Database Develop

Omapscaler: auto over/under cloking

http://www.oliy.co.uk/download.asp?dir=\software\OmapScaler
it is working fine in my wizard but some ppl complain about causing soft resets or not working at all:
about resets: ver .11 have a "max clock step" option, try to reduce it, my wizard works fine with 50 but some ppl r using 25 or even 20...
about not working at all:
1- unpack
2- copy to windows folder in device
3- copy also omapclock.exe to windows folder in device
ftp://xda:[email protected]/Wizard/Misc/OmapClock_v0.2.zip
4- create an icon or run manually
5- set ur desired options and press save
6- copy the .lnk created to windows\startup so it stats with windows everytime (but test it carefully before doing so or u can end with a neverending soft-reset cycle)
my config (working)
- overclock marked and 240mhz (some ppl report soft resets over 220,
reduce this number also)
- above 70%
- underclock marked and 120mhz
- below 30%
- default speed 214mhz (this is the speed center, used if u dont mark
overclock or underclock)
- max clock step 50 (if u have soft resets reduce this number)(mine
work fine at 50 but resets over 70 as example)
- update period 2000 i guess this is the time between system
monitoring of speed in miliseconds, but please correct me if im wrong
- clock threshold 2 (i guess this is the step between clock changes,
but im not sure what is a period...maybe an internal clock tick?)
ok, i tested this one, if u increase it it takes more time between clock changes,
another thing to try to solve soft-resets, i tested 3 and 4...and takes more time
between changes...
some comment: it speeds up depending on the cpu usage but for some reason it dont get along fine with agile messenger in my setup...if agile is connecting it appear as if the device hang but in fact it resumes operation in 10 secs or so...so try to test it first without loaded applications...then load 1 by 1 and test...
lets talk about this and try to solve problems for the ones that have them...
btw...im using molskis rom and working flawlesly...

hi, there,
I followed your instruction fully.
however, when I clicked Omapscaler, I got an "error exception".
I wanted to show you the details..... but I couldn't "copy & paste" the message.
Noted that, I have been running smartskey + Omapclock at 252 stably for long time, and now with Risidoro 2.26 ROM.
hope to get it working.......
thanks,
ecchi

Cool, I'll give this a try ASAP and report back

Activesync quits
I installed the Omapscaler. It seems to work fine. I have it set fairly conservative - middle at 200, underclock at 170 and overclock at 230.
If I am connected to activesync, it quits after a few minutes of idle time. Have you noticed this?
-Kal.

Mine does the same, but in my case it puts a red cross over the green circle, it didnt sync anymore but i can explore the wizard files...and it does it also without the omapscaler, i guess that that have some relation maybe with the proccessor clock losing sync with the pc...dunno...but in my case is not related to omapscaler...but related with the overclock...

Yeah, I saw the same problem with ActiveSync and rom updates.
Just disable the omapscaler, run omapclock and set the speed to 200 or so, then try again.
Oliy, great program, thank you

Again...in my case i tested activesync just using omapclock (NO omapscaler) and does exactly the same...

Since I got the "error exception" with Omapscalar.witrh my current ROM (risidoro), I have started trying XCPUScalar. However, it is not any more stabler as some ppl had mentioned.
Bottom line:
Omap + smartskey ==> most stable, tried at 252, but about 1 day of battery only.
XCPUscalar ==> fixed at 232, stable. fixed at 247 reboot sometimes. Autoscaling b/t 117 and 247 reboot 50% of the time when you launch/close/activesync. I am testing now autoscaling between 117 and 232. Hope it will work.

soft resets with max clock step at 10
I had the max clock step at the default 50, and got a soft reset. Then, per the advice on this thread, I reduced it all the way down to 10. It still soft resets. I have top speed at 230, default at 200, and bottom at 170.
So, I am giving up on this program.

just a q. it stays at least some seconds or die instantly? omapclock alone works?

Yes, it stays up for a while. It seems like it stays up until it goes into standby mode, or I put it in standby manually. A bit after that, it soft resets.
I have not used Omapclock by itself. I have never felt the need for it. But, something that automatically scales up and down to suit the load seems very useful to me; so I tried it. I am willing to keep trying it, if there are bug fix versions.
-Kal.

Maybe ur problem is with the overclock per se, not omapscaler...try using the device with omapclock fixed...if ur device resets then omapscaler is not the faulty...also, what rom do u have? maybe we can pinpoint the problem to a specific rom or set of roms...

hmm... the problem cannot be overclocking, because it goes into the soft reset directly from standby mode. I would think it would be *underclocking* when it is on standby mode, no? Let me do some experimentation with both overclocking and underclocking using smarskey with omapclock, and see what I can figure out. But somehow I doubt that is the issue.
I have standard stock Cingular 2.25.11 rom.
-Kal.

I have been running Omapclock at 230MHz for a while now, with no problems. The Omapscaler would have definitely reset by now.

And what device exactly and with what rom do u have?

