Extreme heat increase after upgrading to 2.6.35

zippy · 2010-08-18 23:20:02

Yep, YAHT. I just upgraded this afternoon to .35. I went from idling at 50C to 65C. I'm dying here. Fans are working, cpufreq, in a dual core laptop that ranges from 800mhz-2ghz. Im sitting here watching sensors with 1-2% load @ 800mhz and both cores are over 64C. Edit:x86_64 btw

Looking around, i stumbled upon this which is interesting because every kernel seems to get a little hotter-
https://bugs.launchpad.net/ubuntu/+sour … bug/524281
* I know its the deban kernel but deb/ubu and even redhat have always ran 5C or so cooler than arch

So they fixed that in .35 mainline? If i understand this commit they reverted in .35.2?

http://git.kernel.org/?p=linux/kernel/g … 724a4bbb07

IDK, im just grasping at straws at this point, i was looking forward to .35 and the supposed intel graphics improvements.

Last edited by zippy (2010-08-18 23:20:50)

zippy · 2010-08-18 23:59:38

I downgraded and noticed when i ran sensors, my critical temp is 85C, and when running sensors in .35, it was 100C. Did something change in the kernel about temp reporting? It reminds me of the iphone wifi bug fix.

Inxsible · 2010-08-19 00:10:06

Moving to testing as 2.6.35 is still in testing...

nvm --- just did a pacman Syu and it is in core now. ...so I'll move it back...

Skripka · 2010-08-19 00:10:38

Did you machine feel hotter to the touch? It could be the new kernel was throwing lm-sensors for a loop.

zippy · 2010-08-19 00:44:11

To be honest, not really. I have sort of a stand that keeps it off of my lap and lets air underneath. I ran my hand under the laptop and it didnt feel any hotter than normal, but my eyes have more decision making power than my sense of touch i guess. I just totally freaked when makepkg was compressing virtualbox_bin for an upgrade it spiked to 90C and i said f*ck this im out, shut it down for 20 min then downgraded.

Makes sense, I'l look around and see if i can find anything. Should I have ran sensors-setup again (after kernel upgrades?)

Kosmonavt · 2010-08-19 06:52:10

It looks like local problem. On my machine temp is fully nominal (laptop HP Compaq 6530b, CPU Intel Core2Duo P8600), 40 C in idle, also Arch x86_64. Have just recompiled VBox module, temp wasn't higher than 50 C. So the kernel is probably ok, looks like you need to use sensors-setup.

ScottArch · 2010-08-19 22:41:02

I run a laptop so the temps you are talking about are normal to me, but it sounds like a sensor fault. If you are worried I would rally look in to installing PHC-intel from AUR. The underclocking of my Core2Duo T7300 leaves me running at 47C as I type this and as low as 32C in idle with no drop in performance (see PHC Wiki). Great for the battery life (and the life or the processor). I am running KDE 4.5 too.

ScottArch · 2010-08-19 22:43:00

See you are on a laptop also, sorry

Edit: just peaked at 53C while recompiling VB module

Last edited by ScottArch (2010-08-19 22:57:36)

zippy · 2010-08-20 00:52:27

Wow i can finally log in again.. anyways, im at a loss. I have a T5800 in a Dell Inspiron 1318, cpuinfo on both kernels are exactly the same, but lm_sensors reports 85C critical on .34 and 100C critical on .35. Ive tried different versions of lm_sensors. Ive tried different 35.2 kernels, all increase the critical and it seems that its running at the same percentages, i.e -
50C idle to 83C high load max..sensors reports 85C is the critical temp ---.34 kernel
65C idle to 96C high load max..sensors reports 100C is the critical temp ---.35 kernel

Ive tried different boot parameters on the 35 kernel, no difference.

I'm going to give up unless someone else has this issue and stick with 34 for now. IDK if its .35, sensors, or my crappy bios.

Last edited by zippy (2010-08-20 00:54:27)

GRem · 2010-08-20 02:21:01

I have the same issue. T5550 here.

Updated to .35, now both high and crit temperature levels are 100ºC instead of 85/100ºC.
The main temperature seems to have increased by that very same 15ºC overall...

.34 idling: 55-58ºC
.35 idling: 70-73ºC

However, it doesn't feel that hot to the touch, when I put the system under a heavy load it claims 90-93ºC.
I also have another sensor "temp1" which is currently 53ºC, and on .34 it was always a few degrees lower than the CPU. On .35, temp1's reading is still normal, 52-55ºC idling, about 67-70ºC under heavy load.

