lm sensors and funky Core temperatures..

b0uncyfr0 · 2008-05-30 12:20:11

Ive been using it a lot lately..(funky) not sure why.

Currently running 2.6.25 and lm sensors 3.0.2. During 2.6.24 the coretemp module would not load and now its ok but the core temperatures seem to be off. Right now the CPU is at 27 C whilst both cores are at 37 C each. That cannot be right.

Real-Temp in Windows gives me 30 C idle for both cores which seems more accurate because the heatsink is barely warm. This is on a Xeon 3110 aka E8400. Maybe its just me or lm sensors isnt doing something right.

Last edited by b0uncyfr0 (2008-05-30 13:46:56)

iBertus · 2008-05-30 16:47:03

37C sounds right. My X3220 Quad-Core shows temps in that range for each core under idle.

ST.x · 2008-06-02 13:44:46

actually my lm_sensors has been reporting the wrong coretemps lately for a c2d e6400@2.6ghz. It should be around low 40's for each core from experience but they show:
Not sure when it happened, might have started when I turned testing repo on and got 2.6.25, but I don't get why that would cause it.

$ sensors
w83627dhg-isa-0290
Adapter: ISA adapter
VCore:       +1.19 V  (min =  +0.00 V, max =  +1.74 V)   
in1:        +12.14 V  (min =  +0.00 V, max =  +1.11 V)   ALARM
AVCC:        +3.28 V  (min =  +0.51 V, max =  +0.10 V)   ALARM
3VCC:        +3.28 V  (min =  +0.00 V, max =  +1.18 V)   ALARM
in4:         +1.34 V  (min =  +0.00 V, max =  +0.27 V)   ALARM
in5:         +1.59 V  (min =  +1.02 V, max =  +0.16 V)   ALARM
in6:         +4.28 V  (min =  +3.69 V, max =  +5.53 V)   
VSB:         +3.28 V  (min =  +0.02 V, max =  +0.19 V)   ALARM
VBAT:        +3.22 V  (min =  +0.05 V, max =  +0.70 V)   ALARM
Case Fan:      0 RPM  (min = 337500 RPM, div = 4)  ALARM
CPU Fan:    1350 RPM  (min = 5273 RPM, div = 8)  ALARM
Aux Fan:    3770 RPM  (min = 5075 RPM, div = 2)  ALARM
fan4:          0 RPM  (min = 21093 RPM, div = 4)  ALARM
fan5:          0 RPM  (min =    0 RPM, div = 4)  ALARM
Sys Temp:    +31.0°C  (high = +34.0°C, hyst = -30.0°C)  sensor = thermistor
CPU Temp:    +32.5°C  (high = +80.0°C, hyst = +75.0°C)  sensor = diode
AUX Temp:   +127.0°C  (high = +80.0°C, hyst = +75.0°C)  ALARM  sensor = thermistor

coretemp-isa-0000
Adapter: ISA adapter
Core 0:      +51.0°C  (high = +84.0°C, crit = +100.0°C)  

coretemp-isa-0001
Adapter: ISA adapter
Core 1:      +55.0°C  (high = +84.0°C, crit = +100.0°C)

Last edited by ST.x (2008-06-02 13:48:01)

b0uncyfr0 · 2008-06-03 11:02:43

I knew something wasn't right. How can the core temperatures be so high. I wasn't getting any temperatures with my cores on 2.6.24 kernel but im getting them now and they are wrong.

revelc · 2008-06-03 16:15:57

Got the same problem.

With a zalman cpu-fan I got normally around 30C (E6550 @ 2.33 GHz)in the summer, I overclock it till 2.80 GHz, and normally it gets in end 30, low 40, but since the upgrade to 2.6.25 the core1 and core0 temps go thru the roof! I got 60C on them! while the CPU overall remains at around 40, BIOS says the same,

I turned on the fan a little harder (but noisier) and now it's back to below 50, but remains high to what I'm used to,

On the other hand there are all these high and critical messages (ALARMS all over the place) in sensors, is there a place where I can correct those borders?

paul2lv · 2008-06-03 23:04:49

Yeah, it's a known problem in the 2.6.25 kernel.
http://lkml.org/lkml/2008/4/18/190

ST.x · 2008-07-18 11:22:19

kewl so hopefully 2.6.26 should fix it. I need to monitor temps soon as i've started folding again.