It's a Cingular 8125 device, ROM version 2.25.11.1 WWE (stock ROM that came with the device from Cingular).
Omapclock has been running for almost a day now, with no problems, so the problem must be with the scaler.
-Kal.

A theory...
In case the developer is monitoring this thread:
I saw in the other thread on this forum that someone reported that XCPUScaler has the same problem as I have with resetting when in standby mode on a Cingular 8125 with stock rom as well. It may be that these units have some kind of daemon processes that run even while the unit is standby (I know for instance that my unit retreives email when in standby mode), and that causes activity that pushes the scaler to try to scale up. And maybe the unit (or the rom os version) doesn't like the processer speed being mucked around when in standby mode. Or maybe, even that the API call to sense the CPU load craps out when in standby.
This is a conjecture of course, but it seems to me that there can't be very heavy activity during standby. So, could the developer please try to disable scaling when in standby mode? This may solve the problem. Or maybe, just make that an option that people with this kind of a problem can select.
Just a suggestion...
-Kal.

error exception log
I can run omapclock with smartskey @252 stable.
But when I reconfig my wizard to use omapscaler, I got the following exception on launch:
=============================================
OmapScaler.exe
Exception
CompareInfoRef::.ctor+0x35
CompareInfoCache::GetCompareInfoRef+0x9c
CompareInfo::GetCompareInfo+0xc
CultureInfo::get_CompareInfo+0x19
DataTable::RecomputeCompareInfo+0x11
DataTable::Reset+0x6
DataTable::.ctor+0x113
DataTable::.ctor+0x6
ConfigDB::setupConfigDB+0x32
ConfigDB::Load+0x2a
ConfigManager::setupConfigDB+0x14
ConfigManager::Get+0xc
ConfigManager::get_ocLoad+0x15
CpuClocker::readConfigSettings+0xb
CpuClocker::.ctor+0x56
MainForm::MainForm_Load+0xb
Form::OnLoad+0x15
Form::_SetVisibleNotify+0x1d
Control::set_Visible+0x1f
Application::Run+0x7
MainForm::Main+0xa
==========================================
I have a Molski Rom WWE (30MB) G3 Wiza, installed with CE Star (language tool), and some 7 today plug-ins.
any idea?

astromusic said:
In case the developer is monitoring this thread:
I saw in the other thread on this forum that someone reported that XCPUScaler has the same problem as I have with resetting when in standby mode on a Cingular 8125 with stock rom as well. It may be that these units have some kind of daemon processes that run even while the unit is standby (I know for instance that my unit retreives email when in standby mode), and that causes activity that pushes the scaler to try to scale up. And maybe the unit (or the rom os version) doesn't like the processer speed being mucked around when in standby mode. Or maybe, even that the API call to sense the CPU load craps out when in standby.
This is a conjecture of course, but it seems to me that there can't be very heavy activity during standby. So, could the developer please try to disable scaling when in standby mode? This may solve the problem. Or maybe, just make that an option that people with this kind of a problem can select.
Just a suggestion...
-Kal.
Click to expand...
Click to collapse
i have read around that there r several levels of standby, full, partial, etc...maybe he only need to modify it to work only on "full on"...
but maybe we r both wrong thinking about it a lil longer...maybe he already thought about that and omapscaler dont work when on any kind of partial standby, in any case the device is in standby, u r not going to notice if it takes 20 more seconds downloading the mail as example...
and btw, xcpuscalar didnt worked for me also, random softresets, but omapscaler worked like a charm (now im using batterystatus solution and also works fine)

[Q] Screen on Delay?

Hi,
I am having an issue where the screen takes a solid 1-2 seconds to power on after the phone has been sleeping/idling. Sometimes it's not so bad but it's becoming quite annoying, and sometimes I think the screen turns off again before it can show the lockscreen and I have to pull out the battery to fix it!
Cm7.1 stable w/cm7 kernel. Was happening on stock too.
Does anyone know a way around this? Widgetlocker doesn't seem to fix it
Thanks.

i am having a similar issue on Topogigi's v1.9, no issue on his 1.4 version, very annoying!

noticed with Topogigi's ROM since v1.9:
This issue is even worse when a call comes in: The fon will vibrate about 4-5 seconds before the screen turns on to shows the caller. Until then you can not see whos calling.

On Horsepower rc1 kernel and topo 1.9 I can't say I have noticed this issue.

Used fon was already running on v2.0 with RC1 kernel.
Also did a clean flash of the rom again to eliminate possible issues of other apps running. But this has no effect on the problem. It still persists: When a call comes in, the fon vibrates up to 5 seconds before the screen turns on. Even then, you first see the unlock screen and about a second later the screen with the incomming call. Reproducability 10/10 times. Notice that vibration for incomming calls has to be set as "always on" in settings to reproduce problem!
Extremely annyoing because call can not be answered or dismissed while fon keeps ringing! Already missed calls because called hung up before screen unlocked.
Don't know if its a kernel or rom issue...

Advice
The best way for you not to have troubles is to pick a stable rom and flash it!

