[2.6.35+] Laptop overheats.

[RobinJ]

At first I thought this was just an issue with Ubuntu and Nouveau, but I noticed this happens with all distro's wich feature a kernel version higher than 2.6.35 (around that, I know it *does* occur in 2.6.39, but not in 2.6.35). My graphic card gets very hot. So does the rest of my laptop's components, but less so. In all of the distro's I've tested I must come to the conclusion it's a bug in the Linux kernel that started occuring around 2.6.36 or something like that. I've got aa graphic card capable of 3D compositing until Ubuntu 10.10, in 11.04 this didn't work anymore all of a sudden, same for all distro's based on Ubuntu.
The sensors indicate very high temperatures (and so do my hands and knees when holding the laptop or putting it on my knees) for any distro a kernel version higher than 2.6.35, and this became a lot worse with the 3.x series somewhere (Ubuntu 11.10 is the milestone for me). I've already heard there are more people than only me with this problem.

My laptop's a Dell Latitude D620, Intel Core 2 Duo 2GHz, 2GB RAM, NVidia Quadro NVS 110M (64MB video RAM).

--
Old report:
I was just trying out Ubuntu 11.10 from my USB drive. I used it for about 10 minutes, and then my laptop cooler started spinning a little louder (still not as loud as it normally does in 11.04). I installed lm-sensors, ran sensors-detect, en checked the output of the sensors command... Glad I was sitting in an armchair or I would have fallen off.
http://minus.com/mdRt76f6P#16f
This seems to me like a critical bug, as this could kill users their hardware. My laptop was indeed getting quite hot on my knees (though I wouldn't have dreamed of 108°C!). I ran sensors again, just to be sure, all that happened was that the temperature had risen by anoter 1°C.
I'm just using Unity 2D, as my graphic card (NVidia Quadro NVS 110M) has been blacklisted for the normal Unity.
I wasn't doing anything in particular. Only playing some music using Banshee (by the way, the music lens doesn't show any songs in Unity 2D? Also, the font in the Unity 2D Panel changes after a few minutes use, as you can see in the screenshot).

Dell Latitude D620: Intel Core 2 Duo 2Ghz, 2GB RAM, NVidia Quadro NVS110 M.
Ubuntu 11.10 AMD64 Daily (25th of september, 2011), booted from a USB drive.

[RobinJ (robinj)]

For comparison, here's the output of sensors in 11.04:
robin@Robin-Latitude-D620:~$ sensors
acpitz-virtual-0
Adapter: Virtual device
temp1: +60.5°C (crit = +126.0°C)

coretemp-isa-0000
Adapter: ISA adapter
Core 0: +61.0°C (high = +100.0°C, crit = +100.0°C)

coretemp-isa-0001
Adapter: ISA adapter
Core 1: +61.0°C (high = +100.0°C, crit = +100.0°C)

[RobinJ (robinj)]

I put the 26-09-11 daily build on my USB drive and booted it. The first thing I did was enabling the right repositories and installing/configuring lm-sensors; The temperature was a 100°C. All I did was boto up my computer! So my colclusion is that it generates a lot of heat when booting up. When doing (almost) nothing for about 10 minutes the temperature reported by nouveau in lm-sensors went down to about 86°C. I tried running Unity 3D manually by typing "unity" in the terminal, and it worked fairly well (with the nouveau drivers!), but the temperature went up again. I logged out and loggen in again with Unity 2D.

These are the current values:
acpitz-virtual-0
Adapter: Virtual device
temp1: +61.5°C (crit = +126.0°C)

nouveau-pci-0100
Adapter: PCI adapter
temp1: +86.0°C (high = +100.0°C, crit = +110.0°C)

coretemp-isa-0000
Adapter: ISA adapter
Core 0: +61.0°C (high = +100.0°C, crit = +100.0°C)
Core 1: +60.0°C (high = +100.0°C, crit = +100.0°C)

These were the values right after installing and configuring lm-sensors:
acpitz-virtual-0
Adapter: Virtual device
temp1: +78.5°C (crit = +126.0°C)

nouveau-pci-0100
Adapter: PCI adapter
temp1: +102.0°C (high = +100.0°C, crit = +110.0°C)

coretemp-isa-0000
Adapter: ISA adapter
Core 0: +76.0°C (high = +100.0°C, crit = +100.0°C)
Core 1: +76.0°C (high = +100.0°C, crit = +100.0°C)

[RobinJ (robinj)]

Would someone take a look at this? It's a critical bug that can kill (or at least drastically shorten the lifetime of) hardware!

