[om2008.9/FDOM] Charging or Long Power On = Crashing?

[Korbinian Rosenegger]

Hi

On Mon, 2008-10-13 at 22:20 -0700, Boris Wong wrote:
> I'm working on a theory of prolonged charging leading to the crashing of
> the Freerunner in FDOM/om2008.9. I'm not saying it's definite, but I
> have had some experiences of my phone ending up in kernel panic and
> where the screen would not turn on after overnight charging without
> suspension.
> 
> Here is the scenario:
> I plug in my Freerunner to the wall adaptor, to let it charge overnight.
> I do not use suspension, but I do have the screen turn off
> automatically. I sometimes don't unplug/plug it in again to make the
> charging indicator (red light and the charging symbol on the screen)
> appear. In the morning (6+ hours later), the phone will be in kernel
> panic or the device will be on without the screen (calls and things get
> through, monitor does not come on).
> 
> The questionable theory:
> Is there any chance that, since the kernel is not aware that the device
> is charging (even though the device still charges to full), will create
> crashes or reverse the loop of the battery indicators? I'm aware that
> this is a smart battery which can dish out statistics. If the FR does
> not know that it is charging, but the power is going up, I might think
> that that may cause problems with the logging and power handling from
> the kernel.
> 
> Just a theory, but I really want to get rid of kernel panics and crashes
> after not shutting down for a long time.
> 
> Any other theories floating out there that we might want to consider?
> 
> -Boris


I have the same problem, charging my Freerunner from <40-50% remaining
battery let it freeze almost reliable. And just like you I (normally)
don't use suspend. But sometimes I also have freezes when it runs some
time, and sometimes after rebooting it it freezes again within minutes.
This mostly happened when it got power over USB (500mA over car adapter,
battery was fully charged at that time IIRC), and it was noticeable warm
at the back and front. Another freeze was a few weeks ago on a warm
sunny day when I was outside with the freerunner in my pocket, and it
also got pretty warm. At all times I had GPS turned on (except for the
freezes at charge time. 
So one of my theories is a thermal problem, but a short test was not
able to prove that, although the Freerunner froze after reaching 100%
(started at about 40%), the day before it froze at 25%, charging was
started at 20%. At that time I ran battery.py (available somewhere in
the wiki), also I'm running this program mostly when charging, but the
freezes seem to be more unlikely when not running it.
This brings me to my second theory, something like a race condition in
the battery driver or somewhere, since battery.sh polls the bat status
quite often.

If it should matter, I'm running Om2008 with the latest updates, but my
Freerunner with freezes Om2008 since I have it. I can't tell anything
about Om2007 because I installed Om2008 right after unpacking and eating
my gummi bears ;) Thanks Christoph! :)


cu Korbi


-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 197 bytes
Desc: This is a digitally signed message part
Url : http://lists.openmoko.org/pipermail/community/attachments/20081016/19e539d6/attachment.pgp