If you are on Topogigi ROM, This has already been explaine in the FAQ.
Topogigi said:
Ladies and gentlemen, guys and gals, experts and noobs, here you will find a F.A.Q. collection for our ROM. Hope it will help.
Q6.But I'm still experimenting lags with overclock daemon on wakeup: how can I tune it to fit my needs at best?
A6.One of the best features of overclock daemon is that you can tune it as you like. If you don't want to see lags on wakeup, just raise a little the max frequency for sleep profile. If you want to mock SGS2 owners with astonishing Quadrant scores, raise the frequency to the max for wake profile and your phone will turn into a raging bull. It's up to you. Your interest is raising? Let's see how to do it:
Fire up ES file explorer (you will find it in my rom beginning with v. 1.6 release), open settings and check "Use Root Mode" and "Mount Filesystem" (you need to do it only once). Then exit settings and hit the highest icon on the left: you will be in the root directory of your phone. Navigate to /etc/andrev_oc, you will find six files there:
-sleep_gov: indicates which governor to use during sleep mode ("performance" or "powersave")
-wake_gov: governor for awake state ("performance" or "powersave")
-sleep(or wake)_max: max frequency to allow during sleep or wake (see frequencies below)
-sleep(or wake)_min: min frequency to allow during sleep or wake (see frequencies below)
Here there are the frequency values available on current kernel (Neoblaze 28092011):
1408000
1216000
1100000
1015000
800000
503000
389000
216000
Make your changes to the files listed above and save them.
Since overclock daemon is already up and running and it monitors these files at every screen state change, it should be enough to hit the sleep button, wait some seconds, and hit it again to wake your phone. Your new settings will be applied immediately.
You may want to check the actual CPU frequencies by using OS Monitor or SystemPanel App (download the freely from the Market).
Click to expand...
Click to collapse

Look in my signature. Download the Andrev OC app.
Set asleep maximum frequency to 800 or 1016.
Problem solved.

There is also an app for the Topogig ROM http://forum.xda-developers.com/showthread.php?p=18273457#post18273457
---------- Post added at 08:22 PM ---------- Previous post was at 08:21 PM ----------
Hahaha, same timing

Problem solved here like described. Thank you guys!
p.s.: I really appreciate the work of topogigi and spica for roms and kernel. They run 1000x better than the stock roms so I would never follow an inappropriate advice to fall back on those. Must be the urge to reach the amount of 10 posts for some ppl to get access to the dev area...

[Q] Small lags here and there

Hello guys, how are you?
I would like to ask if you also have the same "problem" with your nexus 6. First of all, I come from an OnePlus One which was very smooth . I used to run CM + Franco Kernel but even stock CM (11 or 12) run nicely on it.
Now I'm using a Nexus 6 with the stock rom and the latest version of franco kernel (r26) but, for example, when I swipe the multitasking windows, I can see clearly a lag especially with the last one. Scrolling up and down is also laggy sometimes. Are you experiencing the same? I am running without the encryption and I already tried to disable Fsync; the launcher is the Google one.
Do you have any other sugestions?
Regards,
Rafael

I had this issue on Franco kernel. I went back to stock and the lags went away and battery life almost doubled

I have small lags here and there (sometimes seconds long freezes) but that's mostly due to me messing with kernel settings, trying out custom kernels and nightly ROM builds.

where you able to find out what was causing the lag and how to fx it? im having the same issue.

angourakis said:
Hello guys, how are you?
I would like to ask if you also have the same "problem" with your nexus 6. First of all, I come from an OnePlus One which was very smooth . I used to run CM + Franco Kernel but even stock CM (11 or 12) run nicely on it.
Now I'm using a Nexus 6 with the stock rom and the latest version of franco kernel (r26) but, for example, when I swipe the multitasking windows, I can see clearly a lag especially with the last one. Scrolling up and down is also laggy sometimes. Are you experiencing the same? I am running without the encryption and I already tried to disable Fsync; the launcher is the Google one.
Do you have any other sugestions?
Regards,
Rafael
Click to expand...
Click to collapse
phyba said:
I had this issue on Franco kernel. I went back to stock and the lags went away and battery life almost doubled
Click to expand...
Click to collapse
brando56894 said:
I have small lags here and there (sometimes seconds long freezes) but that's mostly due to me messing with kernel settings, trying out custom kernels and nightly ROM builds.
Click to expand...
Click to collapse
aniel300 said:
where you able to find out what was causing the lag and how to fx it? im having the same issue.
Click to expand...
Click to collapse
@ ALL who have this problem:
i had it too, and i was thinking that i have been the only one..went crazy because i didnt find the reason for this.
Solution:
1) You need the paid version of FKU
2) Open FKU, go into "CPU Manager"
3) Go into Governor Control
4) click on "input_boost_duration"
5) Change the value to: 1000000
6) Turn the button on for set on boot (so this value will be saved and applys again after a reboot)
7) then reboot the device, and voila
explanation:
on the stock kernel there is following setting:
if you touch the display, the Cpu will stay at 1497 Mhz for 3 Seconds and thats the reason why stock is smooth!
Franco changed this value for saving battery. his standard value is 0.04 Seconds on 1497 MHz, then it goes down to 300 if the cpu dont has to handle something.
So for example: if you scroll in the settings, and the CPU frequency goes so quickly down (after the 0.04 secs) there will be stutters
and with the 7 steps i mentioned above, you change the boost duration to 1 second. This means the device will not lag, because the CPU stays for 1 sec on 1497 Mhz and the battery impact isnt that big!
Please try this out and respond to me if it helped

