* Re: [lm-sensors] VIA Epia ML6000EAG vt1211 CPU temperature wrong value
2012-06-06 13:51 [lm-sensors] VIA Epia ML6000EAG vt1211 CPU temperature wrong value kernel
@ 2012-06-06 15:19 ` Guenter Roeck
2012-06-11 7:55 ` kernel
` (3 subsequent siblings)
4 siblings, 0 replies; 6+ messages in thread
From: Guenter Roeck @ 2012-06-06 15:19 UTC (permalink / raw)
To: lm-sensors
On Wed, 2012-06-06 at 09:51 -0400, kernel@noltedennis.de wrote:
> Hello everyone,
>
> It seems the cpu temperature of the VIA EPIA ML6000EAG chip vt1211
> seems to be a bit off in
> sensors version 3.1.2 with libsensors version 3.3.2.
>
> The CPU is a VIA Nehemiah with a passive cooling setup.
>
> All tests happend at approx. 20 degree celsius with no active cooling
> inside the case.
>
>
> Here is some hopefully relevant data:
>
>
> /etc/sensors3.conf
> snipped:
>
> chip "vt1211-*"
>
> label in5 "+3.3V"
>
> label temp2 "SIO Temp"
> #compute temp1 (@*@*0.0046) - (@*0.088) - 0.748, (@*0.9686) + 65
> #compute temp1 (@ - 65) / 0.9686, (@ * 0.9686) + 65
> #kernel.org on C3 Ezra
> compute temp1 (@ - 83.869) / 0.9528, (@ * 0.9528) +83.869
>
>
>
> # sensors
> vt1211-isa-6000
> Adapter: ISA adapter
> in0: +1.95 V (min = +0.00 V, max = +2.63 V)
> in1: +1.99 V (min = +0.00 V, max = +2.63 V)
> in2: +1.03 V (min = +0.00 V, max = +2.63 V)
> in3: +1.97 V (min = +0.00 V, max = +2.63 V)
> in4: +2.12 V (min = +0.00 V, max = +2.63 V)
> +3.3V: +3.33 V (min = +0.00 V, max = +4.18 V)
> fan1: 0 RPM (min = 0 RPM, div = 2)
> fan2: 0 RPM (min = 0 RPM, div = 2)
> temp1: +34.8 C (high = +179.6 C, hyst = -88.0 C)
> SIO Temp: +56.0 C (high = +204.0 C, hyst = +0.0 C)
> cpu0_vid: +1.750 V
>
> this output is generated with the above compute value.
>
> the raw value for the temperature are as followed:
>
> # cat /sys/devices/platform/vt1211.24576/temp*_input
> 117000
> 56000
>
> this is with a cpu load of <5 percent.
>
> with approx. 100 % cpu load the temp1 value increases slightly to 121000.
>
>
> I did try to find additional info for the modul , settings for the
> sensors3.conf and came up with the 3 above mentioned compute values.
>
> As they seem to be for a different processor im not sure if those values
> are correct for my system at all.
>
>
> #cat /proc/cpuinfo
> processor : 0
> vendor_id : CentaurHauls
> cpu family : 6
> model : 9
> model name : VIA Nehemiah
> stepping : 8
> cpu MHz : 666.644
> cache size : 64 KB
> fdiv_bug : no
> hlt_bug : no
> f00f_bug : no
> coma_bug : no
> fpu : yes
> fpu_exception : yes
> cpuid level : 1
> wp : yes
> flags : fpu vme de pse tsc msr cx8 sep mtrr pge cmov pat mmx
> fxsr sse rng rng_en ace ace_en
> bogomips : 1333.28
> clflush size : 32
> cache_alignment : 32
> address sizes : 32 bits physical, 32 bits virtual
> power management:
>
> please let me know if i can give additional information to help solving
> this "problem".
>
The driver code says "temp1 is an intel thermal diode which is scaled in
user space". That means the reading depends on the board (ie how the
sensor is connected) and on the CPU type. So anything you can come up
with is as "correct" as it can be. Best would be, if that is possible,
to compare readings with values reported by the BIOS and try to get a
match.
temp1_hyst looks wrong, though. It should not be negative. I don't know
the driver or the chip, and the data sheet is only available under NDA,
so I can not verify it. What is its raw reading ?
