You are not logged in.
My laptop now overheats (105 C) within 10 minutes of being booted into arch. It burns my hands touching the bottom. This problem only started a few days ago. Before that it generally stayed below 50 C on all sensors according to acpi -t. It was always cool to the touch, no matter how long it had been on. I've had arch running on it for over a year with no problems. I'm pretty sure it isn't a hardware issue because I can play system intensive games in Windows XP for hours on it without it getting nearly that hot. Yet, in arch, running pacman is enough to overheat it. I'm using cpu_freq and laptop_mode-tools. I'll be out for the day, but when I get back I will post my configs. Until then, I would be grateful for any suggestions on how to triage this problem further.
Thanks for you help.
Offline
the last time my laptop started overheating, i realized there was a ton of dust clogging up all the fans and everything else. have you tried cleaning it out to see if that helps to drop the temperatures any?
Offline
I would use powertop to see what's using your processor the most, maybe. There're too many possibilities for where the problem lies. Could be a kernel issue, a module not loading properly, messed up configs etc. I wouldn't start guessing without more info.
Also, isn't 105 a bit hot for a processor to run? Maybe I'm living in the past too much, and too used to mobile-computing focused chips, but I thought 85-95 is the standard range without risking irreparable damage.
As a note, I too, had overheating problems, but mine were fixed by cleaning out the comp and reapplying thermal paste.
Offline
Here is my modules array from rc.conf:
MODULES=(tg3 ipw2200 acpi-cpufreq cpufreq_ondemand cpufreq_powersave ac97_bus snd-mixer-oss snd-pcm-oss snd-page-alloc snd-pcm snd-timer snd snd-ac97-codec snd-intel8x0 snd-intel8x0m soundcore wacom usblp !btusb fuse)
Here are my daemons:
DAEMONS=(syslog-ng acpid laptop-mode cpufreq @network netfs crond @alsa @sshd @cups @hal @openntpd @xinetd !bluetooth @gpm skvm)
Assuming it is not a hardware issue, what other info would be useful in diagnosing the problem?
Offline
Hmm... cpufreq-seg -g powersave/ondemand?
top to see if any application is using 100% CPU
My victim you are meant to be
No, you cannot hide nor flee
You know what I'm looking for
Pleasure your torture, I will endure...
Offline
I would also try what Dheart says, cpufreq-set -g powersave/ondemand or I would try to check if there isn't too much dust clogging the fins of the heatsink as most probably that can be the cause of problems for older hardware.
R00KIE
Tm90aGluZyB0byBzZWUgaGVyZSwgbW92ZSBhbG9uZy4K
Offline
My own problem may be related: the fan is continually firing (and noisy) after last upgrade (a big one after leaving some 2 weeks). Battery autonomy has dropped (3h30 to 2h55). My untrained eyes see nothing obviously wrong, but:
- powertop shows 25% wakeups (48-55 out of 110-130) due to "<kernel core> : hrtimer_start_range_ns (tick_sched_timer)", is it normal?
- htop shows a screenful of "/usr/sbin/console-kit-daemon" lines... are those normal?
- CPU usage is low but uses both cores (2% and 4% at 800MHz), before if I remember well only one was used (I mean with just X and Xmonad running)
Thanks for any hint. (If not related maybe I should open a new topic?)
Eric
Offline
I also noticed the powertop message and found this bug which seems to be related => https://bugs.launchpad.net/ubuntu/+sour … bug/373245
Currently I'm installing kernel26-git (2.6.31-rc7) from aur to see if this problem still occurs.
strange, this seems to be an rather old bug..
EDIT: powertop-output is still the same (and the patch which should solve this solution was already applied for shure):
Top causes for wakeups:
29.1% (160.2) <kernel core> : hrtimer_start_range_ns (tick_sched_timer)
Last edited by martin_herndl (2009-08-26 13:29:38)
Offline
Powertop shows me the same high wake ups you're getting with hrtimer_start_range_ns. cpufreq-set powersave doesn't prevent the problem. I think it's interesting that I don't get the cpu_critical message until shutdown. acpi -t tells me 50C or lower until then. Its as if the cpu just stops working all of a sudden, but comes back on when I reboot.
Edit:
I have this in kernel.log:
kernel: fan PNP0C0B:01: registered as cooling_device1
kernel: ACPI: Fan [C266] (off)
kernel: fan PNP0C0B:02: registered as cooling_device2
kernel: ACPI: Fan [C266] (off)
kernel: fan PNP0C0B:03: registered as cooling_device3
kernel: ACPI: Fan [C266] (off)
kernel: fan PNP0C0B:04: registered as cooling_device4
Edit:
Okay, so I finally remembered to check htop and it is reporting constantly high CPU usage but I don't know where its coming from. I tried sorting by cpu usage but nothing shows using more than 1.0. What other programs could I use to find the culprit?
EDIT: I found the culprit for the CPU usage. Now I'm trying to see whether that was it, or if it was my battery.
Last edited by kandrews (2009-08-27 01:31:01)
Offline