[Bryce Harrington (bryce)]

Hey robinj,

Hi, thanks for reporting this issue during the development period of
Ubuntu.

I notice there's not been further comments to the bug report since the
release came out, would you mind updating us on the status of it in the
release?

Are you still able to reproduce the issue? If not, do you think the bug
report can be closed, or do you think we should continue tracking it?

[RobinJ (robinj)]

The computer still gets rather hot (and really hot when forcing Unity on), but it's not a 110°C anymore.
Now the hottest it gets (using Unity 2D) is around 90 to 100°C.

My personal opinion is that it would be best to get to the source of this heat production, as I've had this problem on another computer as well running Ubuntu 11.10. The final decision is up to you of course. Only thing I can say is that I'm not going to risk this again, so I've switched to another distro for the time being, until this issue gets a permanent fix.

[RobinJ (robinj)]

This might be a kernel bug. This seems to happen with every version of the Linux kernel past 2.6.35, and it's a lot worse in the 3.x series.
Adding `pcie_aspm=force` to the kernel line in grub.cfg also killed my battery in a few weeks time.

[RobinJ (robinj)]

Ok, I just tried some different distro's with kernel >2.6.38 and <2.6.35. It seems the problem only occurs with kernel versions higher than 2.6.35. Changing the package to the kernel.

[Joseph Salisbury (jsalisbury)]

Would it be possible for you to test the latest upstream kernel? It will allow additional upstream developers to examine the issue. Refer to https://wiki.ubuntu.com/KernelMainlineBuilds . If possible, please test the latest v3.2-rcN kernel (Not a kernel in the daily directory). Once you've tested the upstream kernel, please remove the 'needs-upstream-testing' tag(Only that one tag, please leave the others). This can be done by clicking on the yellow pencil icon next to the tag located at the bottom of the bug description and deleting the 'needs-upstream-testing' text.

If this bug is fixed by the mainline kernel, please add the following tag 'kernel-fixed-upstream-KERNEL-VERSION'. For example, if kernel version 3.2-rc1 fixed and issue, the tag would be: 'kernel-fixed-upstream-v3.2-rc1'.

If the mainline kernel does not fix this bug, please add the tag: 'kernel-bug-exists-upstream'.

Thanks in advance.

[RobinJ (robinj)]

Problem is that I don't really fancy sacrificing my laptop... I've just got a new battery and I don't want it to be killed again. Testing a "beta version" of a kernel seems to me as a great risk.

[lljccoael (lljccoael-deactivatedaccount)]

Thank you for trying to bring attention to this serious problem, RobinJ.

I too have suffered through all of this and used to frequently participate in three other bug reports on here regarding the same issue over the last year, as it has been known for some time. But I just cannot deal with it anymore and risk shortening the life of my hardware in an atmosphere of stonewalling and altogether too much "skepticism" (for example, look at the hostile response from developers every time Phoronix tries to report about the overheating/power consumption problem).

I will add only this: amazingly enough, since 11.10, adding "acpi_osi=\"Linux\" pcie_aspm=force" through GRUB doesn't work anymore, so it is actually getting worse. I have no idea what's causing it. I'm going to try using the old 2.6.35 kernel, but if that doesn't work out for whatever reason, I'll have to give up Linux in general. The lack of community and developer support to tackle or even admit the existence of this problem has caused me to fear something is seriously amiss.

[RobinJ (robinj)]

Doesn't seem to occur anymore for me in Ubuntu 12.04 Beta.