higher temps on coretemp with 2.6.25* pls report (intel)

[kernelOfTruth]

Hi everybody,

I noticed much higher temps with 2.6.25* based kernels (currently rc8)

45 & 42 vs. 59 & 62 °C

(.24) (.25)

anyone noticed the same raise of temperature ?

I also noticed much louder noise due to faster-running cpu-fan/case-fans because of that high temperatures (3-4°C higher via cpu-cooler case-sensors)

cpu-model: Intel Core 2 Duo 6600

other core 2 duo users have reporrted to have encountered same issues

Intel-users please report !

thanks

p.s.:

if more are experiencing this, perhaps a report to lkml.org would be good, in order for a fix before 2.6.25-final release ?

[dberkholz]

Are you on at least rc8?

[kernelOfTruth]

Are you on at least rc8?

++

yeah,

I'm currently using 2.6.25-rc8 based kernel-sources (zen-sources)

several others are reporting the same behavior

I need to test vanilla-sources a few days later (currently not much time)

[dberkholz]

I think something might have gone into rc8-git3 or so. Check out git-sources.

[kernelOfTruth]

I think something might have gone into rc8-git3 or so. Check out git-sources.

thanks !

[kernelOfTruth]

*bump*

unfortunately they are still happening with latest state from 2 days ago

anyone experiencing the same behavior - too high temperatures - ?

input is highly appreciated since 2.6.25 will go final soon & sadly I don't have time to test since when this happens (exams )

is there anyone who is willing to test-drive the releases between 2.6.24 & 2.6.25-rc8 / 2.6.25-rc8-git7 to find out since when this happens ? or any input
related to this problem ...

... I'm writing too much

thanks in advance

[gerardo]

Same here on Intel Core 2 Duo on 2.6.25-r9-zen0: temp under idle is 59°C
It might be the reading which isn't correct because my fan doesn't turn any harder?
I don't know if it's the same problem on 2.6.24-zen4 or zen5 because I can't seem to get those working (hard lock on udevd while booting)

[paulbiz]

in gentoo-sources .24 my E6600 is 25C at idle and 37C max under full load (compiling). I will try zen-sources and report my results!

[paulbiz]

in gentoo-sources .24 my E6600 is 25C at idle and 37C max under full load (compiling). I will try zen-sources and report my results!

Hmm, I did zen-sources-9999 but it's still only showing 2.6.24 -- what am I missing?

[siddesu]

are you guys also experiencing a large number of wakeups when running two cores (as reported by powertop?)

[paulbiz]

Okay, now I'm using gentoo-sources-2.6.25 and coretemp definitely does show much much higher temps than before. It is showing 47C but the CPU temp from uguru3 shows 29C (previously, uguru and coretemp would almost always match or be within 1 degree).

[kernelOfTruth]

Same here on Intel Core 2 Duo on 2.6.25-r9-zen0: temp under idle is 59°C
It might be the reading which isn't correct because my fan doesn't turn any harder?
I don't know if it's the same problem on 2.6.24-zen4 or zen5 because I can't seem to get those working (hard lock on udevd while booting)

it isn't (at least for me) - I've started experiencing this "behavior" since 2.6.25-rc* kernel-releases

2.6.24-zen* releases showed all the same behavior in terms of coretemp & sensors readout

[kernelOfTruth]

latest news on this issue:

http://forums.gentoo.org/viewtopic-p-50 ... ml#5065782

[Dairinin]

Thats because new kernel (lm_sensors, etc?, etc?)incorrectly thinks our cpu's Tj is 100C, whereas for all core 2 duo's it is 85C (100C is for C0 stepping and above, AFAIK). Substract 15 and you'll get real temps.

