![]() |
Re: [Announce] kernel-power stable v50 in Extras
Quote:
Quote:
So I removed it. But it had the systemui-actdead-ui or whatever it was called as a dependant. So it had to leave my N900 as well. But then ACTDEAD became so broken that I considered removing ACTDEAD for good. Now at least I can reliably power off my N900 by teaching DSME to behave (= realize that the N900 is not charging, hence no need for ACTDEAD rebooting). I can live with charging only when "Maemo" (so to speak) is up and running. But if the changes I've done to Maemo are to be more or less useful to other people, ACTDEAD will have to stay there. I don't see though why ACTDEAD needs to load X only to blank the screen and sleep (it is even rumoured that the notorious dots continue to animate while blanked. Just the thought makes me shiver! :) So I thought I'd replace ACTDEAD with something "lighter" and actually useful, like making it jump to the recovery console.. or better, to some program that actually does the "actdead" part (blank, sleep, flash leds) but if you stop it (ctrl-C) you drop to Pali's recovery console, or to bootmenu, or whatever. Actually there should be an ACTDEAD entry in bootmenu. This way you remove all the logic from getbootstate, dsme and preinit, and let the user decide upon booting. Have to think more about it (ideas welcome, of course). |
Re: [Announce] kernel-power stable v50 in Extras
Quote:
I still don't know what exactly bq27x00 does. Seems to collect data from i2c and give it to HAL. Given that (1) HAL is on my hitlist and (2) *we* can also collect data from i2c, I don't see the need for bq27x00 [ someone correct me/inform me if bq27x00 is useful for anything else ] So with only bq2415x_charger you can normally charge the N900. For those who want a battery indicator, I'm using DCE for running a script which calls i2cget (actually the script is more flexible, will work directly with i2c or with hal depending on what's available), so there's no need for HAL reporting anything. [ which leads me to the question to all: are there any programs that explicitly check the battery level? (thinking navigation and so), and if so, what's the standard way (if there is one). ] |
Re: [Announce] kernel-power stable v50 in Extras
Reportedly, BME is also useful for charging from a PC. I don't know if it's true, as I can easily access mass-storage mode with bme disabled, and then, modified version of charging script/module (that would draw no more than 500 mA from USB port) should be OK.
BME is also responsible for low battery warnings and shutting device down gently on low power (via rather complicated guess-working of "when voltage down spikes could piss of cellular modem", instead of some simple voltage threshold), but this could be re-written in kernel module. /Estel |
Re: [Announce] kernel-power stable v50 in Extras
Quote:
Last time I checked was on saturday. I just had a look now it had turned off. Didn't turn on. Put it on the wall charger I got some quick-flashing in red, like every second or so. After a couple of minutes the led turned solid yellow/orange (more orange than yellow, I'd say). A couple of minutes later it did some quick-red flashes, and then again yellow/orange. I'll leave it like that and see if/when U-boot shows up (should it turn on automatically when the emergency charging is done?). Anyway, we definitely need some way of gently informing the user and then turning off the N900 if bme is removed. Add.: wow. eventually I got a dimmed U-boot, which presumably lead to ACTDEAD mode. I then booted and i2c reported 2831mV (!). Shortly after that the value is around 2930mV. Question: can the battery (or the N900!) be physically damaged if you let the voltage go down so low? (I realize that an instant shutdown due to insufficient voltage/power may lead to file system corruption. That's another issue which doesn't worry me.) |
Re: [Announce] kernel-power stable v50 in Extras
No worries, down to 2,5V is absolutely safe, and N900 "lose consciousness" at ~2,8V, due to not enough power for anything.
Just be sure to not leave battery in such state for long unused, as it will easily discharge itself below that level, and even to absolute 0 (which, in turn, is very lethal for LiIon). At below 3V charge inside is very low, and it takes really quick to self-discharge. /Estel |
Re: [Announce] kernel-power stable v50 in Extras
hello,
I got a problem when trying to download Linux kernel for power user ( linux kernel updater (kernel power 2.6.28.10)). The reason is conflict with : linux-backports-modules-power (1:2.6.28.10-power50). Can somebody instruct me what to do. Thanks in advanceä Best regards to you all |
Re: [Announce] kernel-power stable v50 in Extras
hi folks, since when 2 days ago I installed kernel power 50 I have problems with reboot. The following procedure doesn't work:
sudo gainroot reboot The n900 switches off and just starts rebooting (just the screen with Nokia apperars), but soon after switches off permanently. In order to switch it on, I have to press the button. Same thing happens if I try to reboot with Healthcheck or CleaN900. If someone has a solution please say to me. Thank you |
Re: [Announce] kernel-power stable v50 in Extras
No one can help? I think I am not the only one to have this problem.
|
Re: [Announce] kernel-power stable v50 in Extras
Quote:
|
Re: [Announce] kernel-power stable v50 in Extras
reboot command points to
/bin/busybox march 12 2010 Is that the right executable? |
Re: [Announce] kernel-power stable v50 in Extras
Quote:
|
Re: [Announce] kernel-power stable v50 in Extras
Quote:
Anyway on another thread a user said this issue could be related with kernel-power-settings that messes up when loading custom settings on startup, but I'm not sure about that (the user actually had problems in booting up the device, anyway could be a similar issue). |
Re: [Announce] kernel-power stable v50 in Extras
From your desktop, notebook, or whatever:
flasher-3.5 --set-rd-flags=force-power-key Problem solved. /Estel // edit In case You're wondering, it isn't reflash - it just set r&d flag (not even full r&d mode) to always return bootreason as pwr_key. As side effect, reboot->shutdown problem is fixed. |
Re: [Announce] kernel-power stable v50 in Extras
Thank you, anyway it seems to me a workaround that does not resolve the problem -> system hangs trying to reboot...
Is it known why does it happen? |
Re: [Announce] kernel-power stable v50 in Extras
No idea, for me it *never* hangs on reboot, with force-power-key.
|
Re: [Announce] kernel-power stable v50 in Extras
Quote:
Any idea about it? Output of flasher is: Code:
rhm9bw@domp:/var/lib/vdr$ sudo flasher-3.5 --query-rd-mode |
Re: [Announce] kernel-power stable v50 in Extras
Strange, are You sure that You haven't explicitly enabled R&D mode, before?
On my device, enabling force-power-key flag doesn't result in any other symptom of R&D (other than wrench key in boot screen), i.e. no keyboard flickering on activity, and definitely no diodes behind keyboard lit, when screen isn't slided out. I can't comment on it much, as it's my first 1st hand experience with R&D flags or mode at all (!) - never had need for using it. Maybe someone more experienced with R&D could explain it? /Estel |
Re: [Announce] kernel-power stable v50 in Extras
reboot/shutdown is a bit flaky here as well. sometimes reboot works, other times it don't. shutdown, from power menu, is usually ok but sometimes results in the device rebooting instead.
|
Re: [Announce] kernel-power stable v50 in Extras
Quote:
How about yours? CU & HANW Michael |
Re: [Announce] kernel-power stable v50 in Extras
Only toolbox, (in fact it's Nokia 7xx/N8xx and wrench on it), no mentions of kernel version.
|
All times are GMT. The time now is 16:09. |
vBulletin® Version 3.8.8