Messages in this thread | | | Date | Wed, 21 Apr 2004 14:03:25 -0400 (EDT) | From | Zwane Mwaikambo <> | Subject | Re: flooded by "CPU#0: Running in modulated clock mode" |
| |
On Wed, 21 Apr 2004, Alexey Mahotkin wrote:
> I'm running 2.6.5, compiled for i686 architecture, on Celeron 1.8GHz. > Motherboard is ASUS P4S800-MX. The machine has 1U form-factor. > > Usually processor runs at ~40C, according to lm-sensors (and BIOS > agrees with it). > > When I start CPU hog (prime-net client), the temperature rises to ~62C > in several minutes. After that I'm getting a swamp of messages: > > > CPU#0: Running in modulated clock mode > CPU#0: Temperature/speed normal > CPU#0: Temperature above threshold > CPU#0: Running in modulated clock mode > CPU#0: Temperature/speed normal > CPU#0: Temperature above threshold > [ ...endless... ] > > When I stop it, messages stop immediately, and processor cools down > back to 40C. > > However, according to google, maximum Celeron temperature must be > somewhere near 80C. Also, AFAIU, if it were overheating, I'd get a > single message about "Temperature above threshold". > > Why is it flooding me on KERN_EMERG level? How can I change the > temperature threshold?
I'll consider switching it to KERN_INFO or perhaps find some way of throttling the output. The threshold is set so that its below the catastrophic shutdown threshold. so ~60C sounds right considering your processor will probably shutdown at 80C.
> Should I try noapic? Should I try acpi=off?
It's actually CONFIG_X86_MCE_P4THERMAL you could always disable that, but then you wont get notification when your processor is overheating. Is the cooling setup in the room housing the server also adequate? You could also try looking into better cooling solutions for the processor. - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |