Fan running continuously after kernel switch

Fan running continuously after kernel switch

[Florian Echtler]

Hello everyone,

yesterday, I switched from kernel 2.6.25-gentoo-r8 to 2.6.27-gentoo-r7
on my Thinkpad T60. Since then, the fan has been running continuously. I
have loaded the thinkpad_acpi module and could, of course, switch to
some userspace fan controller, but I'm always a bit reluctant about
those. I've reviewed the Gentoo patches, and none of them seem to have
to do anything with ACPI. So now I'm looking for suggestions on how to
debug this..

Many thanks, Yours, Florian
-- 
0666 - Filemode of the Beast

Re: Fan running continuously after kernel switch

[Zhao Yakui]

On Thu, 2009-01-15 at 16:42 +0800, Florian Echtler wrote:
> Hello everyone,
> 
> yesterday, I switched from kernel 2.6.25-gentoo-r8 to 2.6.27-gentoo-r7
Please attach the output of acpidump so that we can confirm whether the
fan device is controlled by thinkpad acpi driver or generic fan driver.

Will you please check the kernel version of 2.6.27-gentoo-r7?
How about this issue if you only update the kernel?

Of course you can file a bug report at
http://bugzilla.kernel.org/enter_bug.cgi?product=ACPI and attach the
output of acpidump, dmesg, /proc/acpi/fan/*/state
thanks.
> on my Thinkpad T60. Since then, the fan has been running continuously. I
> have loaded the thinkpad_acpi module and could, of course, switch to
> some userspace fan controller, but I'm always a bit reluctant about
> those. I've reviewed the Gentoo patches, and none of them seem to have
> to do anything with ACPI. So now I'm looking for suggestions on how to
> debug this..
> 
> Many thanks, Yours, Florian

Re: Fan running continuously after kernel switch

[Zhao Yakui]

On Thu, 2009-01-15 at 17:32 +0800, Florian Echtler wrote:
> Hello Zhao,
> 
> > > yesterday, I switched from kernel 2.6.25-gentoo-r8 to 2.6.27-gentoo-r7
> > Please attach the output of acpidump so that we can confirm whether the
> > fan device is controlled by thinkpad acpi driver or generic fan driver.
> Dumpfile is attached.
thanks for the acpidump.
>From the dumpfile it seems that the fan is not controlled by the generic
ACPI fan driver. It seems that it is controlled by thinkpad_acpi driver.
cc: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
> 
> > Will you please check the kernel version of 2.6.27-gentoo-r7?
> If you mean what stock kernel this is based on: it's 2.6.27.10.
> 
> > How about this issue if you only update the kernel?
> I'm quite sure I didn't update anything besides the kernel..
Please double check it again. 
Thanks.
> 
> Thanks, Yours, Florian

Re: Fan running continuously after kernel switch

[Henrique de Moraes Holschuh]

On Fri, 16 Jan 2009, Zhao Yakui wrote:
> On Thu, 2009-01-15 at 17:32 +0800, Florian Echtler wrote:
> > > > yesterday, I switched from kernel 2.6.25-gentoo-r8 to 2.6.27-gentoo-r7
> > > Please attach the output of acpidump so that we can confirm whether the
> > > fan device is controlled by thinkpad acpi driver or generic fan driver.
> > Dumpfile is attached.
> thanks for the acpidump.
> >From the dumpfile it seems that the fan is not controlled by the generic
> ACPI fan driver. It seems that it is controlled by thinkpad_acpi driver.
> cc: Henrique de Moraes Holschuh <hmh@hmh.eng.br>

The thinkpad-acpi driver exports an interface that can be used to control
the fan, but you do it on your own risk, and you have to EXPLICLTY opt-in by
giving it the fan_control=1 parameter.

Even if you do give thinkpad-acpi the fan_control=1 parameter, it won't
touch the fan at all unless you command it to through procfs or sysfs.

The driver doesn't do any sort of automated control, the EC does it by
itself.

What can help is for you to tell us the temperatures that thinkpad-acpi
report.  If any of them is above 43C, the fan will be left on continuously
by the EC.

-- 
  "One disk to rule them all, One disk to find them. One disk to bring
  them all and in the darkness grind them. In the Land of Redmond
  where the shadows lie." -- The Silicon Valley Tarot
  Henrique Holschuh

Re: Fan running continuously after kernel switch

[Florian Echtler]

> The thinkpad-acpi driver exports an interface that can be used to control
> the fan, but you do it on your own risk, and you have to EXPLICLTY opt-in by
> giving it the fan_control=1 parameter.
Yes, I read about that option, but I'd rather like to avoid it..

> What can help is for you to tell us the temperatures that thinkpad-acpi
> report.  If any of them is above 43C, the fan will be left on continuously
> by the EC.
Here's the temperatures:

cat /proc/acpi/ibm/thermal 
temperatures:	43 37 35 45 50 -128 29 -128 37 46 43 -128 -128 -128 -128
-128

However, there's obviously a bit more to this.

a) when I load the coretemp module and try to read the temperature via
lm-sensors, I get two values around 60 °C, while thinkpad-acpi shows the
above. Same applies to the standard ACPI thermal interface. I have a
total of 13 different thermal measurements.. :-)

b) the really weird part: when I remove the power supply and let the
laptop run on battery, the fan immediately stops! Some minutes later, it
starts again, but I'm pretty sure that the temperatures are not the only
factor here. When I plug the power cord back in after some minutes, the
fan stops again..

Yours, Florian
-- 
0666 - Filemode of the Beast

--
To unsubscribe from this list: send the line "unsubscribe linux-acpi" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: Fan running continuously after kernel switch

[Henrique de Moraes Holschuh]

On Fri, 16 Jan 2009, Florian Echtler wrote:
> cat /proc/acpi/ibm/thermal 
> temperatures:	43 37 35 45 50 -128 29 -128 37 46 43 -128 -128 -128 -128
> -128

That 45 next to the 50, and the 46 and 43 near the end tail of -128 are
likely going to keep your fan on, yes.

That 50 is a probably a battery pack telling us something that we don't
know, but since the EC does, it doesn't matter (i.e. if it is to be ignored,
the EC will know that and do just that).   I asked Lenovo, and they refused
to explain it.

> However, there's obviously a bit more to this.

Yes.

> a) when I load the coretemp module and try to read the temperature via
> lm-sensors, I get two values around 60 °C, while thinkpad-acpi shows the
> above. Same applies to the standard ACPI thermal interface. I have a
> total of 13 different thermal measurements.. :-)

Heh.  The only ones you can somehow figure out without messing with the
hardware are coretemp (they're exactly what they should be), and the ACPI TZ
ones (read the AML disassembly dump).

All the others require that you open your thinkpad and make good use of a
can of cold air to locate what sensor changes when you cool each part of the
planar card.

> b) the really weird part: when I remove the power supply and let the
> laptop run on battery, the fan immediately stops! Some minutes later, it

Yes, the EC (or BIOS, who knows.  Much of the thinkpad firmware is SMM code)
changes the fan/thermal control profile based on many things.  It also tells
ACPI to limit the CPU speed, and other honky things.

It all depends on how you configured your BIOS.

-- 
  "One disk to rule them all, One disk to find them. One disk to bring
  them all and in the darkness grind them. In the Land of Redmond
  where the shadows lie." -- The Silicon Valley Tarot
  Henrique Holschuh
--
To unsubscribe from this list: send the line "unsubscribe linux-acpi" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html