Re: [Announce] kernel-power v47
 

Ok, next step, can you put here result of this command now (from a terminal with root access) :

A++

Re: [Announce] kernel-power v47
 

100% and 11%

Edit: and now 100% and 79%

Re: [Announce] kernel-power v47
 

Final step :

Open a terminal as root, then do :


For details, please read this thread from here : http://talk.maemo.org/showthread.php?t=71879&page=54

A++

Re: [Announce] kernel-power v47
 

I don't understand how this is going to help. The battery indicator is obviously not following the Nokia bme binary numbers and it's the battery indicator that suggests the battery is draining faster now. Could you please explain?

Re: [Announce] kernel-power v47
 

What you don't understand ?

NEVER load the 'b27x00_battery' driver, as it breaks bme by exclusively occupying the I2C bus !!!

Like stated by 'joerg_rw' here : http://talk.maemo.org/showpost.php?p...&postcount=534

So, there is no interest at all for the 'bq27x00_battery' driver 'pali' backported to this release.

Also, in this thread, 'pali' say, to apply patch to be able to use the values reported by 'bq27x00_battery' driver he backported, but since the driver is unusable, no need to patch anything, and keep using bme.

Hope it help ...

A++

Re: [Announce] kernel-power v47
 

Sorry, i read those posts a few days ago and i read them again now and it still makes no sense in the context of what appears to be an increase in battery consumption. Or are you saying that the b27x00 driver causes the battery consumption to increase?

Re: [Announce] kernel-power v47
 

The thing I'm sure of, is the v47 doesn't consume more or less than v46.

You have to find why you see battery drain elsewhere, maybe you have installed another program power consuming since v47 ...

Also, since you have the 'b27x00_battery' loaded, then chances are bme crash like 'joerg_rw' explained ...

After this stated, reading Wiki, bme seem also to control the battery charge, stopping himself during the charge of the battery (if I understand properly).

And I'm pretty sure this step is required to properly calibrate battery charging ...

A++

Re: [Announce] kernel-power v47
 

ironic how colin is now asking ppl to run commands on root to diagnose a problem after vi_* screwed a newb who asked for help for running commands he did not understand on root,

... its not only the newbs who are problematic around here...

ps - thanks colin for trying to geniuely help.

Re: [Announce] kernel-power v47
 

Difference between him and me, is I'm trying to not screw the audience with dangerous commands, at least I'll do my best ...



A++

Re: [Announce] kernel-power v47
 

vi_ there's no need to take this personally. One of the newbs who ran your command did say their phone wasn't booting anymore, shortly after you posted the ash compatible version. Depends on how you define screwed, of course. That might not qualify in your book. *Shrug* Point is, there's no need to get emotionally invested in other people referring back to it. It happened, your reputation is thus changed for those who are aware of it. It probably will come up as time goes on.

Meanwhile, why IS the bq27x00_battery module loading by default? Last I checked all the other kernel power versions included it, but didn't load it by default unless you explicitly added it to the /etc/modules file.

Re: [Announce] kernel-power v47
 

hey im also getting this problem. the battery widget shows like 50 percent and my status bar battery shows 12 percent.
i have tried this

echo "blacklist bq27x00_battery" >>/etc/modprobe.d/blacklist
then i reboot and it is still the same at 50 percent, the widget.
i tried this as well
modprobe -r bq27x00_battery
after this, the battery widget is back to 12 percent but it goes back to 50 percent after reboot.

Re: [Announce] kernel-power v47
 

Here's to hoping it comes up less rather than more. Frankly, I was kind of hoping a mod would clean up the thread.

Re: [Announce] kernel-power v47
 

hey and by the way, THIS BUG, messes with the battery usage? because i have been feeling it. weirdness in the battery life on this bq27x00_battery bug

Re: [Announce] kernel-power v47
 

Yea, so just because You bloat this thread instead of using PM to send him Your best wishes, i should consider You a noob, then steal Your N900 in some back alley and screw it with... Well, screwdriver, cause you're not worth enough to use it, eh?

*facepalm*

The worst thing is when you do something totally out of line, then you keep insulting people that are REALLY contributing to that thread. By the way, i was - just like ndi - hoping that mod will clean this thread - actively hoping, cause reported all insulting/harmful posts - but mods seem to be totally absent here for a quite long time.

---

Now into real thing - i followed Colin's discussion with Pali and I'm not sure if Pali understand why this module should NOT be loaded. Probably thats why, in opposite to other kernel versions, suddenly we got it loaded until blacklisted.

The question is, why yasirrfc got this module loaded even if blacklisted?

yasirrfc, can you please do:

leafpad /etc/modprobe.d/blacklist

And check if module is really blacklisted without any typos there? Please, check triple for typos, cause it seems like that is the case.

Also, it comes to my mind that maybe You wasn't rot when editing and ignored "permission denied" reply, so module isn't blacklisted for real? Anyway, content of Your blacklist will tell us, so please copy it from N900 and publish here.

