* Re: [lm-sensors] coretemp output changed after loading acpi_cpufreq
2011-02-09 9:48 [lm-sensors] coretemp output changed after loading acpi_cpufreq Owl
@ 2011-02-09 10:19 ` Jean Delvare
2011-02-09 14:04 ` Owl
` (2 subsequent siblings)
3 siblings, 0 replies; 5+ messages in thread
From: Jean Delvare @ 2011-02-09 10:19 UTC (permalink / raw)
To: lm-sensors
On Wed, 09 Feb 2011 12:48:51 +0300, Owl wrote:
> gentoo ~ # uname -a
> Linux gentoo 2.6.36-gentoo-r7 #1 SMP Fri Jan 14 14:34:23 MSK 2011 i686
> Intel(R) Atom(TM) CPU N550 @ 1.50GHz GenuineIntel GNU/Linux
>
> gentoo ~ # sensors
> coretemp-isa-0000
> Adapter: ISA adapter
> Core 0: +18.0°C (crit = +100.0°C)
>
> coretemp-isa-0001
> Adapter: ISA adapter
> Core 1: +22.0°C (crit = +100.0°C)
>
> gentoo ~ # modprobe acpi_cpufreq
> gentoo ~ # sensors
> coretemp-isa-0000
> Adapter: ISA adapter
> Core 0: +37.0°C (crit = +100.0°C)
>
> coretemp-isa-0001
> Adapter: ISA adapter
> Core 1: +41.0°C (crit = +100.0°C)
>
> It's normal?
Definitely not.
> I think, second output is much closer to BIOS values.
This reminds me of
https://bugzilla.kernel.org/show_bug.cgi?id\x17291
That case was a little different (problem was reportedly introduced by
cpuidle drivers, rather than solved by cpufreq) but both share the fact
that apparently digital thermal sensors in Intel CPUs are very
sensitive to the CPU state. The root cause may be the same.
Len, any idea how to investigate and hopefully fix this?
--
Jean Delvare
_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors
^ permalink raw reply [flat|nested] 5+ messages in thread* Re: [lm-sensors] coretemp output changed after loading acpi_cpufreq
2011-02-09 9:48 [lm-sensors] coretemp output changed after loading acpi_cpufreq Owl
2011-02-09 10:19 ` Jean Delvare
@ 2011-02-09 14:04 ` Owl
2011-02-09 16:06 ` Philip Pokorny
2011-02-09 17:12 ` Owl
3 siblings, 0 replies; 5+ messages in thread
From: Owl @ 2011-02-09 14:04 UTC (permalink / raw)
To: lm-sensors
09.02.2011 13:19, Jean Delvare пишет:
> On Wed, 09 Feb 2011 12:48:51 +0300, Owl wrote:
> > gentoo ~ # uname -a
> > Linux gentoo 2.6.36-gentoo-r7 #1 SMP Fri Jan 14 14:34:23 MSK 2011 i686
> > Intel(R) Atom(TM) CPU N550 @ 1.50GHz GenuineIntel GNU/Linux
> >
> > gentoo ~ # sensors
> > coretemp-isa-0000
> > Adapter: ISA adapter
> > Core 0: +18.0°C (crit = +100.0°C)
> >
> > coretemp-isa-0001
> > Adapter: ISA adapter
> > Core 1: +22.0°C (crit = +100.0°C)
> >
> > gentoo ~ # modprobe acpi_cpufreq
> > gentoo ~ # sensors
> > coretemp-isa-0000
> > Adapter: ISA adapter
> > Core 0: +37.0°C (crit = +100.0°C)
> >
> > coretemp-isa-0001
> > Adapter: ISA adapter
> > Core 1: +41.0°C (crit = +100.0°C)
> >
> > It's normal?
>
> Definitely not.
>
> > I think, second output is much closer to BIOS values.
>
> This reminds me of
> https://bugzilla.kernel.org/show_bug.cgi?id\x17291
>
> That case was a little different (problem was reportedly introduced by
> cpuidle drivers, rather than solved by cpufreq) but both share the fact
> that apparently digital thermal sensors in Intel CPUs are very
> sensitive to the CPU state. The root cause may be the same.
>
> Len, any idea how to investigate and hopefully fix this?
Additional info, maybe it help:
dmesg: http://codepad.org/2CXxLyLv
kernel config: http://codepad.org/vBsXFiXM
lsmod: http://codepad.org/NY7SixKo
_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [lm-sensors] coretemp output changed after loading acpi_cpufreq
2011-02-09 9:48 [lm-sensors] coretemp output changed after loading acpi_cpufreq Owl
2011-02-09 10:19 ` Jean Delvare
2011-02-09 14:04 ` Owl
@ 2011-02-09 16:06 ` Philip Pokorny
2011-02-09 17:12 ` Owl
3 siblings, 0 replies; 5+ messages in thread
From: Philip Pokorny @ 2011-02-09 16:06 UTC (permalink / raw)
To: lm-sensors
On Feb 9, 2011, at 1:48 AM, Owl <owl@umail.ru> wrote:
> gentoo ~ # uname -a
> Linux gentoo 2.6.36-gentoo-r7 #1 SMP Fri Jan 14 14:34:23 MSK 2011
> i686 Intel(R) Atom(TM) CPU N550 @ 1.50GHz GenuineIntel GNU/Linux
>
> gentoo ~ # sensors
> coretemp-isa-0000
> Adapter: ISA adapter
> Core 0: +18.0°C (crit = +100.0°C)
>
> coretemp-isa-0001
> Adapter: ISA adapter
> Core 1: +22.0°C (crit = +100.0°C)
>
Can you also read fan speeds? Is there a fan on the CPU? Some Atom
systems have a fan on the northbridge but not on the CPU itself.
Are there any other temperature sensors in the system for comparison?
> gentoo ~ # modprobe acpi_cpufreq
> gentoo ~ # sensors
> coretemp-isa-0000
> Adapter: ISA adapter
> Core 0: +37.0°C (crit = +100.0°C)
>
> coretemp-isa-0001
> Adapter: ISA adapter
> Core 1: +41.0°C (crit = +100.0°C)
>
> It's normal?
> I think, second output is much closer to BIOS values.
Again, if we knew more about other sensors or fans in the system, this
might be explained by acpi also stopping the fan when the CPU is idle
because CPU power is known to be so low and without a fan, thermal
margins are stil met (100degC)
If you don't have sensors, can you simply look or listen to the system
when you load the acpi-cpufreq module? Does anything change?
Phil P.
--
Philip Pokorny, RHCE
Chief Hardware Architect
PENGUIN COMPUTING, Inc
www.penguincomputing.com
_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [lm-sensors] coretemp output changed after loading acpi_cpufreq
2011-02-09 9:48 [lm-sensors] coretemp output changed after loading acpi_cpufreq Owl
` (2 preceding siblings ...)
2011-02-09 16:06 ` Philip Pokorny
@ 2011-02-09 17:12 ` Owl
3 siblings, 0 replies; 5+ messages in thread
From: Owl @ 2011-02-09 17:12 UTC (permalink / raw)
To: lm-sensors
09.02.2011 19:06, Philip Pokorny пишет:
> Can you also read fan speeds? Is there a fan on the CPU? Some Atom
> systems have a fan on the northbridge but not on the CPU itself.
>
> Are there any other temperature sensors in the system for comparison?
At this time - no. We wait adding support of Fintek F71869FG chip (which
is in Hans TODO list).
Anyway, I think it is not root of problem, because my system is fanless.
MB - Jetway NC9C-550-LF - http://www.jetwaycomputer.com/NC9C.html
> > gentoo ~ # modprobe acpi_cpufreq
> > gentoo ~ # sensors
> > coretemp-isa-0000
> > Adapter: ISA adapter
> > Core 0: +37.0°C (crit = +100.0°C)
> >
> > coretemp-isa-0001
> > Adapter: ISA adapter
> > Core 1: +41.0°C (crit = +100.0°C)
> >
> > It's normal?
> > I think, second output is much closer to BIOS values.
>
> Again, if we knew more about other sensors or fans in the system, this
> might be explained by acpi also stopping the fan when the CPU is idle
> because CPU power is known to be so low and without a fan, thermal
> margins are stil met (100degC)
If I run (after reboot), for example, cpuburn - and run commands, liste
above, I have got same results - coretemp get values, that about 2 time
more.
> If you don't have sensors, can you simply look or listen to the system
> when you load the acpi-cpufreq module? Does anything change?
No.
_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors
^ permalink raw reply [flat|nested] 5+ messages in thread