Page 3 of 9

Re: [hlte][2014/SEP/02][deCAF] Yank555.lu kernel v1.0c

PostPosted: 04 Sep 2014, 16:15
by Hundsbuah
Running ur kernel for two days now. No single reboot and very good performance and battery life.

Just want to report a bit. But unfortunately I got very very few reboots with the caf kernel. So I don't know if this is a good report :D But until now, no single reboot

Re: [hlte][2014/SEP/02][deCAF] Yank555.lu kernel v1.0c

PostPosted: 04 Sep 2014, 16:30
by capnkrunch
yank555 wrote:Liekly not affect it at all ...

The split is just radio stuff, the rest looks to be still "unified".

JP.


Yes but since I have a hltevzw if will03 only keeps up his gsm builds my radio is going to be broken.

Re: [hlte][2014/SEP/02][deCAF] Yank555.lu kernel v1.0c

PostPosted: 04 Sep 2014, 22:38
by yank555
capnkrunch wrote:Yes but since I have a hltevzw if will03 only keeps up his gsm builds my radio is going to be broken.


Oh crap, yes, that's bad news indeed , then :(

JP.

Sent from my 'proudly eFused' Note 3 (n9005) running W03 deCAF SlimKAT v26 on Yank555.lu deCAF-full W03 Slim v1.0c kernel.

Re: [hlte][2014/SEP/02][deCAF] Yank555.lu kernel v1.0c

PostPosted: 06 Sep 2014, 16:38
by Xavihernandez
Hi @yank555,

Thanks for your work on this!

I have had big battery drain tonight. Wifi was on and I suspect those kernel wakelocks are related to WiFi.
Attached some screenshots.

Re: [hlte][2014/SEP/02][deCAF] Yank555.lu kernel v1.0c

PostPosted: 06 Sep 2014, 18:32
by yank555
Xavihernandez wrote:Hi @yank555,

Thanks for your work on this!

I have had big battery drain tonight. Wifi was on and I suspect those kernel wakelocks are related to WiFi.
Attached some screenshots.


Partial wakelocks are userspace wakelocks (ROM, apps).

From what I see, you lost about 20%, that's something around 600mA, that is huge, that's like 1.5h screen on time !

A wakelock is not enough for that, that means keeping your CPUs on spinng high for hours (given the display way off)...

Don't think this to be wifi related, rather some rogue app.

JP.

Re: [hlte][2014/SEP/02][deCAF] Yank555.lu kernel v1.0c

PostPosted: 06 Sep 2014, 18:39
by Xavihernandez
Thanks for your answer. Yep I know partial wakelocks is userspace , it is the reason why I have also included kernel wakelocks screenie. And we can see numerous of eventsXX wakelock.

Re: [hlte][2014/SEP/02][deCAF] Yank555.lu kernel v1.0c

PostPosted: 06 Sep 2014, 19:26
by yank555
Xavihernandez wrote:Thanks for your answer. Yep I know partial wakelocks is userspace , it is the reason why I have also included kernel wakelocks screenie. And we can see numerous of eventsXX wakelock.


If I remember well, those are wakelocks held by processes not running anymore, though I'm not 100% certain, can't remember exactly if those were called like this...

JP.

Sent from my 'proudly eFused' Note 3 (n9005) running W03 deCAF SlimKAT v26 on Yank555.lu deCAF W03 v1.0c full kernel.

Re: [hlte][2014/SEP/02][deCAF] Yank555.lu kernel v1.0c

PostPosted: 07 Sep 2014, 00:42
by capnkrunch
Xavihernandez wrote:Hi @yank555,

Thanks for your work on this!

I have had big battery drain tonight. Wifi was on and I suspect those kernel wakelocks are related to WiFi.
Attached some screenshots.

eventXX sounded familiar to me and so I thought it was mentioned in T.J. Bender's wakelock guide on xda (http://forum.xda-developers.com/showthr ... ?t=1827676). I checked and it is but as a part of a different wakelock so I'm not sure if his suggestions will help (you can skip dmseg part and start with the "ps" command). Couldn't hurt to try though.

suspend_backoff: This is a difficult one to nail down. Very difficult. To make a very long story short, this wakelock is caused when your phone wants to sleep, but a running process blocks it from doing so. Typically, this is going to be Wifi. Make sure your Wifi is set to stay awake when the phone sleeps. Otherwise, you risk the Wifi going out of its way to keep the phone from sleeping. This can also, unfortunately, be caused by apps, making it hard to pin down.
To fix: Do the Wifi thing, then reboot into recovery and wipe your caches. If it persists after that, you have no choice but to look at a dmesg output to pin it down. Go into a terminal emulator app and type the following commands, one per line:

su
dmesg >/mnt/sdcard/dmesg.txt

Next, transfer the .txt file onto your computer and open in WordPad or a similar program that automatically cuts line (i.e., no Notepad). Search the document looking for the word "event". You should find a lot of strings that look like this:
eventX-XXXX
Where the X's are all numbers. See those last four numbers after the dash? You should see them repeated again and again. Write them down, then go back into your terminal emulator. They are your process ID (PID). Once there, type:

ps

This will show you all processes that have run since your last boot. If you look at the second column from the left hand side of the page, and you should see a sequence of four numbers. These are listed sequentially, lowest to highest, as you come down the page. Scroll around until you find the sequence of four numbers that matches the one you wrote down. The line below those four numbers is the name of the process causing the wakelock. If it's instantly recognizable as an app, delete the app and see what happens. If it's a hardware thing that can easily be fixed, like Wifi, change your settings to accommodate it. If it's com.android.process.acore, I smell a wipe in your future. If you don't know what the process is, don't go screwing with things to find out. You can brick your phone in terminal emulators. Instead, write it down and Google it. Someone has already written about what this process is and what it does. If it's something you can easily fix, go for it. If it's a deeply-embedded system process, I'd just do a full wipe and clean reinstall.


Funny unrelated story, I was looking at yank's open source report card and it said "It seems like Yank555.lu is—or should be—friends with willl0344." I got a kick out of that. Also, coming up on 4 days of no reboots and working HD youtube.

UPDATE: Of course after saying that I got a reboot. It looks different from the old mdss reboots but I think it's still mdss related.

Re: [hlte][2014/SEP/02][deCAF] Yank555.lu kernel v1.0c

PostPosted: 07 Sep 2014, 08:14
by Raver099
Still WIP, this will even get better...

Re: [hlte][2014/SEP/02][deCAF] Yank555.lu kernel v1.0c

PostPosted: 07 Sep 2014, 13:01
by yank555
capnkrunch wrote:UPDATE: Of course after saying that I got a reboot. It looks different from the old mdss reboots but I think it's still mdss related.


Yes, that had to happen, does each time I day something thelike, Murphy's law ;)

Which kernel / ROM were you on ?

Reboot was indeed mdss related, but 'different' indeed.

JP.

Sent from my Nexus 5 running official SlimKAT 7.0 on Yank555.lu v1.0 kernel.