//Edit

As for Your question - it shouldn't increase battery life, but it is possible (though rather uncommon) that it interfere with charging calibration and charging at all.

By the way, if You feel that battery is draining faster, please measure it with dr_frost graph, or script that monitor current_now - other way it didn't exist (i agree here with vi_ - yea, we can still "work together" even that You (vi_) are d*ckish, you know?).

Most of the times when someone report higher battery drain - and someone report it ALWAYS when we talk about more popular kernels/.programs/tweaks/new N900 leather cases ;) / whatever - it is subject feeling / higher usage cause you play with new options/ something else is draining battery, like new unneeded widget on totally forgotten 7th desktop ;) etc... "Placebo'ish" of that is exactly the same as reports that phone become "snappier".

Re: [Announce] kernel-power v47
 

hey it says:
blacklist mousedev
blacklist bq27x00_battery
blacklist bq27x00_battery
blacklist bq27x00_battery
blacklist bq27x00_battery
blacklist bq27x00_battery
blacklist bq27x00_battery

Re: [Announce] kernel-power v47
 

hey what do u mean by dr_frost_graph? and one more thing. before like overclocking, when my phone showed like 0.4 percent remaining battery, it used to play 3 songs MAX then dead. after overclocking and some battery bq27x00, the battery shows 0.4 percent but it played for 1.5 hours and then 3,4 pic snaps from fcam then it died. overall it was poor battery life but in the end this was the case.

Re: [Announce] kernel-power v47
 

Here's the thing with bq27x00_battery - it is only in power kernel, not in stock kernel. It ALSO reports different values than the normal battery-monitoring module.

This is EXTREMELY visible if you have bq27x00_battery loaded AND have Advanced Power installed (Advanced Power replaces the stock battery status menu applet and entry with a python recode - on that looks for the battery values differently than the stock battery module does).

I have had Advanced Power reporting 0% battery for over an HOUR, with the device running perfectly fine under normal almost constant USE. For whatever reason, bq27x00_battery REALLY underestimates the battery charge left over.

What does this mean? That your N900's battery is NOT discharging any faster than it normally was. The little pictures on the screen that show you battery level are just showing the wrong values. The battery itself is still fine, as I've been suspecting since I first saw you posting about your battery issue and as your last anecdote just proved.

Re: [Announce] kernel-power v47
 

hey yes i do have advanced power installed. so what should i do to fix this bug?

Re: [Announce] kernel-power v47
 

See attached. With phone fairly well charged, I put it in offline mode, locked the screen and went out cycling. Two hours later I heard the beeps of it shutting down (I did not so much as open the keyboard or unlock it in the interim).

I am trying sans bq27x00_battery now, in the hope this fixes the poor performance.

Re: [Announce] kernel-power v47
 

Have you blacklisted bq27x00_battery?

Re: [Announce] kernel-power v47
 

yees read my above posts

Re: [Announce] kernel-power v47
 

I've updated to v47. Also overclocked to 805 and enabled Smartreflex. Sudden reboots happen often. Now I noticed that enabling Smartreflex through QCPUfreq would have something to do with it since i tried 900 frequency and then enabling Smartreflex caused immediate reboot. Anyone similar findings & fix for it?

Re: [Announce] kernel-power v47
 

Try without Smartreflex... been using v47 at 805MHz/ideal and it is very stable here (obviously depends on what your HW can handle)

Re: [Announce] kernel-power v47
 

Just my two cents..

I've never tried SmartReflex, but at least my N900 crashes(*) when using the ideal profile (with 805Mhz max or whatever the default is for ideal).

(*) not immediately, but after leaving it a while in idle I would look at it and it would display a banner saying that an unexpected reboot had happened..

Didn't bother investigating it or raising the voltages or whatever. I use LV now with 850Mhz and works like a charm.

Re: [Announce] kernel-power v47
 

hey it says:
blacklist mousedev
blacklist bq27x00_battery
blacklist bq27x00_battery
blacklist bq27x00_battery
blacklist bq27x00_battery
blacklist bq27x00_battery
blacklist bq27x00_battery

Re: [Announce] kernel-power v47
 

Has there been any consideration of releasing an updated power kernel 47 package without the bq27x00_battery module included?

After blacklisting that sucker here's how my battery graph looks today.

I've zoomed out so you can see yesterday's 3 hours of standby from 75%-ish to 0. And interestingly, overnight it held its charge well also, so my suspicion is not that it drains any faster perhaps, but that the modules messes up the calibration between actual and reported charge, as others have explained. So perhaps the charging process is being told the battery is fully charged when actually it is far from it.

Re: [Announce] kernel-power v47
 

Exactly same scenario, no reboots at all until now.

I wanted to enable smart reflex before V47 with automatic reboots.