ST.x · 2008-07-19 18:38:00

hmm I just read some mailing list on this: http://linux.derkeiler.com/Mailing-List … 12964.html
And it looks like it's not a bug but the devs just changed the TJMAX scales so for some people temps look like they went up but 15C and some went down. Still a bit confusing, dunno has anyone checked this with 2.6.26 yet?

b0uncyfr0 · 2008-07-20 15:14:58

Oh man that is not good. Who thought it would be such a hassle to read temperatures. And why is he in contact with a coretemp developer. Last time i checked coretemp was giving up false results for all 45nm chips and some 65nm (Q6600). The only temp reader giving out accurate readings for any 45nm chip is RealTemp.

The best option is to give the user the ability of setting the Tjmax. Its a better solution than guessing. I know for 45nm chips its 90.

Jst my 2 cents...

Fackamato · 2008-07-21 10:54:15

Coretemp accurately reads the temps of my Penryn CPU. (mobile)

ST.x · 2008-07-21 11:11:49

Fackamato wrote:

Coretemp accurately reads the temps of my Penryn CPU. (mobile)

Yep same here on my lappy(T9300 Penryn), but not for my desktop(Conroe). They were suggesting on that thread the temps we had before might actually be wrong and these could be more correct..hmm. At the moment im just subtracting 16C from what its displaying as they seem to match the ranges that I got even when running windows some years back. These temps are currently with my e6400@2.4ghz and folding. Changing vcore around a bit isn't affecting much either, so yea just subtracting 16 from those gives me temps I remember from using this chip for so long.

Core 0:      +56.0°C  (high = +84.0°C, crit = +100.0°C)  
Core 1:      +57.0°C  (high = +84.0°C, crit = +100.0°C)

AllFather · 2008-07-21 20:17:24

Seems to work ok here:

CPU:         +21.0°C  (high = +85.0°C, hyst = +81.0°C)
                      (crit = +100.0°C, hyst = +96.0°C)  sensor = transistor
System:      +29.0°C  (high = +85.0°C, hyst = +81.0°C)
                      (crit = +100.0°C, hyst = +96.0°C)  sensor = thermistor

coretemp-isa-0000
Adapter: ISA adapter
Core 0:      +31.0°C  (high = +82.0°C, crit = +100.0°C)

coretemp-isa-0001
Adapter: ISA adapter
Core 1:      +30.0°C  (high = +82.0°C, crit = +100.0°C)

coretemp-isa-0002
Adapter: ISA adapter
Core 2:      +30.0°C  (high = +82.0°C, crit = +100.0°C)

coretemp-isa-0003
Adapter: ISA adapter
Core 3:      +31.0°C  (high = +82.0°C, crit = +100.0°C)

Arch Linux

#1 2008-05-30 12:20:11

lm sensors and funky Core temperatures..

#2 2008-05-30 16:47:03

Re: lm sensors and funky Core temperatures..

#3 2008-06-02 13:44:46

Re: lm sensors and funky Core temperatures..

#4 2008-06-03 11:02:43

Re: lm sensors and funky Core temperatures..

#5 2008-06-03 16:15:57

Re: lm sensors and funky Core temperatures..

#6 2008-06-03 23:04:49

Re: lm sensors and funky Core temperatures..

#7 2008-07-18 11:22:19

Re: lm sensors and funky Core temperatures..

#8 2008-07-19 18:38:00

Re: lm sensors and funky Core temperatures..

#9 2008-07-20 15:14:58

Re: lm sensors and funky Core temperatures..

#10 2008-07-21 10:54:15

Re: lm sensors and funky Core temperatures..

#11 2008-07-21 11:11:49

Re: lm sensors and funky Core temperatures..

#12 2008-07-21 20:17:24

Re: lm sensors and funky Core temperatures..

Board footer