[maemo-developers] Where are the N900 "too much time at 600Mhz" safeguards?

From: Matan Ziv-Av matan at svgalib.org
Date: Thu Jan 28 01:25:44 EET 2010
On Wed, 27 Jan 2010, Javier S. Pedro wrote:

> When I got my N900, one of the first things I noticed is that (as measured by
> powertop) I could never get a 100% ratio at 600 Mhz, but more like 95%. I
> quickly assumed this was the safeguard for the issue Igor Stoppa talked about
> at the Maemo Summit.
> However, I've noticed today (as suggested by a tmo post) that the above is not
> caused by any special modification in the kernel, but rather because of the
> CPU idling while waiting for the SGX / some other hw (so, testing methodology
> failure on my part :) ).
> Thus, given any task bounded by raw CPU throughput, the device will happily
> clock itself at 600Mhz, even for hours. Doesn't that contradict what Igor said
> at the summit?

At least for 30 minutes, there appear to be no 'safeguards':

N900:~# date ; cat /sys/devices/system/cpu/cpu0/cpufreq/stats/time_in_state ; for  i in ` seq 1 1000` ; do bzip2 -c9 /lib/libc-2.5.so  > /dev/null ; done ; cat /sys/devices/system/cpu/cpu0/cpufreq/stats/time_in_state ; date
Wed Jan 27 15:35:32 IST 2010
600000 43683
550000 487
500000 17132
250000 1164197
600000 254442
550000 487
500000 17132
250000 1164200
Wed Jan 27 16:10:39 IST 2010

This represents more than 99.99% of 35 minutes at 600MHz.

Note that I ran this test with no SIM, screen off, not charging and wifi 
connected, but with practically no traffic. The device got only slightly 
warm, but it was hardly noticeable, so I guess that the power draw of 
the CPU, even at 600MHz does not have a large effect the system.

BTW, is there a temperature sensor somewhere in the system like there is 
in the N810?

Matan Ziv-Av.                         matan at svgalib.org

More information about the maemo-developers mailing list