Thanks,
Guenter
_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors
^ permalink raw reply [flat|nested] 6+ messages in thread* Re: [lm-sensors] VIA Epia ML6000EAG vt1211 CPU temperature wrong value
2012-06-06 13:51 [lm-sensors] VIA Epia ML6000EAG vt1211 CPU temperature wrong value kernel
2012-06-06 15:19 ` Guenter Roeck
@ 2012-06-11 7:55 ` kernel
2012-06-11 13:39 ` Guenter Roeck
` (2 subsequent siblings)
4 siblings, 0 replies; 6+ messages in thread
From: kernel @ 2012-06-11 7:55 UTC (permalink / raw)
To: lm-sensors
Hello and thank you for helping.
The raw readings on following config settings:
compute temp1 (@ - 83.869) / 0.9528, (@ * 0.9528) +83.869
# cat /sys/devices/platform/vt1211.24576/temp1_input
99000
# cat /sys/devices/platform/vt1211.24576/temp2_input
41000
for comparison the sensor output:
#sensors
...
fan1: 0 RPM (min = 0 RPM, div = 2)
fan2: 0 RPM (min = 0 RPM, div = 2)
temp1: +15.9 C (high = +179.6 C, hyst = -88.0 C)
SIO Temp: +41.0 C (high = +204.0 C, hyst = +0.0 C)
cpu0_vid: +1.750 V
without the compute line the hyst value seems to be ok, but the temp is
pretty high.
temp1: +102.0 C (high = +255.0 C, hyst = +0.0 C)
Raw value: 104000
The BIOS of this device does not have an option to check temperature
directly so i cannot confirm/deny the temp.
Anything else i can try?
With regards
Dennis Nolte
> On Wed, 2012-06-06 at 09:51 -0400, kernel@noltedennis.de wrote:
>> Hello everyone,
>>
>> It seems the cpu temperature of the VIA EPIA ML6000EAG chip vt1211
>> seems to be a bit off in
>> sensors version 3.1.2 with libsensors version 3.3.2.
>>
>> The CPU is a VIA Nehemiah with a passive cooling setup.
>>
>> All tests happend at approx. 20 degree celsius with no active cooling
>> inside the case.
>>
>>
>> Here is some hopefully relevant data:
>>
>>
>> /etc/sensors3.conf
>> snipped:
>>
>> chip "vt1211-*"
>>
>> label in5 "+3.3V"
>>
>> label temp2 "SIO Temp"
>> #compute temp1 (@*@*0.0046) - (@*0.088) - 0.748, (@*0.9686) + 65
>> #compute temp1 (@ - 65) / 0.9686, (@ * 0.9686) + 65
>> #kernel.org on C3 Ezra
>> compute temp1 (@ - 83.869) / 0.9528, (@ * 0.9528) +83.869
>>
>>
>>
>> # sensors
>> vt1211-isa-6000
>> Adapter: ISA adapter
>> in0: +1.95 V (min = +0.00 V, max = +2.63 V)
>> in1: +1.99 V (min = +0.00 V, max = +2.63 V)
>> in2: +1.03 V (min = +0.00 V, max = +2.63 V)
>> in3: +1.97 V (min = +0.00 V, max = +2.63 V)
>> in4: +2.12 V (min = +0.00 V, max = +2.63 V)
>> +3.3V: +3.33 V (min = +0.00 V, max = +4.18 V)
>> fan1: 0 RPM (min = 0 RPM, div = 2)
>> fan2: 0 RPM (min = 0 RPM, div = 2)
>> temp1: +34.8 C (high = +179.6 C, hyst = -88.0 C)
>> SIO Temp: +56.0 C (high = +204.0 C, hyst = +0.0 C)
>> cpu0_vid: +1.750 V
>>
>> this output is generated with the above compute value.
>>
>> the raw value for the temperature are as followed:
>>
>> # cat /sys/devices/platform/vt1211.24576/temp*_input
>> 117000
>> 56000
>>
>> this is with a cpu load of <5 percent.
>>
>> with approx. 100 % cpu load the temp1 value increases slightly to
>> 121000.
>>
>>
>> I did try to find additional info for the modul , settings for the
>> sensors3.conf and came up with the 3 above mentioned compute values.
>>
>> As they seem to be for a different processor im not sure if those values
>> are correct for my system at all.
