"Critical temperature reached" issue

[cgorac]

Got Thinkpad R51 machine that was running Slackware Linux for almost a
year without any problems. Recently I installed new version of
Slackware, and then compiled 2.6.18 kernel on my own (am regularly using
hand-compiled kernels). From this point, machine is going down from
time to time (always after being on for full days and in the middle of a
large compilation) with following messages in /var/log/syslog:
-------
Oct 3 20:10:59 foo kernel: ACPI: Critical trip point
Oct 3 20:10:59 foo kernel: Critical temperature reached (98 C), shutting down.
-------
I wasn't able to find any definite info on this problem on the net.
There are people claiming that ACPI measuring are plain wrong, and seems
to me that this could be the case here. I never investigated
temperature issues before, but I can see now ACPI is reporting 65-70 C
CPU temperature on my machine upon boot, and going 90-95 C for each
compilation or other heavy CPU usage period (large numerical
calculations, image transformations etc.). On the other side, when I
put my hand on bottom of my laptop, it doesn't seem that much hot (it is
hot, but I can cure hold my hand on its bottom for large period of
time). Also, I can hear fan noise and it is monotonic, thus I guess it
seems to be working regularly (fan spead reported trough
/proc/acpi/ibm/fan, when ibm-acpi kernel module is loaded with
"experimental=1" option, is around 3700rpm, and it seems to be working
in "level 7", which means fastest, speed. So - any suggestion, should I
just ignore it or my CPU is in real danger here?

Thanks.

[tfflivemb2]

Heat can damage the CPU or even the board.

Is the R51 still under warranty? If it is, you could always send it in to Lenovo to be looked at.

If it isn't under warranty, then I would suggest applying a little Arctic Silver to the CPU, to help keep the temps down.

[K0LO]

and check the heatsink fins/fan for dust accumulation

[cgorac]

Thanks for the replies. After opening my notebook, I found indedd lots of dust on fan, and also that practically there is no heat paste left between processor and cooler. So I cleaned the dust, and applied AS5 over processor. Seems like there exist slight improvement for now, but I'll have to wait couple days to make any conclusion (I found on my AS tube that it takes 200 hours or so for it to achieve optimum temperature conductivity). Another piece of advice for people in alike situation would be probably to check their battery - my battery is detached from my notebook all the time (because I'm always on AC), but in that case there is a rather large hole on the back that makes it possible for dust to enter into notebook case. So I've put some duct tape over.