Great! Do this kernel have the "calling-problem" fixed?

What does the ignore_nice_load do? If it is equal to "1", does it mean ignoring all priority? Please explain!

it locks the CPU to 600MHz (or the next highest frequency) only during call initiation until the other side picks up.
If you ring for a long time (e.g., accidentally calling a number w/o answering machine) you may fry your device.
Maybe we should file a bug report (if someone can confirm my observation with a stock kernel)?

no, it's an older version. I could update it as soon as the workaround is stable.

It ignores any background process in calculation of CPU load. So, tracker (which runs in background) would not trigger high CPU frequency and it would save you battery.

is it like this scaling_max_freq850 or this scaling_max_freq_850 or scaling_max_freq 850?
http://talk.maemo.org/showpost.php?p...postcount=1259

It would be nice if some volunteers could summarize and link relevant posts in this
thread on http://wiki.maemo.org/Overclocking

PR1.1/PR1.1.1 stock kernel, before call:

Nokia-N900-42-11:/sys/devices/system/cpu/cpu0/cpufreq# cat scaling_max_freq scaling_cur_freq scaling_min_freq cpuinfo_max_freq cpuinfo_cur_freq cpuinfo_min_freq
600000
250000
250000
600000
249600
250000

During call until other end answers:

Nokia-N900-42-11:/sys/devices/system/cpu/cpu0/cpufreq# cat scaling_min_freq scaling_max_freq scaling_cur_freq cpuinfo_min_freq cpuinfo_max_freq cpuinfo_cur_freq
600000
600000
600000
250000
600000
600000

And after it answered:

Nokia-N900-42-11:/sys/devices/system/cpu/cpu0/cpufreq# cat scaling_min_freq scaling_max_freq scaling_cur_freq cpuinfo_min_freq cpuinfo_max_freq cpuinfo_cur_freq
250000
600000
250000
250000
600000
249600

EDIT: I guess, it is error tone recognition...

I have summarised quite a bit. Link in my sig.

Hello to all.
Todays night i've done somes speed tests when booting up the N900, using diferent kernel images. They was:

1-Test With StockImage 600 Mhz.

Since, Click Power Botton to "Enter Pin" Screen.
00:00:21:834
Since, Click "Advance" on "Enter Pin" Screen to All Widgets appear on desktop.
00:00:50:651
Total time = 00:00:21:834 + 00:00:50:651


2-Test With Lehto 900 Mhz Stock DSP:

Since, Click Power Botton to "Enter Pin" Screen.
00:00:17:032
Since, Click "Advance" on "Enter Pin" Screen to All Widgets appear on desktop.
00:00:40:235
Total time = See first exemple

3-Test With Lehto 900 Mhz 500 DSP:

Since, Click Power Botton to "Enter Pin" Screen.
00:00:17:045
Since, Click "Advance" on "Enter Pin" Screen to All Widgets appear on desktop.
00:00:42:445
Total time = See first exemple

Ok, it's not 100% accurate\fiable\credible (Dont know how to say it well in english, sorry), because i used my two hands to double click the mouse (i used http://stopwatch.onlineclock.net/) and the N900, but i've tried to touch them at the same time.

I've done the test 3 times, and the results was similar, so, i post here one of them.

I can resume that the boot speed of those two 900 Mhz images are similar, and the 600 Mhz stock more slow. (Nothing new..just to confirm with test results!)

If anyone want to try it .

Here are my 4 Desktops (a lot of things, i think lol)


(just edited this one because of the pics of friends and famaly).