>>
>>
>> #cat /proc/cpuinfo
>> processor : 0
>> vendor_id : CentaurHauls
>> cpu family : 6
>> model : 9
>> model name : VIA Nehemiah
>> stepping : 8
>> cpu MHz : 666.644
>> cache size : 64 KB
>> fdiv_bug : no
>> hlt_bug : no
>> f00f_bug : no
>> coma_bug : no
>> fpu : yes
>> fpu_exception : yes
>> cpuid level : 1
>> wp : yes
>> flags : fpu vme de pse tsc msr cx8 sep mtrr pge cmov pat mmx
>> fxsr sse rng rng_en ace ace_en
>> bogomips : 1333.28
>> clflush size : 32
>> cache_alignment : 32
>> address sizes : 32 bits physical, 32 bits virtual
>> power management:
>>
>> please let me know if i can give additional information to help solving
>> this "problem".
>>
> The driver code says "temp1 is an intel thermal diode which is scaled in
> user space". That means the reading depends on the board (ie how the
> sensor is connected) and on the CPU type. So anything you can come up
> with is as "correct" as it can be. Best would be, if that is possible,
> to compare readings with values reported by the BIOS and try to get a
> match.
>
> temp1_hyst looks wrong, though. It should not be negative. I don't know
> the driver or the chip, and the data sheet is only available under NDA,
> so I can not verify it. What is its raw reading ?
>
> Thanks,
> Guenter
>
>
>
_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors
^ permalink raw reply [flat|nested] 6+ messages in thread* Re: [lm-sensors] VIA Epia ML6000EAG vt1211 CPU temperature wrong value
2012-06-06 13:51 [lm-sensors] VIA Epia ML6000EAG vt1211 CPU temperature wrong value kernel
2012-06-06 15:19 ` Guenter Roeck
2012-06-11 7:55 ` kernel
@ 2012-06-11 13:39 ` Guenter Roeck
2012-06-11 13:49 ` Guenter Roeck
2012-06-11 15:35 ` Guenter Roeck
4 siblings, 0 replies; 6+ messages in thread
From: Guenter Roeck @ 2012-06-11 13:39 UTC (permalink / raw)
To: lm-sensors
On Mon, Jun 11, 2012 at 03:55:07AM -0400, kernel@noltedennis.de wrote:
> Hello and thank you for helping.
>
> The raw readings on following config settings:
>
> compute temp1 (@ - 83.869) / 0.9528, (@ * 0.9528) +83.869
>
> # cat /sys/devices/platform/vt1211.24576/temp1_input
> 99000
> # cat /sys/devices/platform/vt1211.24576/temp2_input
> 41000
>
> for comparison the sensor output:
> #sensors
> ...
> fan1: 0 RPM (min = 0 RPM, div = 2)
> fan2: 0 RPM (min = 0 RPM, div = 2)
> temp1: +15.9 C (high = +179.6 C, hyst = -88.0 C)
> SIO Temp: +41.0 C (high = +204.0 C, hyst = +0.0 C)
> cpu0_vid: +1.750 V
>
temp1 is too low. Just don't have a good idea how to adjust it. Maybe measure
at low and high load and come up with a better approximation.
>
> without the compute line the hyst value seems to be ok, but the temp is
> pretty high.
>
> temp1: +102.0 C (high = +255.0 C, hyst = +0.0 C)
> Raw value: 104000
>
I think I know the problem. hyst is supposed to show report absolute temperature,
not the hysteresis value itself. Looks like displays the hysteresis value itself
(ie the temperature difference/variance) here. The equation is applied to hyst,
which results in the negative value. The raw hysteresis value should be 104000,
just like the raw temperature (assuming no hysteresis is programmed).
> The BIOS of this device does not have an option to check temperature
> directly so i cannot confirm/deny the temp.
>
Too bad.
> Anything else i can try?
>
Not much; all you can do is play around with the compute values to get reasonable
temperatures.
I'll get the hysteresis bug fixed.
Guenter
_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [lm-sensors] VIA Epia ML6000EAG vt1211 CPU temperature wrong value
2012-06-06 13:51 [lm-sensors] VIA Epia ML6000EAG vt1211 CPU temperature wrong value kernel
` (2 preceding siblings ...)
2012-06-11 13:39 ` Guenter Roeck
@ 2012-06-11 13:49 ` Guenter Roeck
2012-06-11 15:35 ` Guenter Roeck
4 siblings, 0 replies; 6+ messages in thread
From: Guenter Roeck @ 2012-06-11 13:49 UTC (permalink / raw)
To: lm-sensors
On Mon, Jun 11, 2012 at 09:39:38AM -0400, Guenter Roeck wrote:
> On Mon, Jun 11, 2012 at 03:55:07AM -0400, kernel@noltedennis.de wrote:
> > Hello and thank you for helping.
