I have basically never had heating issues with my Compaq Evo N600c until a recent upgrade to Ubuntu's Jaunty 2.6.28. Now, the computer appears to overheat and shut down regularly. In doing some checking, these are the assigned trip points: cat /proc/acpi/thermal_zone/TZ1/trip_points critical (S5): 108 C passive: 98 C: tc1=1 tc2=2 tsp=100 devices=C000 active[0]: 80 C: devices=C1E1 active[1]: 70 C: devices=C1E2 active[2]: 55 C: devices=C1E3 Which seem quite high. I cannot change the trip points manually, and have played with the DSDT, though it is far above my ability. I know this is an incomplete bug report, so maybe the next question is: What else do I need to make it more complete? tarek : )
please attach the acpidump output using the latest pmtools here: http://www.lesswatts.org/projects/acpi/utilities.php please attach the content of "grep . /proc/acpi/thermal_zone/TZ1/*" both in 2.6.28 and in the non-overheating kernel.
Created attachment 21129 [details] ACPI Dump from n600c Hello, Thank you for your prompt reply. Here is the ACPI dump from the N600c
>> Name (_CRT, 0x0EE4) From the acpidump output, we can see that the critical trip point is hardcoded to 108C, which means that it's the same in the kernel which used to work. Plus, a critical trip point of 108C is also normal in other laptops. this is rather a thermal management regression than a trip point bug to me. can you make sure that 1. the overheating problem goes away if you switch back to the old kernel? 2. the laptop is hotter when you are running 2.6.28? Plus, please attach the full dmesg output and the result of "grep . /proc/acpi/thermal_zone/TZ1/*" both in 2.6.28 and in the non-overheating kernel.
Created attachment 21134 [details] "grep . /proc/acpi/thermal_zone/TZ1/*" in 2.6.28 Here is the output from 2.6.28. I shall go back to the old kernel shortly to get all the information from there as well. As well, I'll be tracking the logs to see if there are any contributory messages and post them here. tarek : )
Created attachment 21135 [details] dmesg from 2.6.28 after fresh boot dmesg output immediately after boot.
What is the latest version of the kernel that works properly? Can you reproduce this failure using a kernel.org kernel? upon the faiure, what is the reported temperature. Is the reading jumping around? you can monitor it with watch grep . /proc/acpi/thermal_zone/*/* /proc/acpi/fan/*/* also, when the temperature exceeds the active trip points in the output above, do the fans run (you should be able to see "on" in the output above, and hear them as well) finally, for testing, you can disable the actual shutdown by booting with thermal.nocrt=1 which will still print out, but not shutdown on CRT trip.
ping tarek.
did you feel the system really heat or just the ACPI report the temperature is high?
ping Tarek...
Apologies. Things have been a bit hectic on this end. I'll put some updated information here later today. tarek : )
ping tarek again. :)
please re-open if this issue is reproducible.