[lm-sensors] fan change between 3.1.0-rc6 and commit b172e38

[lm-sensors] fan change between 3.1.0-rc6 and commit b172e38

[James Cloos]

Did something change in the f71889ed support between rc6 and today?

Up until rc6 my am3+phenom board ran its fan at about 800 Hz most of the
time, with occasional bursts to 1200 Hz or so when under heavy load.

With b172e38 it runs at around 2048 Hz.  *Loudly*.

(The only fan monitored is connected via the board's cpufan socket, and
is a 12cm 3-wire fan attached to a 12x12x6 (or so) cm heat sink.)

The cpu is running about 16 Kelvins colder than the in-case ambient,
instead of around 2 Kelvins colder.

But the noise is worse than a window fan.

The only change was the reboot into the newer kernel.

I haven't run into the particular presentation before, but it seems
worrysome.  Should I worry?

pwmconfig is not able to stop the fan, if that is relevant.

-JimC
-- 
James Cloos <cloos@jhcloos.com>         OpenPGP: 1024D/ED7DAEA6

_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

Re: [lm-sensors] fan change between 3.1.0-rc6 and commit b172e38

[Guenter Roeck]

On Sat, Sep 24, 2011 at 07:53:12PM -0400, James Cloos wrote:
> Did something change in the f71889ed support between rc6 and today?
> 
> Up until rc6 my am3+phenom board ran its fan at about 800 Hz most of the
> time, with occasional bursts to 1200 Hz or so when under heavy load.
> 
> With b172e38 it runs at around 2048 Hz.  *Loudly*.
> 
> (The only fan monitored is connected via the board's cpufan socket, and
> is a 12cm 3-wire fan attached to a 12x12x6 (or so) cm heat sink.)
> 
> The cpu is running about 16 Kelvins colder than the in-case ambient,
> instead of around 2 Kelvins colder.
> 
> But the noise is worse than a window fan.
> 
> The only change was the reboot into the newer kernel.
> 
> I haven't run into the particular presentation before, but it seems
> worrysome.  Should I worry?
> 
> pwmconfig is not able to stop the fan, if that is relevant.
> 
I don't recall a relevant change, nor did I find one in the commit log.
Did pwmconfig ever work for you ?

Thanks,
Guenter

_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

Re: [lm-sensors] fan change between 3.1.0-rc6 and commit b172e38

[James Cloos]

>>>>> "GR" = Guenter Roeck <guenter.roeck@ericsson.com> writes:

GR> I don't recall a relevant change, nor did I find one in the commit log.

OK.  Perhaps it was a change in another subsystem....

GR> Did pwmconfig ever work for you ?

I never tried it before.  I'll have to reboot to the previous kernel and check.

-JimC
-- 
James Cloos <cloos@jhcloos.com>         OpenPGP: 1024D/ED7DAEA6

_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

Re: [lm-sensors] fan change between 3.1.0-rc6 and commit b172e38

[Jean Delvare]

On Tue, 27 Sep 2011 19:37:07 -0400, James Cloos wrote:
> >>>>> "GR" = Guenter Roeck <guenter.roeck@ericsson.com> writes:
> 
> GR> I don't recall a relevant change, nor did I find one in the commit log.
> 
> OK.  Perhaps it was a change in another subsystem....
> 
> GR> Did pwmconfig ever work for you ?
> 
> I never tried it before.  I'll have to reboot to the previous kernel and check.

Please do, if nothing else just to confirm that this really makes the
fans spin down again.

As we did not touch the driver, it could be an issue with ACPI or the
thermal subsystem, or things like cpuidle or cpufreq. As the number of
commits is relatively small and you seem to know how to build a kernel
from the git repository, your best chance is probably to bisect the
problem and then contact the author of the faulty commit directly.

-- 
Jean Delvare

_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

Re: [lm-sensors] fan change between 3.1.0-rc6 and commit b172e38

[James Cloos]

D'Oh.  I had forgotten that I had upgraded the board's firmware (which
flashrom(8) makes supremely easy).  It turns out that the manuf defaults
the fan control to max rather than dynamic.

Apologies for the noise.

-JimC
-- 
James Cloos <cloos@jhcloos.com>         OpenPGP: 1024D/ED7DAEA6

_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors