* [lm-sensors] testing w83627hf treiber?
@ 2006-12-04 8:06 Jean Delvare
2006-12-04 8:14 ` Udo van den Heuvel
` (3 more replies)
0 siblings, 4 replies; 5+ messages in thread
From: Jean Delvare @ 2006-12-04 8:06 UTC (permalink / raw)
To: lm-sensors
Hi Thomas,
(Huh, please don't use URLs as post subjects.)
On Sun, 03 Dec 2006 21:20:01 +0100, Thomas Dohl wrote:
> I've found a strange bug or what ever in the Module. :)
> I'll try to explain it.
> If I boot Linux with the module "w83627hf", everything
> is ok. But if I load the module with the workaround, like
> "modprobe w83627hf reset=1" I get a 23?C higher temperature
> of the second sensor.
>
> At boottime:
> temp1: 34?C
> temp2: 33?C
>
> after modprobe with reset=1
> temp1: 34?C
> temp2: 56?C (23?C more than it is)
>
> This doesn't change after rmmod w83627hf, modprobe w83627hf.
> Only after reboot, the values return to normal.
>
> What could that be?
Easy. The W83627*F chips have temperature offset registers. The BIOS
must write an offset for the temp2 (CPU?) temperature sensor, to
account for a different thermal diode model for example, or just to
make it look cooler to the user... When the w83627hf driver resets the
chip, the offset is reset to 0 as well, so you get the "raw" reading
again.
See, this is exactly the reason why we don't reset the chip by default:
to preserve BIOS settings.
--
Jean Delvare
^ permalink raw reply [flat|nested] 5+ messages in thread
* [lm-sensors] testing w83627hf treiber?
2006-12-04 8:06 [lm-sensors] testing w83627hf treiber? Jean Delvare
@ 2006-12-04 8:14 ` Udo van den Heuvel
2006-12-04 8:16 ` Jean Delvare
` (2 subsequent siblings)
3 siblings, 0 replies; 5+ messages in thread
From: Udo van den Heuvel @ 2006-12-04 8:14 UTC (permalink / raw)
To: lm-sensors
Jean Delvare wrote:
> See, this is exactly the reason why we don't reset the chip by default:
> to preserve BIOS settings.
In that case Thomas really has a problem:
Either his fans won't work or the temperature will be off.
(in my situation it looks like it is just the fans)
So how can we fix this?
Udo
^ permalink raw reply [flat|nested] 5+ messages in thread
* [lm-sensors] testing w83627hf treiber?
2006-12-04 8:06 [lm-sensors] testing w83627hf treiber? Jean Delvare
2006-12-04 8:14 ` Udo van den Heuvel
@ 2006-12-04 8:16 ` Jean Delvare
2006-12-04 12:54 ` Jean Delvare
2006-12-04 17:28 ` Udo van den Heuvel
3 siblings, 0 replies; 5+ messages in thread
From: Jean Delvare @ 2006-12-04 8:16 UTC (permalink / raw)
To: lm-sensors
(Huh, please don't use URLs as post subjects.)
On Sun, 03 Dec 2006 09:49:38 +0100, Udo van den Heuvel wrote:
> So we might have the same problem.
> In my situation I have a VIA EK8000 which has the fans running during a
> boot until the kernel loads/starts. Then the fans stop. So is it a Linux
> issue?
> When the w83627hf module is loaded with the reset=1 parameter the fans
> start again. Fancontrol is functional (although I do not use it).
>
> What can we do to find out more?
I guess the BIOS is enabling the automatic fan speed control, with
parameters such that the fan is stopped by default. I suspect the fan
would start spinning after some time when the temperature is high
enough. Check your BIOS settings for "Smart Fan" or something similar.
Using reset=1 is no good as you lose all the BIOS settings. This should
only be used as a last resort. So look in the BIOS if you can disable
the automatic fan speed control. Or just live with it, it should work
after all.
The w83627hf driver should be updated to let the user disable the
automatic fan speed control at runtime. I think there was a patch doing
that floating around some times ago, but it never got merged.
--
Jean Delvare
^ permalink raw reply [flat|nested] 5+ messages in thread
* [lm-sensors] testing w83627hf treiber?
2006-12-04 8:06 [lm-sensors] testing w83627hf treiber? Jean Delvare
2006-12-04 8:14 ` Udo van den Heuvel
2006-12-04 8:16 ` Jean Delvare
@ 2006-12-04 12:54 ` Jean Delvare
2006-12-04 17:28 ` Udo van den Heuvel
3 siblings, 0 replies; 5+ messages in thread
From: Jean Delvare @ 2006-12-04 12:54 UTC (permalink / raw)
To: lm-sensors
On Mon, 04 Dec 2006 09:14:53 +0100, Udo van den Heuvel wrote:
> Jean Delvare wrote:
> > See, this is exactly the reason why we don't reset the chip by default:
> > to preserve BIOS settings.
>
> In that case Thomas really has a problem:
>
> Either his fans won't work or the temperature will be off.
> (in my situation it looks like it is just the fans)
>
> So how can we fix this?
As I said already: look into the BIOS menu for options to adjust or
disable the automatic fan speed control.
--
Jean Delvare
^ permalink raw reply [flat|nested] 5+ messages in thread
* [lm-sensors] testing w83627hf treiber?
2006-12-04 8:06 [lm-sensors] testing w83627hf treiber? Jean Delvare
` (2 preceding siblings ...)
2006-12-04 12:54 ` Jean Delvare
@ 2006-12-04 17:28 ` Udo van den Heuvel
3 siblings, 0 replies; 5+ messages in thread
From: Udo van den Heuvel @ 2006-12-04 17:28 UTC (permalink / raw)
To: lm-sensors
Jean Delvare wrote:
> On Mon, 04 Dec 2006 09:14:53 +0100, Udo van den Heuvel wrote:
>> Jean Delvare wrote:
>>> See, this is exactly the reason why we don't reset the chip by default:
>>> to preserve BIOS settings.
>> In that case Thomas really has a problem:
>>
>> Either his fans won't work or the temperature will be off.
>> (in my situation it looks like it is just the fans)
>>
>> So how can we fix this?
>
> As I said already: look into the BIOS menu for options to adjust or
> disable the automatic fan speed control.
Will try this later this week, since the Ek8000 having this issue is my
only firewall.
I will find a moment to investigate. Maybe I can check the manual for
such an option as well.
What system does Thmas run?
Udo
^ permalink raw reply [flat|nested] 5+ messages in thread
end of thread, other threads:[~2006-12-04 17:28 UTC | newest]
Thread overview: 5+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2006-12-04 8:06 [lm-sensors] testing w83627hf treiber? Jean Delvare
2006-12-04 8:14 ` Udo van den Heuvel
2006-12-04 8:16 ` Jean Delvare
2006-12-04 12:54 ` Jean Delvare
2006-12-04 17:28 ` Udo van den Heuvel
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.