> >
> > The raw readings on following config settings:
> >
> > compute temp1 (@ - 83.869) / 0.9528, (@ * 0.9528) +83.869
> >
> > # cat /sys/devices/platform/vt1211.24576/temp1_input
> > 99000
> > # cat /sys/devices/platform/vt1211.24576/temp2_input
> > 41000
> >
> > for comparison the sensor output:
> > #sensors
> > ...
> > fan1: 0 RPM (min = 0 RPM, div = 2)
> > fan2: 0 RPM (min = 0 RPM, div = 2)
> > temp1: +15.9 C (high = +179.6 C, hyst = -88.0 C)
> > SIO Temp: +41.0 C (high = +204.0 C, hyst = +0.0 C)
> > cpu0_vid: +1.750 V
> >
> temp1 is too low. Just don't have a good idea how to adjust it. Maybe measure
> at low and high load and come up with a better approximation.
>
> >
> > without the compute line the hyst value seems to be ok, but the temp is
> > pretty high.
> >
> > temp1: +102.0 C (high = +255.0 C, hyst = +0.0 C)
> > Raw value: 104000
> >
> I think I know the problem. hyst is supposed to show report absolute temperature,
> not the hysteresis value itself. Looks like displays the hysteresis value itself
> (ie the temperature difference/variance) here. The equation is applied to hyst,
> which results in the negative value. The raw hysteresis value should be 104000,
> just like the raw temperature (assuming no hysteresis is programmed).
>
Correction: hyst is related to max, not to the current temperature. So the raw
reported value should be 255000, not 104000. Sorry for the confusion.
Guenter
_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors
^ permalink raw reply [flat|nested] 6+ messages in thread* Re: [lm-sensors] VIA Epia ML6000EAG vt1211 CPU temperature wrong value
2012-06-06 13:51 [lm-sensors] VIA Epia ML6000EAG vt1211 CPU temperature wrong value kernel
` (3 preceding siblings ...)
2012-06-11 13:49 ` Guenter Roeck
@ 2012-06-11 15:35 ` Guenter Roeck
4 siblings, 0 replies; 6+ messages in thread
From: Guenter Roeck @ 2012-06-11 15:35 UTC (permalink / raw)
To: lm-sensors
On Mon, 2012-06-11 at 09:49 -0400, Guenter Roeck wrote:
> On Mon, Jun 11, 2012 at 09:39:38AM -0400, Guenter Roeck wrote:
> > On Mon, Jun 11, 2012 at 03:55:07AM -0400, kernel@noltedennis.de wrote:
> > > Hello and thank you for helping.
> > >
> > > The raw readings on following config settings:
> > >
> > > compute temp1 (@ - 83.869) / 0.9528, (@ * 0.9528) +83.869
> > >
> > > # cat /sys/devices/platform/vt1211.24576/temp1_input
> > > 99000
> > > # cat /sys/devices/platform/vt1211.24576/temp2_input
> > > 41000
> > >
> > > for comparison the sensor output:
> > > #sensors
> > > ...
> > > fan1: 0 RPM (min = 0 RPM, div = 2)
> > > fan2: 0 RPM (min = 0 RPM, div = 2)
> > > temp1: +15.9 C (high = +179.6 C, hyst = -88.0 C)
> > > SIO Temp: +41.0 C (high = +204.0 C, hyst = +0.0 C)
> > > cpu0_vid: +1.750 V
> > >
> > temp1 is too low. Just don't have a good idea how to adjust it. Maybe measure
> > at low and high load and come up with a better approximation.
> >
> > >
> > > without the compute line the hyst value seems to be ok, but the temp is
> > > pretty high.
> > >
> > > temp1: +102.0 C (high = +255.0 C, hyst = +0.0 C)
> > > Raw value: 104000
> > >
> > I think I know the problem. hyst is supposed to show report absolute temperature,
> > not the hysteresis value itself. Looks like displays the hysteresis value itself
> > (ie the temperature difference/variance) here. The equation is applied to hyst,
> > which results in the negative value. The raw hysteresis value should be 104000,
> > just like the raw temperature (assuming no hysteresis is programmed).
> >
> Correction: hyst is related to max, not to the current temperature. So the raw
> reported value should be 255000, not 104000. Sorry for the confusion.
>
After looking into the data sheet, it is not clear to me if the
hysteresis registers store an absolute temperature or a relative
temperature. So it may as well be that the hysteresis value is not
programmed, or not programmed correctly, on your board. Difficult to say
which one it is without experimenting with the board itself.
Guenter
_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors
^ permalink raw reply [flat|nested] 6+ messages in thread