For now I'm using temp1 as a "reference" to the CPU's temperature by adding a few degrees, since its area definitely doesn't feel 15ºC hotter.
I did have the CPU on ~95ºC once in a really hot day and a very heavy load, but that was back on the "normal" behaving kernels, and I'm sure that was actually 95, since the ventilation grids were practically a powerful hair dryer. That doesn't happen on .35 when it claims similar temperatures.

(Also, hello Arch forums!)

Last edited by GRem (2010-08-20 02:21:42)

gtklocker · 2010-08-20 04:59:15

I don't have any heating problem at all.

I have an ATi, and I can say the fan has finally shut up. I'm now idling at ~53C and the laptop isn't hot. It's actually more cooler than with Windows and Catalyst

lgolebio · 2010-08-21 06:21:30

Same problem here. I have Dell xps M1530 with [Core(TM)2 Duo CPU T5750 @ 2.00GHz] and I'm experiencing the same issue.
On 2.6.34 temp of CPU in iddle, was about 58-60*C, and after upgrade they went up to 71*C. But the laptop is not hotter, so the readings are just incorrectly read by lm-sensors. It might by a kernel bug or maybe we have to wait for lm-sensors upgrade.

combuster · 2010-08-21 08:20:07

sensors
coretemp-isa-0000
Adapter: ISA adapter
Core 0:      +50.0°C  (high = +100.0°C, crit = +100.0°C)  

coretemp-isa-0001
Adapter: ISA adapter
Core 1:      +50.0°C  (high = +100.0°C, crit = +100.0°C)  

uname -a
Linux vostro 2.6.35.3-DELL #1 SMP PREEMPT Sat Aug 21 01:38:46 CEST 2010 x86_64 Intel(R) Core(TM)2 Duo CPU T5670 @ 1.80GHz GenuineIntel GNU/Linux

So, no problems here, it was always idling at 50-55C. Dell Vostro 1310, GM965 chipset, coretemp module for sensors reading.

zippy · 2010-08-21 13:38:46

Combuster do you remember what your critical was on .34?

broch · 2010-08-21 16:41:39

this might be related to specific hardware combined with kernel as it looks like not everybody with 2.6.35 has this issue. Not to mention that it may not be related to kernel at all, but another recent upgrade.

I have t2500 (dual core 2.o GHz) HP notebok with nVidia Go 7400:
x.34.x - 48C (light usage but not iddle)
x.35.x - 46C (see above)

so anything above 50C seems pretty excessive (particularly in the case of portable computers)

Kilzool · 2010-08-21 17:57:40

I have a Dell XPS 400 - dual core - ICH7 Intel 82801G (not sure if you need info)

# System: Dell Inc. Dell DXP051
# Board: Dell Inc. 0FJ030

And although I can't seem to get any temps or sensors to work on this machine,
I have noticed my load averages go into 1s and 2s far more often, after this last kernel upgrade (pacman -Syu).

I have also noticed, my cage fan kicking on more.

With all that being said, I am somewhat concerned of overheating the cpu as I can't actually see the temp.

So any recommendations would be helpful.. or a status update as to whether to downgrade, etc.

combuster · 2010-08-21 19:00:19

@zippy

Can't really remember but I think it was always 100C cause when my fan was chocked with dust, when I compiled kernel with j3 switch it gets hot as 93C. Now it's all nice and dandy and it doesn't get hotter than 75C after cleaning it up on 100% load...

Last edited by combuster (2010-08-21 19:00:48)

Roline · 2010-08-22 19:49:27

Idling on 62C, with 2.6.34 on 55C.

architech · 2010-08-23 01:34:24

Another Solution that works 100%

Step 1 Enable ACPI
Step 2 Enable cpufreq
Step 3 Enable powerknowd

I went from 72 c idle to 40 c idle . Also I can play flash via npviewer running like hell and the temperature will never go above 62 c
--------------------------------------
I found the solution/problem

I believe the culprit is/was the vbox kernel module. After recompiling the virtualbox kernel modules and rebooted, my systems temperature dropped almost 35 degrees. I also installed powernowd with the config of "OPTIONS="-u 95 -l 65", for additional temperature control.

----------------------------------------
tavianator, TigTex, and to who ever else that thinks this is a reporting error, it's not. I can physically tell the difference. My laptop is brushed alluminium, so the difference is highly recognizable. When your laptop runs so hot that the room temperature changes when running a flash video , come talk some logic.