aniel300 said:
where you able to find out what was causing the lag and how to fx it? im having the same issue.
Click to expand...
Click to collapse
Turns out, most of my problems were caused by the "Deep Darkness" CM theme. Builds of BlissPop and BenzoROM would be fine for a little while and then start to screw up, getting pissed off I would boot into CleanROM which didn't have any themes applied (I use MultiROM) and everything would work as expected for hours, no lags at all! I decided to switch to another theme in BlissPop to test my theory and voila no more lag!

GregNexus said:
@ ALL who have this problem:
i had it too, and i was thinking that i have been the only one..went crazy because i didnt find the reason for this.
Solution:
1) You need the paid version of FKU
2) Open FKU, go into "CPU Manager"
3) Go into Governor Control
4) click on "input_boost_duration"
5) Change the value to: 1000000
6) Turn the button on for set on boot (so this value will be saved and applys again after a reboot)
7) then reboot the device, and voila
explanation:
on the stock kernel there is following setting:
if you touch the display, the Cpu will stay at 1497 Mhz for 3 Seconds and thats the reason why stock is smooth!
Franco changed this value for saving battery. his standard value is 0.04 Seconds on 1497 MHz, then it goes down to 300 if the cpu dont has to handle something.
So for example: if you scroll in the settings, and the CPU frequency goes so quickly down (after the 0.04 secs) there will be stutters
and with the 7 steps i mentioned above, you change the boost duration to 1 second. This means the device will not lag, because the CPU stays for 1 sec on 1497 Mhz and the battery impact isnt that big!
Please try this out and respond to me if it helped
Click to expand...
Click to collapse
BEEN SEARCHING FOR A SOLUTION TO THIS FOR DAYS NOW AND THIS WORKED! THANKS! :good:

Custom ROM breaking serious GPS bug. Help needed.

My Original Post:
Phone runs hot. No deep sleep. High cpu usage from AndroidSystem app. Sometimes screen doesn't turn off after pressing power button. Phone runs slow.
All of this happen after any GPS data request from any app that can use device only gps data.
When GPS is off with a fresh boot, there is no problem.
When gps is turned on but not any data requested from it, there is no problem.
If you turn on device only gps and start using a gps app. CPU usage goes abnormally high and never falls down even after force closing the gps app and gps itself. Only a reboot turns the phone to normal.
Another thread I've found describing the same problem without solution:
http://forum.cyanogenmod.org/topic/94185-gps-bug-phone-unstable/
I've tested:
cm-11-20140916-SNAPSHOT-M10-p880
cm-11-20150517-NIGHTLY-p880
cm-11-20150524-NIGHTLY-p880
Euphoria-OS-1.1-p880-UNOFFICIAL-20150516
Euphoria-OS-1.1-p880-UNOFFICIAL-20150529
aosplp-20150423-p880
MK44.4-p880-150131-HISTORY (Mokee History)
LS-KK-v3.2-2014-10-14-p880 (Mourta Ed. with kernel 3.4)
LS-LP-v4.0-2015-03-27-liquid_p880 (Mourta Hybrid Kernel 3.4)
BeanStalk-4.4.4050-20140905-p880CMAD
Kernels:
Optimus_Prime_CM11-V3.0
iodakX3_v10_CM
WWJB_v009_anykernel-20140429
WWJB_v010_anykernel-2014-12-24_1303
I've tried listed roms and kernels with required wipe operations switching between them.
I've rebooted several times after installation and didn't give up in first boot.
I have CWM Official v6.0.4.5 as recovery.
Tried different governors and schedulers. No luck.
I'm using the phone with CM 10.1.3 Stable & WWJB_v009_anykernel-20140429 and I don't have this problem with it.
GPS apps never use high amount of CPU and when you turn off GPS apps, CPU usage falls to idle levels.
I've flashed same WWJB_v009_anykernel-20140429 kernel over a fresh CM11 M10 install and gps bug is still there.
I think this is not about kernel or rom but something different. I can't move forward from Jellybean.
Any idea what's causing this?
Any clues, ideas or suggestions are appreciated.
Click to expand...
Click to collapse
Update: I still have this annoying problem and I'm trying to solve this.
I need you to test this on your devices and tell me the result. Please.
How to reproduce the bug
1- Prepare your SD card with the custom Rom(Kitkat or later preferably with built-in file manager to install apks) to flash and apk files of F-Droid Satstat, CPU-Z for monitoring the cpu load.
2- Before flashing the rom, REMOVE your SIM card. The important thing is, do not let the phone go online after the fist boot. I will download the Agps data immediately and it may work without problem.
3- Flash X-cm12_1_1-20161131-ufoFILES-p880 or any other Rom which is 4.4+ or later. I've tried many and all of them have this problem.
4- After the first boot,(ignore the no "sim card inserted restart" notification just touch some where out of that box) install satstat and cpu-z.
5- Turn on your GPS if it is off.
6- Launch the GPS test app satstat. It will try to connect to GPS and Glonass satellites (if you have a v20 baseband) It is not important to keep the app running until you get a lock just wait for 15 seconds.
7- Close the gps app.
8- At this point you should be seeing a serious lag. Pushing the power button won't response. The phone won't go the sleep. If it does, it won't wake up easily. You will be pushing the power button several times to make it work. Battery will be drained fast.
7- Launch the CPUZ and watch the CPU load and frequency. One or two cores will spike up and down constantly and the cpu will stay at 1000-1500mhz forever.
8- There is nothing to do but restarting the phone. Sometimes my phone gets totally frozen before I'm able to restart it from the power menu. After the restart it will work normally. Provided if you keep the gps closed and not requesting info from it.
Could you test this please?