[kernelOfTruth]

Thats because new kernel (lm_sensors, etc?, etc?)incorrectly thinks our cpu's Tj is 100C, whereas for all core 2 duo's it is 85C (100C is for C0 stepping and above, AFAIK). Substract 15 and you'll get real temps.

but why does it first happens since 2.6.25*

[Dairinin]

Maybe because of this (hwmon/coretemp.c, 2.6.25):

Code: Select all

static int __devinit adjust_tjmax(struct cpuinfo_x86 *c, u32 id, struct device *dev)
{
        /* The 100C is default for both mobile and non mobile CPUs */

        int tjmax = 100000;
....
       if ((c->x86_model == 0xf) && (c->x86_mask < 4))
....
       if (ismobile) {

                err = rdmsr_safe_on_cpu(id, 0xee, &eax, &edx);
                if (err) {
                        dev_warn(dev,
                                 "Unable to access MSR 0xEE, for Tjmax, left"
                                 " at default");
                } else if (eax & 0x40000000) {
                        tjmax = 85000;

(hwmon/coretemp.c, 2.6.24):

Code: Select all

        if (((c->x86_model == 0xf) && (c->x86_mask > 3)) ||
                (c->x86_model == 0xe))  {
                err = rdmsr_safe_on_cpu(data->id, 0xee, &eax, &edx);
                if (err) {
                        dev_warn(&pdev->dev,
                                 "Unable to access MSR 0xEE, Tjmax left at %d "
                                 "degrees C\n", data->tjmax/1000);
                } else if (eax & 0x40000000) {
                        data->tjmax = 85000;
                }

They set 85C for all cpu's with model 15 and stepping(not sure) <4. Core2duo E6600 has stepping 6...
In previous kernel all Core's with stepping >3 got 85C which obviously confused quad and 45nm owners, I think.

PS: Well, details are surely wrong, but the main idea is wrong cpu detection

[whitesouls]

I'm experiencing this where my laptop temperature used to be at 42C max. Now it can go up 90C

Here are the details

Code: Select all

SkullCrusherX ~ # uname -a
Linux SkullCrusherX 2.6.26-rc8-zen1 #1 SMP PREEMPT Sun Jun 29 23:42:49 MYT 2008 i686 Intel(R) Pentium(R) M processor 1.60GHz GenuineIntel GNU/Linux
SkullCrusherX ~ # cat /proc/cpuinfo
processor       : 0
vendor_id       : GenuineIntel
cpu family      : 6
model           : 13
model name      : Intel(R) Pentium(R) M processor 1.60GHz
stepping        : 6
cpu MHz         : 1600.000
cache size      : 2048 KB
fdiv_bug        : no
hlt_bug         : no
f00f_bug        : no
coma_bug        : no
fpu             : yes
fpu_exception   : yes
cpuid level     : 2
wp              : yes
flags           : fpu vme de pse tsc msr mce cx8 sep mtrr pge mca cmov pat clflush dts acpi mmx fxsr sse sse2 ss tm pbe up bts est tm2
bogomips        : 3191.16
clflush size    : 64
power management:

Code: Select all

SkullCrusherX ~ # acpitool -e
  Kernel version : 2.6.26-rc8-zen120080321   -    ACPI version : 20080321
  -----------------------------------------------------------
  Battery #1     : slot empty

  AC adapter     : on-line
  Fan            : enabled
  Fan Speed      : 3490 RPM

  CPU type               : Intel(R) Pentium(R) M processor 1.60GHz
  Min/Max frequency      : 600/1600 MHz
  Current frequency      : 1600 MHz
  Frequency governor     : userspace
  Freq. scaling driver   : acpi-cpufreq
  Cache size             : 2048 KB
  Bogomips               : 3191.16
  Processor ID           : 0
  Bus mastering control  : yes
  Power management       : yes
  Throttling control     : yes
  Limit interface        : yes
  Active C-state         : C0
  C-states (incl. C0)    : 4
  Usage of state C1      : 0 (0.0 %)
  Usage of state C2      : 904117 (100.0 %)
  Usage of state C3      : 0 (0.0 %)
  T-state count          : 8
  Active T-state         : T0


  Thermal zone 1 : ok, 60 C
  Trip points :
  -------------
  critical (S5):           98 C
  passive:                 94 C: tc1=5 tc2=4 tsp=600 devices= CPU


  wakeup devices : <not available>

I've taken out the battery so that it wont contribute towards the temp. raising thingy.. any fix on this?

[pdw_hu]

Just a quick fix for 90% of the topic s/2.6.25/2.6.26 :)

[whitesouls]

any updates on this?