Also, my sh!t started with kernel 2.6.34

On a good day, 68C - 72C Idle!
Nvidia Graphics card underclocked by half
Athlon(tm) X2 Dual-Core underclocked @ 1000Mhz per core + conservative w/ 1.2Ghz max

sensors
k10temp-pci-00c3
Adapter: PCI adapter
temp1:       +68.6°C  (high = +70.0°C)                  

acpitz-virtual-0
Adapter: Virtual device
temp1:       +69.0°C  (crit = +100.0°C) 

lspci | grep -i GeForce\ 9
02:00.0 VGA compatible controller: nVidia Corporation G96 [GeForce 9600M GT] (rev a1)

uname -a
Linux 2.6.34-ARCH #1 SMP PREEMPT Mon Jul 5 22:12:11 CEST 2010 x86_64 AMD Athlon(tm) X2 Dual-Core QL-62 AuthenticAMD GNU/Linux

cpufreq-info 
cpufrequtils 008: cpufreq-info (C) Dominik Brodowski 2004-2009
Report errors and bugs to cpufreq@vger.kernel.org, please.
analyzing CPU 0:
  driver: powernow-k8
  CPUs which run at the same hardware frequency: 0
  CPUs which need to have their frequency coordinated by software: 0
  maximum transition latency: 1000 ns.
  hardware limits: 1000 MHz - 2.00 GHz
  available frequency steps: 2.00 GHz, 1000 MHz
  available cpufreq governors: conservative, performance
  current policy: frequency should be within 1000 MHz and 1000 MHz.
                  The governor "conservative" may decide which speed to use
                  within this range.
  current CPU frequency is 1000 MHz.
analyzing CPU 1:
  driver: powernow-k8
  CPUs which run at the same hardware frequency: 1
  CPUs which need to have their frequency coordinated by software: 1
  maximum transition latency: 1000 ns.
  hardware limits: 1000 MHz - 2.00 GHz
  available frequency steps: 2.00 GHz, 1000 MHz
  available cpufreq governors: conservative, performance
  current policy: frequency should be within 1000 MHz and 1000 MHz.
                  The governor "conservative" may decide which speed to use
                  within this range.
  current CPU frequency is 1000 MHz

Last edited by architech (2010-08-27 20:50:09)

Kilzool · 2010-08-23 01:50:42

Kilzool wrote:

I have a Dell XPS 400 - dual core - ICH7 Intel 82801G (not sure if you need info)
# System: Dell Inc. Dell DXP051
# Board: Dell Inc. 0FJ030
And although I can't seem to get any temps or sensors to work on this machine,
I have noticed my load averages go into 1s and 2s far more often, after this last kernel upgrade (pacman -Syu).
I have also noticed, my cage fan kicking on more.
With all that being said, I am somewhat concerned of overheating the cpu as I can't actually see the temp.
So any recommendations would be helpful.. or a status update as to whether to downgrade, etc.

So could anyone tell me how I could get a readout for the cpu temp, on this comp?
Linux server7 2.6.35-ARCH #1 SMP PREEMPT Sat Aug 21 07:23:08 UTC 2010 i686 Intel(R) Pentium(R) D CPU 2.80GHz GenuineIntel GNU/Linux

Skripka · 2010-08-23 02:17:36

Kilzool wrote:

So could anyone tell me how I could get a readout for the cpu temp, on this comp?
Linux server7 2.6.35-ARCH #1 SMP PREEMPT Sat Aug 21 07:23:08 UTC 2010 i686 Intel(R) Pentium(R) D CPU 2.80GHz GenuineIntel GNU/Linux

http://wiki.archlinux.org/index.php/Lm_sensors

TigTex · 2010-08-23 02:30:45

people, it's not a kernel issue. The TJmax was incorrect on some cpu's (it was suposed to be 100 and 105ºC and they had 85º) in <2.6.35. Since they changed it, the temperatures aren't higher: they are being CORRECTLY REPORTED now, your pc isn't getting hotter than before: IT'S THE SAME, but now it's correctly reported to you. Careful with +80ºC on laptops, open them and clean the dust every 6 months.

Recent cpu's use DTS to thermal monitoring, older use ACPI sensors (like pentium 4 and below). The DTS instead of the real temperature, it reports the distance to tjmax (to max temperature). If your tjmax was 85ºC and your cpu was at max themperature: DTS = 0, it was reported as 85ºC but the real cpu temperature was 105ºC (because tjmax is 105ºC on that cpu, T9x00 for example). So, if you had tjmax set to 85ºC and now you have 100ºC, all your temperatures are 15ºC higher (but they are correct).

