Active Topics
-
Extra softwares in Sailfish using CLI, repositories, etc (117)
to SailfishOS by nieldk - 1 day, 8 hrs ago -
Firefox with Leste (6)
to Maemo 7 / Leste by teroyk - 2 days, 5 hrs ago - more...
misiak |
2012-07-30
, 10:14
|
|
Posts: 804 |
Thanked: 1,598 times |
Joined on Feb 2010
@ Gdynia, Poland
|
#71
|
Originally Posted by peterleinchen
Best would be to include an /etc/init.d script in kp package which would set this symlink on every phone reboot, imagine having 2 or 3 versions of kp installed at once, then one would have e.g. /opt/packet-injection-modules/2.6.28.10-power51 and /opt/packet-injection-modules/2.6.28.10-power52 , sometimes booting kp 51, sometimes kp 52 - it should be set like /lib/modules/current - on every reboot. Or, maybe even better, just create a symlink to /opt/packet-injection-modules in e.g. /lib/modules/2.6.28.10-power51/packet-injection-modules, then they could be accessed under path /lib/modules/current/packet-injection-modules - any comments about that from someone more experienced in kernel development? would it be bad, good?
|
|
2012-07-30
, 10:43
|
|
Posts: 1,808 |
Thanked: 4,272 times |
Joined on Feb 2011
@ Germany
|
#72
|
@misiak,
I think the best would be simply to copy (not symlink) the modules on /lib/modules/whatever-kernels-you-want
I don't see the point in having kernel modules under /opt or Mydocs or who knows where.
I think the best would be simply to copy (not symlink) the modules on /lib/modules/whatever-kernels-you-want
I don't see the point in having kernel modules under /opt or Mydocs or who knows where.
| The Following User Says Thank You to reinob For This Useful Post: | ||
|
|
2012-07-30
, 11:49
|
|
Posts: 804 |
Thanked: 1,598 times |
Joined on Feb 2010
@ Gdynia, Poland
|
#73
|
Originally Posted by reinob
MyDocs would be indeed stupid, as this one gets unmounted when you connect N900 to computer as mass memory, but I don't agree that it should be in /lib/modules instead of opt. Many people are not using them and they are >500kb and, as not essential for running kernel with working wifi, I agree with maintainer's decision to place them in /opt. See official packaging guidelines - these modules are now part of kernel power, but I somehow feel it's wrong to place these 500kb in rootfs (and let's not forget they started as a separate package, this is only first stage of their inclusion in official kernel power packages).
@misiak,
I think the best would be simply to copy (not symlink) the modules on /lib/modules/whatever-kernels-you-want
I don't see the point in having kernel modules under /opt or Mydocs or who knows where.
I still believe they could be symlinked, not placed in /lib/modules/proper-kernel-version/packet-injection-modules , but that's just my oppinion based on stuff I wrote earlier in this post.
| The Following User Says Thank You to misiak For This Useful Post: | ||
|
|
2012-07-30
, 11:53
|
|
Posts: 186 |
Thanked: 79 times |
Joined on Feb 2010
|
#74
|
Originally Posted by majaczek
Either I broken my charging chip/batterry or there actually is a bug/incompability in KP51r1.
I upgraded in last minute, and nokia default batterry meter shows broken "almost empty", while bnf confirmed it is still 33% full (it perhaps uses i2c method once when run and closes).
Seems some new driver version miraclously conflict with bme (and now checked my fool steps and were none).
Anyway it seems even faster than before perceptfully. Hope that confused bme wouldn't decide to shutdown.
Edit: it shows "calibration needed : 0" and measured batterry capacity changed, but I didn't run the calibration script, could it have calibrated naturally via long charging or it indicates something broken here (it could broke something after me beeing so blunt to allow 50 celsius and not shutdown instantly to prevent overheat).
It shows "almost empty" again and again, will ccheck after next reboot.
I'm also experiencing battery problem. I was charging it few hours and battery stayed on the same level. after replugging the charger i was able to rechearg battery. Now I'm experiencing increased power drain.
I also installed package with gui for USB host mode, nothing else has changed in my system
Edit:
After a playing a while with different host mode options(and no device connected to usb) i pluged n900 to charger and led was blinking with little almost red light, not normal one - small and barely visible. There is something definitly wrong with new kernel.
Last edited by hardkorek; 2012-07-30 at 12:45.
|
|
2012-07-30
, 12:54
|
|
Posts: 67 |
Thanked: 32 times |
Joined on Oct 2011
|
#75
|
About my batterry measure bug.
It seems some nokia stuff (or were it CSSU?) couldn't read batterry info while charging behaves properly, and bnf was not fooled by this bug.
Both default statusbar widget and custom desktop widget which is set to mimic original batterry meter shows almost empty (the second is supposed to get more detailed info and is showing 5mAh since i felt asleep).
Bnf seems to read batterry info most directly and instantly and shows that overnight it discharged from about 400mAh to about 300mAh, and now is saying properly charging while the charger is connected.
It also seems bme is not fooled enough to force shutdown. It got 40 celsius now while charging.
I really should test my device with powertop... What was proper command to redirect powertop result to file (it would be easier to copy here, and I'm confused with 4 different "send to file" operators in linux shell... Anyway I'm too lazy to rewrite and prefer copy the info by the device)?
It seems some nokia stuff (or were it CSSU?) couldn't read batterry info while charging behaves properly, and bnf was not fooled by this bug.
Both default statusbar widget and custom desktop widget which is set to mimic original batterry meter shows almost empty (the second is supposed to get more detailed info and is showing 5mAh since i felt asleep).
