880144 - Power consumption in flying mode

Status: RESOLVED FIXED

(Firefox OS Graveyard :: General, defect)

Description

[jacocc]

User Agent: Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.22 (KHTML, like Gecko) Ubuntu Chromium/25.0.1364.160 Chrome/25.0.1364.160 Safari/537.22

Steps to reproduce:

I had 38% of battery when I set the alarm and turned the phone in flying mode (and then I went to sleep)


Actual results:

When I woke up in the morning, the alarm didn't work because the phone was turned off caused by empty battery. Battery was 0% and I had to charge it.


Expected results:

Battery consumption should be very low in flying mode. For example, I remember that my old phone used roughly 2-5% of battery consumption during 8 hours in flying mode (not 38%).

Comment 1

[Gene Lian (I already quit Mozilla)]

Alarm will acquire CPU wake lock at the time point when it fires. However, it seems the device has run out of battery before the alarm is firing. This issue might not be related to the alarm stuff.

It might be worth adding a dump at UpdateCpuSleepState() in the hal/gonk/GonkHal.cpp. All the logic go through here if they want to acquire a CPU wake lock.

All the above assumptions are based on the fact of running out of battery could be due to the CPU wake locking. Maybe we have other root causes.

Comment 2

[Gene Lian (I already quit Mozilla)]

This one might be a dupe of Bug 880609.

Comment 3

[jacocc]

Thanks Gene. 
The really problem here is the very high power consumption in flying mode.

Comment 4

[Gene Lian (I already quit Mozilla)]

Yes I know. I mean if at least more than one CPU wake lock is hold and not released, we would have significant power consumption.

Comment 5

[jacocc]

And, is it not possible to release all locks when phone gets flying mode?. I mean, the Mozilla experts are you and I don't know in depth the B2G architecture but from my ignorant point of view I think that.

Comment 6

[jacocc]

By the way, I observe this behavior since I automatically updated my Peak to 2013-06-03 16:56:03 git commit.

Comment 7

[Gene Lian (I already quit Mozilla)]

(In reply to jacocc from comment #5)
> And, is it not possible to release all locks when phone gets flying mode?. I
> mean, the Mozilla experts are you and I don't know in depth the B2G
> architecture but from my ignorant point of view I think that.

Flying mode doesn't mean we have to release all the wake locks. It only implies the device is turning off all the networks, including 3G and wifi. For example, the firing alarm or in-coming calls still need to acquire CPU wake locks to wake up the system from the sleeping mode.

Comment 8

[jacocc]

Ok, I understand. So, when you release all wake locks means that nobody will be able acquire CPU wake locks anymore, even they need it.
By the way, I think in-coming calls shouldn't be possible in flying mode because the telephone network also should be turned off. Dont you think?

Comment 9

[Gene Lian (I already quit Mozilla)]

Ah yes! In-coming call is not a good example here.

Comment 10

[Jamie Charlton (:Jamie_)]

due to the state of the bug that this is linked to bug 881584 has been solved which deals with power consumption i am going to close this as fixed.