FOUND
ACPI Error (psparse-0530): Method parse/execution failed \_SB_.PCI0.GFX0.DD03._BCM] (Node ffff88013fb1d710), AE_NOT_FOUND
ACPI Error (psargs-0358): [BRTW] Namespace lookup failure, AE_NOT_FOUND
ACPI Error (psparse-0530): Method parse/execution failed \_SB_.PCI0.GFX0.DD03._BCM] (Node ffff88013fb1d710), AE_NOT_FOUND
ACPI Error (psargs-0358): [BRTW] Namespace lookup failure, AE_NOT_FOUND
ACPI Error (psparse-0530): Method parse/execution failed \_SB_.PCI0.GFX0.DD03._BCM] (Node ffff88013fb1d710), AE_NOT_FOUND
linux-8z8e:/home/ednilton #

Well, I realized that when using almost 100% of the CPU my laptop it gets the fan in the maximum speed, as I can deduce from the fan noise. But there still the question why the fan speed changes as I run “acpi -t”, mainly when my system starts with fan spinning fast?

If it helps, my laptop is Semp Toshiba IS 1454. I think this kind of laptop is only made in Brazil (Believe me, most things made here are the worst!). It is no good for using with openSUSE 11.1, and probably with any other Linux distribution.

So, still hoping for some help.

Bye.

#4

Toshiba’s are tough with Linux no matter where you get them(LOL). The dmesg errors are a kernel bug. These have been reported in several places and should disappear in the next couple of kernel releases. It is very rare to have 0 dsdt errors. It is strange that the acpi command changes fan speeds. You might want to look at toshset