Unless your TJmax was correct in 2.6.34 and with 2.6.35 is wrong, open a bug report and just ignore the extra "virtual" temperature. If not... be careful with temperature. CPU's throttle down if they are too hot and reduce it's life.

Please read your cpu themal specifications at intel.com

Last edited by TigTex (2010-08-23 04:29:50)

lgolebio · 2010-08-23 07:19:45

@TigTex, you are wrong. I have dualboot with windows vista, and before kernel upgrade (when I was on 2.6.34) the cpu temperature was the same in iddle on both OS, Windows coretemp shows almoust the same temperature on both cores as sensors in linux. So they are not correctly reported.

Last edited by lgolebio (2010-08-23 07:20:39)

tavianator · 2010-08-23 16:26:54

lgolebio wrote:

@TigTex, you are wrong. I have dualboot with windows vista, and before kernel upgrade (when I was on 2.6.34) the cpu temperature was the same in iddle on both OS, Windows coretemp shows almoust the same temperature on both cores as sensors in linux. So they are not correctly reported.

You're making the giant assumption that Windows gets the reporting right...

To the OP: you can see if TigTex is right by seeing if the fans behave the same before and after the upgrade. If they're on now at an idle, and they weren't before the upgrade, your temperature has probably actually increased. Otherwise, I'd assume that the kernel is now reporting the correct temperature. I see temps in the 90s all the time on my thinkpad, don't worry about it. Also, check out the powertop utility, it provides some useful information.

lgolebio · 2010-08-23 17:36:27

@tavaniator, I'm not assuming that Windows gets the reporting right, because programm called "coretemp" http://www.alcpu.com/CoreTemp/, was reporting the same temperatures as lm-sensors in linux on every kerenel, up too 2.6.35.

So I still believe that there is a bug in kernel or lm-sensors. It doesn't matter where is the problem, but 75*C in iddle on arch, when the usage of CPUis about 4-5% is not normal. When I go back to kernel 2.6.34 temperatures where ok.

Arch Linux

#1 2010-08-18 23:20:02

Extreme heat increase after upgrading to 2.6.35

#2 2010-08-18 23:59:38

Re: Extreme heat increase after upgrading to 2.6.35

#3 2010-08-19 00:10:06

Re: Extreme heat increase after upgrading to 2.6.35

#4 2010-08-19 00:10:38

Re: Extreme heat increase after upgrading to 2.6.35

#5 2010-08-19 00:44:11

Re: Extreme heat increase after upgrading to 2.6.35

#6 2010-08-19 06:52:10

Re: Extreme heat increase after upgrading to 2.6.35

#7 2010-08-19 22:41:02

Re: Extreme heat increase after upgrading to 2.6.35

#8 2010-08-19 22:43:00

Re: Extreme heat increase after upgrading to 2.6.35

#9 2010-08-20 00:52:27

Re: Extreme heat increase after upgrading to 2.6.35

#10 2010-08-20 02:21:01

Re: Extreme heat increase after upgrading to 2.6.35

#11 2010-08-20 04:59:15

Re: Extreme heat increase after upgrading to 2.6.35

#12 2010-08-21 06:21:30

Re: Extreme heat increase after upgrading to 2.6.35

#13 2010-08-21 08:20:07

Re: Extreme heat increase after upgrading to 2.6.35

#14 2010-08-21 13:38:46

Re: Extreme heat increase after upgrading to 2.6.35

#15 2010-08-21 16:41:39

Re: Extreme heat increase after upgrading to 2.6.35

#16 2010-08-21 17:57:40

Re: Extreme heat increase after upgrading to 2.6.35

#17 2010-08-21 19:00:19

Re: Extreme heat increase after upgrading to 2.6.35

#18 2010-08-22 19:49:27

Re: Extreme heat increase after upgrading to 2.6.35

#19 2010-08-23 01:34:24

Re: Extreme heat increase after upgrading to 2.6.35

#20 2010-08-23 01:50:42

Re: Extreme heat increase after upgrading to 2.6.35

#21 2010-08-23 02:17:36

Re: Extreme heat increase after upgrading to 2.6.35

#22 2010-08-23 02:30:45

Re: Extreme heat increase after upgrading to 2.6.35

#23 2010-08-23 07:19:45

Re: Extreme heat increase after upgrading to 2.6.35

#24 2010-08-23 16:26:54

Re: Extreme heat increase after upgrading to 2.6.35

#25 2010-08-23 17:36:27

Re: Extreme heat increase after upgrading to 2.6.35

Board footer