Bnf seems to read batterry info most directly and instantly and shows that overnight it discharged from about 400mAh to about 300mAh, and now is saying properly charging while the charger is connected.
It also seems bme is not fooled enough to force shutdown. It got 40 celsius now while charging.
I really should test my device with powertop... What was proper command to redirect powertop result to file (it would be easier to copy here, and I'm confused with 4 different "send to file" operators in linux shell... Anyway I'm too lazy to rewrite and prefer copy the info by the device)?
| The Following 3 Users Say Thank You to majaczek For This Useful Post: | ||
 |
|
2012-07-30
, 16:25
|
|
Posts: 101 |
Thanked: 73 times |
Joined on Oct 2011
@ Lithuania
|
#76
|
Originally Posted by bennypr0fane
http://talk.maemo.org/showpost.php?p...40&postcount=1 If you read this carefully you should find answers for all your questions.
you mean install u-boot-update-bootmenu? or update it?
no, write in terminal u-boot-update-bootmenu
I wasn't aware there was a newer version of U-Boot. I didn't get a notification in HAM - or after running apt-get update.
How do I upgrade? After I do, do I still need to make those changes to kp51, or should it work as-is then?
 |
|
2012-07-30
, 16:56
|
|
Posts: 28 |
Thanked: 24 times |
Joined on Feb 2010
|
#77
|
The phone was turned off due to battery being empty, i plugged the wall charger to charge the phone while it was still turned off ,placed it in a cool place,i did not turn it on at all, after the battery was charged (green LED) i left it connected to the charger for a about 2 hours,
after that time when i picked it up it was really hot, dangerously hot!
the battery was hot and so was the left side of the phone,
i think i never left it charging while it was turned off before to notice this, now i'm doing all of those test (have a bit of free time) i'm noticing strange things
ok so common sense says the phone is off nothing is running, right?
the charging chip should stop charging the battery when it detects full charge (either by measuring temp or by the cell reaching 4.2V like any good lithium battery charger to prevent over charging the battery and reducing it's life or in some cases if the battery protection circuit fails the battery may catch on fire)
so why it is getting hot after it charged the battery, it should cool down, not keep getting hot!
i have KP51 installed for a while now and i'm noticing strange things
after that time when i picked it up it was really hot, dangerously hot!
the battery was hot and so was the left side of the phone,
i think i never left it charging while it was turned off before to notice this, now i'm doing all of those test (have a bit of free time) i'm noticing strange things
ok so common sense says the phone is off nothing is running, right?
the charging chip should stop charging the battery when it detects full charge (either by measuring temp or by the cell reaching 4.2V like any good lithium battery charger to prevent over charging the battery and reducing it's life or in some cases if the battery protection circuit fails the battery may catch on fire)
so why it is getting hot after it charged the battery, it should cool down, not keep getting hot!
i have KP51 installed for a while now and i'm noticing strange things
|
|
2012-07-31
, 08:36
|
|
Posts: 67 |
Thanked: 32 times |
Joined on Oct 2011
|
#78
|
Yeah strange things. I think those are older rare indirect outcoming of bugs. Just now it happens more frequently. There is a deep bug related to charging and kernel which seem to be here since kp49, but until the update to kp51 only 2-4 occurences were catched, including older mine with borked video player when battery were near to entry (reboot and charge up cured the bug temporary then). It seems one of these bug which are conditional bomb (happens only when specific conditions happens, I refer to bug/worm behaviour, not the real bomb). I guess latest CSSU update make such condition much more common AND the bug worsened in kp51. Please remind me proper powertop to file comand because I want to post it here for pali or freemangordon checking if it is good or problem (I'm not that good at reading powertop so I have to post it whole and hope to copy-paste). I have problem at isolating my problem to single cause (there can be at least two causes at once and I'm bad at not doing something coindeincialy).
|
|
2012-07-31
, 08:42
|
|
Posts: 1,808 |
Thanked: 4,272 times |
Joined on Feb 2011
@ Germany
|
#79
|
Originally Posted by misiak
Fine. I just don't see why some non-essential modules come with kernel-power-modules (and are installed under /lib), like xfs (500kb), nfs, ipv6, cifs, reiserfs, ntfs, hfsplus, jbd2, jbd, etc. (to mention only the bigger ones), but at the same time keeping the injection modules at a kernel-independent directory (and not even in rootfs so they could be loaded at boot).
MyDocs would be indeed stupid, as this one gets unmounted when you connect N900 to computer as mass memory, but I don't agree that it should be in /lib/modules instead of opt. Many people are not using them and they are >500kb and, as not essential for running kernel with working wifi, I agree with maintainer's decision to place them in /opt. See official packaging guidelines - these modules are now part of kernel power, but I somehow feel it's wrong to place these 500kb in rootfs (and let's not forget they started as a separate package, this is only first stage of their inclusion in official kernel power packages).
I still believe they could be symlinked, not placed in /lib/modules/proper-kernel-version/packet-injection-modules , but that's just my oppinion based on stuff I wrote earlier in this post.
But yeah, it's a matter of taste.
| The Following 5 Users Say Thank You to reinob For This Useful Post: | ||
|
|
2012-07-31
, 20:04
|
|
Posts: 186 |
Thanked: 79 times |
Joined on Feb 2010
|
#80
|
My charger is getting really hot while charging. I'm going back to kp50.
 |
| Tags |
| bq2415x_charger, bq27200, kernel-power |
«
Previous Thread
|
Next Thread
»
|
All times are GMT. The time now is 14:46.