@nkirk - I think SR is very close to if not stable, a week or so should not matter.

@knobby - try to lower DSP frequency for 805,850 and 900 from 600 to 430 to see if that helps. Or limit max frequency to 850 or 805.

Uploaded on the same place.

It is the whole range 125-900 which is recalculated. I have no idea why, but with stock calibrations voltages were too high, i.e. for 250 SR voltage was about 28-30 while with re-calculated calibrations it stays 21-22. The same for other frequencies. I also changed lowest voltage limit from 20 to 18. That should affect 125MHz, and will (hopefully) have visible effect on battery life if one sets lower frequency to be 125. If someone give me a hint on what is the lowest voltage for 125 (without SR) I can lower that limit further. Also I would like to have others voltages for 250 (manually undervolted) to tweak SR driver further.

And yes, uploaded debs are based on latest KP with only SR driver modified.

I have used 125:20,90 setting for a long time. Without and with SR enabled (SR last 6 months) and device has been stable.

I use Alarmed to put phone into power safe mode during night times (23:00 - 06:45). Enable freqs are 125MHz and 500MHz. This seems to be a bit more power saving mode than 250MHz & 500MHz or single 250MHz or 500MHz when CPU is awake, idle = 0MHz always.

I could try even lower voltages for 125MHz and check if it is still stable.

I have a general question about OMAP PM:
Is it so that DVFS sets OPP according to CPU power needs (when more calculation power is needed, then frequency will be increased also voltage) and SmartReflex tries to autocompensate VDD1 and VDD2 voltages around used OPP (i.e fine tuning based on different paramenters).

- VDD1 domain control MPU&IVA (ARM, DSP, etc..)
- VDD2 domain control on-chip interconnects

Is it possible to see used (current) voltage values, when SR's are enabled?
/etc/power/ -folder files are not showing any changes w/wo SR.

Uptime ~14h with latest KP without any problems.

EDIT: Now testing 125:18,90 setting and N900 seems to work OK (this editing done with that setting and using single 125M freq)

Why would problems with supl.nokia.com affect N900 only? As I said, N82 has no issues.

@Gusse - sr_vdd1_voltage reflects current VDD1 voltage when SR is enabled, but for 125 it won't go bellow 18 because of hard limit.

sr_vdd1_voltage is showing VDD1 voltage OK. For some reason SR was not enabled when I checked values, my misstake.

18 seems to be OK value at 125M and when SR is enabled voltage seems to stay the same value.

@freemangordon

Hey, I don't get one thing. Before re-calibration, device with 900 mhz as max limit and SR enabled, would instantly turn off - it was always told, that this behavior was due to voltage set too low. Now You're telling us, that stock calibration turns voltage too high. Was that just "urban legend", and in fact, voltage was set too high?

Also, aren't You (due to new re-calibrated values) undervolting too much? I quite don't get idea, why do You want to know "lowest" possible values for some frequencies - on other devices, such values may be set too low. We don't want slow filesystem corruption due to "invisible" over-undervolting (situation, when it doesn't result in reboots, but operations are performed with errors, occasionally, due to voltage set too low), don't we?

Of course, it's quite possible that I've misunderstood something, in such case, just correct me please. And, again, thanks for Your awesome work!

@patlak
Some time ago, it was suggested, that Nokia do "something" to drop support for some devices. I.E, at the same time, N900 was able to ping Nokia supl servers, but wasn't able to get satellite data. Meanwhile, newest devices were able to ping and receive data. Also, on exact same time, supl.google.com was working for both N900 and other device.

Some claim, that this Nokians practice was seen before, with other devices. Ho ever, I'm not sure if all of this isn't just guesswork error, so I'm just relaying opinions from (many) threads about supl.nokia.com problems.

The "ultimate" answer for You, would be to try supl.google.com as server - if You get satellite fix very quickly with it, there is no (technical) reason on N900 side, to not get it from supl.nokia.com.

/Estel

@Estel - ok, I will try to explain the SR situation AIUI

From gathered efuse values (posted here by several users and from my devices) it is pretty obvious that there are only two OPPs which have values calibrated - 125MHz and 250MHz. The other frequencies (500,550 and 600) have one and the same values for all the devices so far. Which does not look like there was any calibration, just a random numbers someone (TI most probably) has decided will be OK for all.

The problem is (and that is why I am saying voltages are too high) that based on that calibration values SR puts voltage for 250 MHz to be around 28-30 on my two devices and on another one i borrow. In the same time those devices work without any issue with a voltages around 21-23 for the same 250 MHz frequency.

So what I have done?

First - 125 MHz and 250 MHz calibration is lowered in such a way that SR calculates voltage to be around 21 - 23 for 250 MHz( it still depends on initial calibration, so "weak" device will get higher voltage). The values for higher OPPs are calculated based on 250MHz re-calculated calibration value, so again "weak" device will receive more juice.

The re-calculation is done in such a way that value for OPP(n) is based on value for OPP(n-1) plus a constant. And if DSP is overclocked an additional voltage is applied too(12.5 mV for 520MHz and 25mV for 600 MHz). Of course without having documentation for SmartReflex I cannot be sure whether re-calibrated values are optimal, but so far it seems that they are at least close to. And one of the reasons I ask for SR calculated voltages for different OPPs is that when we have them one can compare his/her manually undervolted profile with SR undervolted thus checking if they are similar/equal.

The last thing is 125MHz - I have an impression that we can run the device with voltage around 14-16 on that frequency, increasing battery life by several hours. But now hardcoded lower limit for voltage is 18 (was 20). That is why I want as many as possible people to test 125 MHz with values in range 14-18, so I can further lower the hardcoded limit.

BTW there is still a minor issue in SR driver - every time a frequency is switched SR starts AVS from highest possible voltage, which is inefficient. For KP50 I plan to optimize that too. And I will check VDD2 SR too, because if VDD1 calibration is too high it is possible that VDD2 is the same.

Re too much undervolting - TBH creeping filesystem corruption seems like an urban myth to me. If your device is undervolted too much DSP will cry loudly in an instant. Could be that I am wrong of course . Anyway, as now I (almost) understand how SR works I can easy fix/tweak any issue that could appear.

EDIT:
There were no calibration values above 600MHz. I such case for frequencies above 600MHz calibration for 600MHz was used. Too low of course. And that is another proof that stock calibrations were set too high - most devices were running 805MHz with calibration for 600MHz.

Could you try lowering it further, i.e. 17-16-... . Sorry that I am not doing it myself here, just need to rest from kernel-development for a while