Re: [Announce] kernel-power v47
 

hey can u please make me understand battery graph? i cannot understand it here nor on my phone. please help wpuld be appreciated

Re: [Announce] kernel-power v47
 

Well, basically, it's a graph. You enable and/or disable certain figures, until you know what you see. The colors near the despription relate to the colors in the graph. Easy way: deselect all properties, then just select one property, like 'charge'. You'll see a graph describing the battery charge. So if you just recharged your phone, you can now see how long it took, to what value it was charged, etc.

Re: [Announce] kernel-power v47
 

does this go from left to right? and shows the current day use? and going top means charging?
hey and im able to disable the options like charging and cpu in battery graph. how do u do it? and can u tell what is on y axis and on x-axis? ie time? numbers?etc

Re: [Announce] kernel-power v47
 

hey i managed to find out. thanks :) can u please tell me if this application drains battery when its closed?

Re: [Announce] kernel-power v47
 

hey and i didnt find out about the x-axis and y-axis as in what is on both axises. and my above post answer :)
thanks

Re: [Announce] kernel-power v47
 

Here's a great place to ask that and your few previous completely off topic posts: http://talk.maemo.org/showthread.php?t=44831

Also, there's an edit button there for a reason and a Power Search link in the right toolbar. Please don't spam.

Re: [Announce] kernel-power v47
 

Hello Pali and thanks for your work

I try to recompile a custom kernel using the debian tree of your git repository in order to fix an IO issue described and hopefully fixed here by Paul Hartman.

So I took the orig.tar.gz file from apt-get source kernel-power, and added your debian tree

But to be able to use dpkg-buildpackage flawlessly, I have to use commit 87cd160f7535d7ed06db26783d367f54812e700c and edit debian/rules in order to redefine DEFCONFIG to rx51_defconfig and comment line 54 "cp debian/$(DEFCONFIG) arch/arm/configs/".

Eventually, the debs are generated and I install them on the N900: kernel-power, kernel-modules, kernel-bootimg and finally kernel-flasher. Everything run smooth, but when kernel-power is selected at startup, my multiboot installation keep complaining about a missing file required for the flasher.

With your packages, everything goes fine, so I guess I'm missing something here. The only modifications I made to your tree are those to the debian/rules file mentionned above. I also modified the debian/patches/nokia-20094803.3+0m5.diff file to integrate the Paul Hartman fix.

What version of your git repository did you use to build the .deb you distribute ?
Is there something I'm missing ?

thanks

Re: [Announce] kernel-power v47
 

Please post output of commands:
And contents of /etc/multiboot.d/something file with power kernel multiboot option. Maybe it's something simple ;)

Re: [Announce] kernel-power v47
 

multiboot
zImage-2.6.28.10-power47

vmlinuz-2.6.28-omap1

2.6.28-omap1
2.6.28.10-power47-g87cd160-dirty
current

00-Maemo-2.6.28-omap1.item
01-Maemo-kernel-power-2.6.28.10-power47.item
README
examples

ITEM_NAME="Maemo kernel-power 2.6.28.10-power47"
ITEM_KERNEL="2.6.28.10-power47"
ITEM_MODULES="ext3"

I hope so :)

Re: [Announce] kernel-power v47
 

Now tree in my git repo is broken. I'm rewriting debian scripts. For sources use: apt-get source kernel-power.

If you have patch, you can send me it and I apply it to git.

Re: [Announce] kernel-power v47
 

So err.. well, this has propably been asked before, but is there any chance the BFS, wireless-patches and the DSP patches will be incorporated into one fully tested package? I myself think this would bring the power_kernel to another level for a lot of users, but I guess a lot of testing/fixing/etc. is required, and in the meanwhile, this thread gets spammed etc. Basically, some power users need to get together in order to be able to reach this. Any chance whatsoever on this?

Re: [Announce] kernel-power v47
 

Yeah it is simple ;) your kernel image is zImage-2.6.28.10-power47 and your modules directory is 2.6.28.10-power47-g87cd160-dirty and in .item file you have line ITEM_KERNEL="2.6.28.10-power47". I wrote in bold parts which should be the same for correct multiboot boot ;) So you have two choices:
1. simpler and more proper (if you would like to have later "clean" power kernel 47 installed for example): rename /boot/zImage-2.6.28.10-power47 to /boot/zImage-2.6.28.10-power47-g87cd160-dirty and change line in .item file to ITEM_KERNEL="2.6.28.10-power47-g87cd160-dirty"
2. dirtier (you will not be able to use kernel power 47 default files at the same time): rename /lib/modules/2.6.28.10-power47-g87cd160-dirty directory to /lib/modules/2.6.28.10-power47 and don't modify rest.

Re: [Announce] kernel-power v47
 

Actually, I tested the second solution as soon as I discovered that the modules directory name was different from the kernel name. But it ended in a forever rebooting loop.
But your first solution fixed everything: many thanks !

@pali: if you talk about the IO issue, i'm not sure it is something that should be included in kernel-power. It seems to concern very few people, and may be related to the sd card used. I still did not test this fix yet