Problem solved. Temporary solution.
My description from CM11 Official thread:
Finally, I'm happy to clarify that both assisted and device only gps is working excellent!
After many sleepless nights, I've found what's causing the problem in my device. I've gone over 15+ custom roms and different kernels. I've changed baseband and custom recovery to only get despair.
The last thing I did before abandoning everything.... has worked.
Assisted GPS has to go on-line at least one time.
I'm not sure about the logic behind this but I can't believe how such a simple thing caused this trouble for me...
I don't have a data plan right now so no there is no data from cellular. I generally use device only gps outside. I've never used this rom with assisted gps when outside. I didn't get a chance.
When at home, I flashed roms with "device-only gps" testing in mind so I didn't bother connecting to wireless.
I tried this just with 20150524 and nothing else. (gapps, kernel etc.)
Enabled high accuracy mode on gps and connected to wifi.
Launched GPS Status & Toolbox app and at first, cpu usage was high but after the first fix, it's reduced to normal. I exited the app and cpu usage fell to idle as it should.
Switched off wireless and switched gps to device-only. Rebooted. Launched the GPS Status & Toolbox again and now it's working normally under device only mode. Other applications also working good.
Click to expand...
Click to collapse
Fused Location Provider, responsible for managing different data sources for providing geolocation data, manages the assisted or device-only gps itself. When you enable GPS, it switches the gps mode to high accuracy itself but if you hadn't gone online before with using assisted gps, gps is always stuck at device-only. I think that creates the high CPU usage.

THANK YOU THANK YOU THANK YOU!!!! I wanted to change custom ROM because of this, and you saved me from doing it! So, if I understood correctly, all I need is to enable High accuracy mode, launch "GPS status & toolbox" app (or can it be any other gps app?) and restart after the app gets a fix?

terirem said:
THANK YOU THANK YOU THANK YOU!!!! I wanted to change custom ROM because of this, and you saved me from doing it! So, if I understood correctly, all I need is to enable High accuracy mode, launch "GPS status & toolbox" app (or can it be any other gps app?) and restart after the app gets a fix?
Click to expand...
Click to collapse
You need to let the device to download A-GPS data from internet. GPS status & toolbox does this easily and you can see the "a-gps age" at the app screen as how old your data is.
Once you get a fresh a-gps data, the phone won't stuck at high cpu frequency when gps is enabled.
I 've found out that after "a-gps age" goes over 20+ days, something happens to the gps, and gps usage starts to cause high cpu usage and frequency stuck again. You have to re-download a-gps data to solve this.
I don't know why but gps in this device is not a true offline gps.

I've already described a temporary solution to my GPS bug.
I'm still trying to find a lasting solution.
Here is what I did in various posts:
http://forum.xda-developers.com/showthread.php?p=69893311#post69893311
http://forum.xda-developers.com/showthread.php?p=69970358#post69970358
http://forum.xda-developers.com/showthread.php?p=69988910#post69988910
http://forum.xda-developers.com/showthread.php?p=70066931#post70066931
http://forum.xda-developers.com/showthread.php?p=70067876#post70067876
http://forum.xda-developers.com/showthread.php?p=70068250#post70068250
http://forum.xda-developers.com/showthread.php?p=70071380#post70071380
http://forum.xda-developers.com/showthread.php?p=70084342#post70084342
I also tried
- replacing fusedlocation.apk from /priv-app folder with CM10.1.3 stable's fusedlocation.apk
- another fusedlocation.apk from Atrix 4g (wiped dalvik cache after priv-app change)
- μg v1.3.2 from >HERE<
New roms:
Near all versions after cm-11-20150524-NIGHTLY-p880 up to the FINAL cm-11-20160815-NIGHTLY-p880
omni-4.4.4-20150520-p880-FINAL
X-cm12_1_1-20161016-ufoFILES-p880
X-cm12_1_1-20161131-ufoFILES-p880
Kernels:
CyodakX3_v11_CM
cyospak12
Cyospak6-7
I also upgraded my V20D baseband to V20c-EUR-XXX-OCT-07-2014 and trying ROM/Kernel combinations with this.
So far I've only tried
cm-11-20160815-NIGHTLY-p880
omni-4.4.4-20150520-p880-FINAL
X-cm12_1_1-20161016-ufoFILES-p880
X-cm12_1_1-20161131-ufoFILES-p880
Nothing works for me so far.
Thanks to @oF2pks , I think this may be ril/baseband problem.
Here is another thing:
When I check the "Settings/AboutPhone baseband" it shows L6260_modem_sic_02.1233.00 but If I disable SIM PIN lock and restart the phone it shows LGP880AT-00-V20C-EUR-XXX-OCT-07-2014.
According to oF2pks, a healthy phone should always have only one baseband reading under Settings/AboutPhone baseband.

