ubuntu-phone team mailing list archive

[Ricky Chan <djghosie@xxxxxxxxx>]

Hi Michael,

I've trialed some replication method for the heating up / battery drain
issue. Having checked in terminal, the two services are taking up most of
the CPU usage:

1. Maliit-server (~35%)
2. qml-phone-shell (~35%)

In order to reproduce the battery drain, I tried the followings:

1. Let the phone to connect to a registered wifi connection
2. Move to an area where there's no wifi network which is registered by the
phone.
3. Then slowly but surely the phone will heat up.

It seems like when the current connection is lost, the phone will try
frantically to find another connection.

It's  happening to many previous builds. I'm currently on build #153 using
Mako.

Hope this helps.


Ricky


On Thu, May 30, 2013 at 12:08 AM, Michael Zanetti <
michael.zanetti@xxxxxxxxxxxxx> wrote:

> Hi,
>
> On Wednesday 29 May 2013 23:09:38 Ricky Chan wrote:
> > There had been more than one instance where my phone gets hot and battery
> > was draining at a rate of 1% per 2 minutes. I would really like to help
> out
> > in sorting these battery draining issue and I know in order for the devs
> to
> > analyze the problem, a logcat would be the best information to relay to.
> > The problem is, I do not know how I could pull it from my phone.
> >
> > So could anyone please let me know the procedures I need to do in order
> to
> > extract the information from my phone.
>
> I think the first thing to do is top log in the phone via ssh and call
> "top" to
> see which process goes wild.
>
> logcat, like Daniel suggested can be useful in certain circumstances too,
> but
> I expect the issue to be on the ubuntu side of things. So the ssh + top is
> more useful as a first step.
>
> That said, lately I have experienced issues where hud-service consumes
> around
> 80% CPU. I think especially if the shell gets restarted for some reason,
> hud-
> service seems to dislike that. Still not really sure how to reproduce
> tho...
>
> Br,
> Michael
>