Attend XDA's Second Annual Developer Conference, XDA:DevCon 2014!
5,740,011 Members 38,908 Now Online
XDA Developers Android and Mobile Development Forum

Stock Android 4.3 'Android OS' permanent wakelock

Tip us?
 
andrewf403
Old
(Last edited by andrewf403; 15th August 2013 at 04:25 PM.)
#1  
Senior Member - OP
Thanks Meter 152
Posts: 161
Join Date: Jul 2007
Location: Odessa
Default Stock Android 4.3 'Android OS' permanent wakelock

Looking at some XDA and Reddit posts, small set of users (my Nexus 4 including, unfortunately) experiencing some sort of bug with stock not rooted Android 4.3 after OTA on all range of Nexus devices.

The bug is Android OS holding device awake all the time and not letting it into deep sleep. It looks like it has not huge impact on battery life, but still it is abnormal. Rebooting devices, for me, curing this, but only for a day or two, after a few days wakelock makes its come back.

Unfortunately WakeLock Detector does not show who is guilty, looks like it is in the kernel. I just have bought BetterBatteryStats, and will look at its data tomorrow. What WLD does show is that my Nexus 4 had 0:00 of deep sleep today:





Here is what Android's battary stats looks like:





And here is what is should look like and does look like on my 2012 Nexus 7, that is unaffected by this. And stats looks the same for a few days after I reboot my Nexus 4, before wakelock occurs again:



I have to assume that it has nothing to do with connectivity, because same thing happens even if I disable wifi and mobile data entirely.

Here is a list of topics about the same issue I could found, but they only have questions, not answers.

Android Police's review of 2013 Nexus 7: http://cdn.androidpolice.com/wp-cont...2-16-14-17.png
XDA thread about another Nexus 4: http://forum.xda-developers.com/show....php?t=2388072
Reddit thread about Nexus 4: http://www.reddit.com/r/nexus4/comme...attery_issues/
Reddit thread about 2012 Nexus 7: http://www.reddit.com/r/Nexus7/comme...roblems_since/
UPD: one more thread about 2013 Nexus 7 http://forum.xda-developers.com/show....php?t=2394813
UPD: one more Nexus 4 thread http://www.reddit.com/r/nexus4/comme...ot_of_battery/

Unfortunately, I also do not have any answers (yet). The purpose of this post is gather all the info I've found in one place, and maybe rise the awareness about this.

Update:

Ok. I had BetterBatteryStats on my phone for 30 minutes, and this whole 30 minutes PowerManagerService held wakelock itself. So, it safe to assume that it is an offender.
Attached Thumbnails
Click image for larger version

Name:	2013-08-04 14.48.47.png
Views:	1105
Size:	93.8 KB
ID:	2164359   Click image for larger version

Name:	2013-08-04 14.53.38.jpg
Views:	1118
Size:	36.1 KB
ID:	2164361   Click image for larger version

Name:	Screenshot_2013-08-04-17-54-27.jpg
Views:	1070
Size:	25.7 KB
ID:	2164366   Click image for larger version

Name:	2013-08-04 15.22.11.png
Views:	987
Size:	105.2 KB
ID:	2164398  
The Following 3 Users Say Thank You to andrewf403 For This Useful Post: [ Click to Expand ]
 
mlj11
Old
#2  
Member
Thanks Meter 7
Posts: 82
Join Date: Sep 2010
Hi, I saw your reply on this thread.

I read some speculation somewhere that the wakelocks might be caused by the device continually trying to poll for location, and apparently this behaviour could be corrected by the latest version of Google Play Services v3.2.25. I don't see the wakelocks now after upgrading to it earlier today, however this could also be because I rebooted my phone... Nevertheless I'm still not getting great screen-on time, just about 3.5 hours compared to other people's 5+ hours.

What's your version of the app? If you don't have 3.2.25 you can either download it directly (Android Police has put up a copy of the apk, but it might not be a Nexus 4-specific one) or download a more recent version of Play Store and wait for it to update Google Play Services automatically.

I'll report back here if I notice any more changes in my phone.
 
andrewf403
Old
#3  
Senior Member - OP
Thanks Meter 152
Posts: 161
Join Date: Jul 2007
Location: Odessa
Quote:
Originally Posted by mlj11 View Post
Hi, I saw your reply on this thread.

I read some speculation somewhere that the wakelocks might be caused by the device continually trying to poll for location, and apparently this behaviour could be corrected by the latest version of Google Play Services v3.2.25. I don't see the wakelocks now after upgrading to it earlier today, however this could also be because I rebooted my phone... Nevertheless I'm still not getting great screen-on time, just about 3.5 hours compared to other people's 5+ hours.

What's your version of the app? If you don't have 3.2.25 you can either download it directly (Android Police has put up a copy of the apk, but it might not be a Nexus 4-specific one) or download a more recent version of Play Store and wait for it to update Google Play Services automatically.

I'll report back here if I notice any more changes in my phone.
Location related wakelock are reflected in Android System and Google Services in the Android's battery stats, and as NLP (and some others) in the BetterBatteryStats. They account for about 1 hour of awake in a day. And they actually DO something.

This wakelock unfortunately does not show in Partial Wakelocks statistics in BetterBatteryStats, so I think it occurs somewhere deep in kernel.