Sorry for the late answer. I didn't give up and still working on my issue.
For easy describing, lets say State X to the the high cpu state which is causing general slowdown, stuttering, lag and freezes for my device, which occurs after GPS usage, without any data connection after fresh rom install and that kicks in again any time if you are not online with the phone frequently.
After many flashes and tests I found some interesting things.
- In State X, the phone actually goes to deep sleep BUT you have to wait for the 1 minute screen timer and don't push the power button to send it to sleep. After waking it up, cpu is still stuck.
- p880-ota-eng.nmf26qz never displays L6260_modem_sic_02.1233.00 but still falling to the State X.
Only custom rom working without State X problem for me is the cm-10.1.3-p880 stable.
I tried:
Switching cm-11-20160815-NIGHTLY-p880's com.android.location.provider.jar with cm-10.1.3-p880 stable's. → Result: State X
Switching cm-11-20160815-NIGHTLY-p880's telephony-common.jar with cm-10.1.3-p880 stable's. → Result: Error at start-up
add line to getprop file: ro.ril.def.agps.mode = 0
→ Result: State X
Switching cm-11-20160815-NIGHTLY-p880's lge-ril.so with cm-10.1.3-p880 stable's. → Result: State X
oF2pks said:
So do you have right baseband v20C when you boot with sim-on, or still firmware number?
Click to expand...
Click to collapse
This is where the things are getting interesting...
- With CM10.1.3-stable, first boot after the rom flash(without sim card) baseband shows L6260_modem_sic_02.1233.00 but in the second boot with sim card, after the very first initialization with gsm tower, it shows LGP880AT-00-V20c-EUR-XXX-OCT-07-2014 and stays like this.
- With my current daily drive cm-11-20160424-NIGHTLY-p880 + WerewolfJB kernel V10
Without sim-> LGP880AT-00-V20c-EUR-XXX-OCT-07-2014
With sim-> LGP880AT-00-V20c-EUR-XXX-OCT-07-2014
Sim with sim pin lock-> L6260_modem_sic_02.1233.00
With different roms, sometimes baseband switches from LGP880AT-00-V20c-EUR-XXX-OCT-07-2014 to L6260_modem_sic_02.1233.00 after gsm initialization.
In one situation, after all the tampering, I flashed CM10.1.3-stable again and this time it was stuck with
L6260_modem_sic_02.1233.00 and never showed the LGP880AT-00-V20c-EUR-XXX-OCT-07-2014 but no State X
With my daily drive, I also checked both states with sibor's Getril. md5sums are the same.
I'm adding screenshots for them.
I think, maybe baseband display is irrelevant with my issue.
Also did format via recovery cleared your previous rom specific settings?
Click to expand...
Click to collapse
All settings clear but I'm not sure abut default gps switch behaviour. Maybe it's irrelevant too.
Somehow gps switch remembers it's old positions.
Today on single(/primary_multirom) rom I have same kdz designation v20D while booting airplane or sim-on.
But on 2nd/multirom I have the difference (due to primary fake transition via kexec ?), AND the gps is working fine : airplane boot (+FasterGps) + Satstat OK and deep sleep after OK.
Click to expand...
Click to collapse
My current daily drive does not have the State X rgiht now. It will stay like this for a while without internet connection, but after some time, something will trigger the State X again and there are two solutions:
1- Don't enable GPS
2- Connect to internet and update the A-GPS data.
Did you check CPU usage after GPS use?
Mine stays at 1000-1400mhz with one or two cores.
IMHO you should pm @Demetris about V20C because he was the one who did cover v20c-2014 :http://forum.xda-developers.com/show...postcount=2727 ; also he's OP for alternative http://forum.xda-developers.com/opti...grade-t2919195 (and many more...)
Click to expand...
Click to collapse
I did the CSE flash method of Demetris for the V20C 2014.
I'll ask him about this gps issue. Thanks for the suggestion.
I'm also adding my getprop result for you. From fresh CM10.1.3 stable.
The biggest deal breaker for me with this phone was the "temp_throttle_skin" (around 40°C) temp limit. It's a real pain to have and the main reason of general slowdown.
At least I can deal with this GPS issue from time to time when it needs correction.
My aim here is to create some consciousness about this elusive issue for the fellow O4X users, so some unexplained battery drain issues may be resolved or dealt with by the user.
I wish everyone a Merry Xmas and a happy new year.

@KSH2014 , can you check few more:
-is imei/software = printed imei under battery
-is GPS OK with cm10's original kernel
-is GPS OK with lg JB v20c
Within hidden menu in lg JB, I remember bunch of options (GPS ?), and you problem could be somewhere in it, maybe @RuedasLocas could help you to re-settle this hidden menu
http://www.lg-phones.org/lg-optimus-4x-hd-hidden-menu.html

oF2pks said:
@KSH2014 , can you check few more:
-is imei/software = printed imei under battery
-is GPS OK with cm10's original kernel
-is GPS OK with lg JB v20c
Within hidden menu in lg JB, I remember bunch of options (GPS ?), and you problem could be somewhere in it, maybe @RuedasLocas could help you to re-settle this hidden menu
http://www.lg-phones.org/lg-optimus-4x-hd-hidden-menu.html
Click to expand...
Click to collapse
I've been busy lately.
- imei matches.
- GPS is OK with CM 10.1.3 Stable release with it's own kernel.
- GPS was OK or I think so with JB V20. I have to check it again.
I'm planning to record and upload a video to show the problem.

[Fix] Lollipop Bluetooth Speaker stuttering Xperia Z, Z3 [Kernel Devs please help :)]

Since upgrading to Lollipop, both Xperia Z and Xperia Z3 stutter on audio playback over bluetooth. It works fine for a minute or so and after it starts stuttering, UI becomes unresponsive and FC, and many times I had to reboot in order to properly use the phone.
I need someone to point me in the right direction with this kernel issue. SysCtl doesn't make too much sense to me.
[WORKAROUND 1 in the second post]
[UPDATE] After more digging I found out that the Bluetooth Stuttering issue is Kernel related and it affects other brands like LG, Motorola. Governor on BT headset/receiver, while streaming jumps from normal 384 Mhz minimum FQ to 702 Mhz locking on that Frequency and that's why I didn't notice any stuttering on the BT receiver. As soon as I changed the min Fq to 384 (as it should be) the choppiness began as soon as I turned off the screen.
>>So, Bluetooth Receiver on BT version 3.0+EDR - on normal conditions (Governor Interactive - min 384 max 1512) stutters on screen off;
>>Bluetooth Speaker on BT version 4.0 - on normal conditions stutters even more aggressive (1 second of music every 3 seconds of nothing) and crashes the UI and renders the phone barely usable until reboot.
>>Another Issue is with an app (music or bluetooth share) that on the first run changes the governor, and with the kernel policy on sleep mode.
??? A fix might be in Kernel policy or sysctl???
[UPDATE 2] I think I'm looking at the issue completely wrong. I don't think the CPU needs to be forced to keep cores active in sleep states but I think the bluetooth driver (must be in devices/platform) must be kept hogging resources in screen off. Anyone with a brilliant idea? I just discovered where the cpu settings are so not much I can do.
[UPDATE 3] [Workaround 2] Disabling MPDecision will keep all 4 cores running with no excessive battery drain - I had no stutter in 2 hours of playback and also The Min FQ didn't change. So now, when I go out, I either turn on the Partial Wakelock or temporary disable the MP Decision. I've got some help from kernel devs and I'll try playing with the hotplugs sometime next week and maybe come with a real fix. If anyone with enough knowledge of kernel wishes to participate I'd be really grateful.
I have very limited knowledge on Kernel and I really need help - I'm now learning stuff. But, I'll compare KitKat SysCtl to Lollipop SysCtl to see if I find something that might help understand the issue and maybe come up with a fix. On KK I never had this kind of issues. If anyone can help me on this please PM.

Maybe this help
Hello
I can remember that is had a lot of stuttering on bluetooth with a speaker or with music regularly what is did, was let the phone run a default 468 Mhz instead of 384 Mhz, after that the problems were over.
I dont know of this will fix it but you can try it

Sovjet-Command said:
Hello
I can remember that is had a lot of stuttering on bluetooth with a speaker or with music regularly what is did, was let the phone run a default 468 Mhz instead of 384 Mhz, after that the problems were over.
I dont know of this will fix it but you can try it
Click to expand...
Click to collapse
I noticed this yesterday while I was playing with the governor. On xperia Z governor settings were jumping to min FQ of 702Mhz which has a huge impact on battery. Even on trickster settings wont stick for long. So far the issue isn't related to SND_Soc audio mapping, it's the crappy kernel implementation on multiple devices running 5.02 (LG, Motorola etc).

Workaround 1
The cause is the kernel after all or at least as far as I know.
The workaround is about creating a Partial Wakelock so the cpu stays active after screen off with this app.
It works. No more stuttering. As for speakers that require high sample rate I just switched the cpu governor to performance (for the Z3 mostly). The partial wakelock should work especially on bluetooth headphones. I also found an app that ads partial wakelocks here's the link : https://play.google.com/store/apps/d...arken.wl&hl=en

SysCtl diff between KK; LL
These are Kernel differences discovered in the SysCtl file - comparison made betwen Kitkat (KK) and Lollipop (LL) are these of any help?
Fs.epoll.max_user_watches KK 169642 LL 164219
Fs.file-max KK 183359 LL 181926
Kernel.dmesg_restrict KK 1 LL 0
Kernel.msgmni KK 877 LL 851
Kernel.threads-max KK 28653 LL 28429

Bump
Sent from my SPH-L900 using Tapatalk

Bump any fix made?

superspeedz32 said:
Bump any fix made?
Click to expand...
Click to collapse
Well, I worked around the issue so further investigation wasn't really necessary since it would take a lot of time.
Here is what I know so far:
Bluetooth streaming works flawlessly with aptX devices
It's kernel related - and mp decision
This is my daily driver for music http://forum.xda-developers.com/gen...h-receiver-t3212201/post63021706#post63021706 which now works flawlessly.
MP-Decision, while music streaming, upscales the min cpu fq to 702mb which means the CPU will drain a lot of battery. So, with Trickster or the cpu app of your choosing, lock the min fq to 384.
To avoid the stuttering (if you're using interactive governor), set the hispeed_freq @ 1134mb and go_hispeed_load of anywhere between 85-90 % default values are too power hogging and what i'm usually trying is to find a ballance between decent performance and battery saving.
In my experience interactive has the tendency to overheat the device much faster, which is why I'm now using the ondemand governor, but since now i'm using an aptx receiver I hardly have any issues with stuttering.
Anyway, feel free to PM me for more infos.

App name
Sebi673 said:
The cause is the kernel after all or at least as far as I know.
The workaround is about creating a Partial Wakelock so the cpu stays active after screen off with this app.
It works. No more stuttering. As for speakers that require high sample rate I just switched the cpu governor to performance (for the Z3 mostly). The partial wakelock should work especially on bluetooth headphones. I also found an app that ads partial wakelocks here's the link :
Click to expand...
Click to collapse
What was the name of the app ? It´s not on the store anymore. Thx in advice

Tenshinryu said:
What was the name of the app ? It´s not on the store anymore. Thx in advice
Click to expand...
Click to collapse
Wake lock power manager. Although I don't have that problem since the last fw update. Package Name eu.thedarken.wl
It helped me a lot so feel free to try it

Sebi673 said:
Wake lock power manager. Although I don't have that problem since the last fw update. Package Name eu.thedarken.wl
It helped me a lot so feel free to try it
Click to expand...
Click to collapse
Thanks I have installed it, but maybe one of you could help me with this idea: I would like to have Tasker execute the Partial_Wake_Lock via the Wake Lock Power Manager when Bluetooth is connected
I have come so far that Tasker starts the Wake Lock Power Manager, but I don´t know how to execute certain routines within the started app. Any ideas ?
Maybe someone at least knows how to start functions within a app with Tasker ?
Thanks in advance
UPDATE: I´ve managed it by myself
For those of you interested how - here´s what you need: Root, Tasker, Execute Locale App :
In Tasker
> Set Profile Bluetooth Connected
Task :
1.Run Application Wake Lock
2.Input > Trackball > Down, 5
3. Plugin > Execute Locale > !Input keyevent 66
That´s it.

Tenshinryu said:
Thanks I have installed it, but maybe one of you could help me with this idea: I would like to have Tasker execute the Partial_Wake_Lock via the Wake Lock Power Manager when Bluetooth is connected
I have come so far that Tasker starts the Wake Lock Power Manager, but I don´t know how to execute certain routines within the started app. Any ideas ?
Maybe someone at least knows how to start functions within a app with Tasker ?
Thanks in advance
UPDATE: I´ve managed it by myself
For those of you interested how - here´s what you need: Root, Tasker, Execute Locale App :
In Tasker
> Set Profile Bluetooth Connected
Task :
1.Run Application Wake Lock
2.Input > Trackball > Down, 5
3. Plugin > Execute Locale > !Input keyevent 66
That´s it.
Click to expand...
Click to collapse
ha awesome dude. you basically manually started the activity. but automatically ))))

Sebi673 said:
ha awesome dude. you basically manually started the activity. but automatically ))))
Click to expand...
Click to collapse
Yeah, I dunno, but I didn´t like the thought of doing that manually all the time I´m using Bluetooth frequently
But, maybe I found another solution to my BT stutter / disconnect problem: I installed Proximity Fix and couldn´t test in my car or my speakers, but my Gear Fit didn´t disconnect until now, like it did before
Update: No, no, no didn´t fix my problem. Wether the proximity fix nor the Wake Lock solution. I also did a factory reset.. I´m really eager to reroll to 5.1.1.