If you use Google Now, or some other location based apps like Aviate, 3.5 hours of screen time is great. I do not turn off Wi-Fi, even when it is not available, use Google Now (without location history). 3 hours of screen on is what I expecting as a best case scenario.

Unfortunately, at days then 'Android OS' wakelock happens it nowhere near 3 hours.
 
mlj11
Old
#4  
Member
Thanks Meter 7
Posts: 82
Join Date: Sep 2010
So you're saying that location polling only affects "Android System" and "Google Services" wakelocks. Ok, I didn't know that.

But I did note that my "Android OS" wakelocks have now stopped. Although I cannot say whether the update to Google Play Services / Google Play Store had anything to do with it, or whether it was simply because I rebooted my phone. Do you know for sure that Play Services has no effect in this respect?

And yes, I do use Google Now and also Field Trip.

In any case, I'm also noticing that my battery drains at a much faster rate (as compared to 4.2.2) when I have my phone turned on.
 
Galaxo60
Old
#5  
Galaxo60's Avatar
Senior Member
Thanks Meter 324
Posts: 832
Join Date: May 2012
Location: Mexico City
Quote:
Originally Posted by andrewf403 View Post
Location related wakelock are reflected in Android System and Google Services in the Android's battery stats, and as NLP (and some others) in the BetterBatteryStats. They account for about 1 hour of awake in a day. And they actually DO something.

This wakelock unfortunately does not show in Partial Wakelocks statistics in BetterBatteryStats, so I think it occurs somewhere deep in kernel.

If you use Google Now, or some other location based apps like Aviate, 3.5 hours of screen time is great. I do not turn off Wi-Fi, even when it is not available, use Google Now (without location history). 3 hours of screen on is what I expecting as a best case scenario.

Unfortunately, at days then 'Android OS' wakelock happens it nowhere near 3 hours.
Quote:
Originally Posted by mlj11 View Post
So you're saying that location polling only affects "Android System" and "Google Services" wakelocks. Ok, I didn't know that.

But I did note that my "Android OS" wakelocks have now stopped. Although I cannot say whether the update to Google Play Services / Google Play Store had anything to do with it, or whether it was simply because I rebooted my phone. Do you know for sure that Play Services has no effect in this respect?

And yes, I do use Google Now and also Field Trip.

In any case, I'm also noticing that my battery drains at a much faster rate (as compared to 4.2.2) when I have my phone turned on.
Possible answer here:

http://forum.xda-developers.com/show...php?p=44361765

Hope it helps.

Phone: LG Nexus 4
ROM: JellyBeer
Android Version: 4.2.2
Kernel: Matrix V10
Recovery: TWRP

= = = If you find the information here to be helpful, take a moment to press the Thanks button = = =
The Following 2 Users Say Thank You to Galaxo60 For This Useful Post: [ Click to Expand ]
 
andrewf403
Old
#6  
Senior Member - OP
Thanks Meter 152
Posts: 161
Join Date: Jul 2007
Location: Odessa
Quote:
Originally Posted by Galaxo60 View Post
Possible answer here:

http://forum.xda-developers.com/show...php?p=44361765

Hope it helps.

It'll help with Google Services drain, I think. Anyway Google Play Services just updated for me (with Android Device Manager), hope it'll fix Android OS drain too. Hope is not high, though )
 
flintdragon
Old
#7  
Senior Member
Thanks Meter 33
Posts: 326
Join Date: Aug 2010
I've recently have this issue as well with "Android OS". Luckily I can still make it through the day easily but annoyed this kind of stuff still happens. I got the latest Google Play Services yesterday and still have this issue today after a full charge. Will reboot and see what happens I guess.
 
steven676
Old
#8  
Senior Member
Thanks Meter 2350
Posts: 579
Join Date: Nov 2011
Quote:
Originally Posted by flintdragon View Post
I've recently have this issue as well with "Android OS". Luckily I can still make it through the day easily but annoyed this kind of stuff still happens. I got the latest Google Play Services yesterday and still have this issue today after a full charge. Will reboot and see what happens I guess.
I've been seeing this on my Galaxy Nexus with AOSP 4.3 as well. It looks like this patch might be relevant: https://github.com/CyanogenMod/andro...d0eacfcd2003b9
 
andrewf403
Old
#9  
Senior Member - OP
Thanks Meter 152
Posts: 161
Join Date: Jul 2007
Location: Odessa
Quote:
Originally Posted by steven676 View Post
I've been seeing this on my Galaxy Nexus with AOSP 4.3 as well. It looks like this patch might be relevant: https://github.com/CyanogenMod/andro...d0eacfcd2003b9
It looks relevant indeed. Although, I would not call the drain "rapid". I'm now looking close to my device statistics and try to pinpoint exact moment this lock occurs to find the cause.

If it is indeed caused by concurrent condition in PowerManagerService there is not much to do, but flash another, patched, ROM. But I eager to stay on stock.
 
morgante
Old
#10  
Junior Member
Thanks Meter 1
Posts: 7
Join Date: Mar 2011
Location: Tucson, AZ
Also seeing this with my Nexus 4. Reboot doesn't do much at all for me.

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes