Database Develop

Post Your Smartbench Scores!

With so many ROMS to choose from, I thought that it would be nice to have people post their Smartbench scores along with pertinent information such as ROM used, VM Heap Size, etc.
I like Smartbench much more than Quadrant since it breaks up the result in 2 categories and are more reliable - since Quadrant scores are all over the map depending on the direction of the wind in Tokyo, Japan
I think that with alot of input, we can more easily determine which ROMS are truly faster or maybe help people figure out if there is a misconfiguration on their phone.
I figure scores in between 800-900 (Productivity) and 1100-1200 (Games) range would be pretty good - since I had got below 700 and 1100 before tweaking.
Anyway, here is my best score so far:
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
ROM..................................: CM 612 0.07-11.01.28
VM Heap Size......................: 48 mb
[email protected]: 1100mhz @ 64 vsel
Memory Management (in AOR): Mild - Keep Launcher Alive
Governor............................: Performance
Battery Level......................: 100%

Here's mine at 900Mhz-56vsel using Moto APAC / Interactive Governer / VM size 28mb

I forgot to mention that your honeycomb theme goes fast for me

Wouldn't be more relevant if we all use the same settings (CPU speed , vm , governor) , just to see the performance differences between the ROMs?

Sure, comparing apples to apples makes sense.
Most people seem to be in the 900 to 1000 mhz range so they can still be comparable.
Say a phone at 900mhz scores higher or the same as a phone at 1000mhz, then it could be that the phone at 900mhz is running a much faster rom.
It would be nice if people would start posting scores from different roms indeed.
I'll be trying a different rom soon so I would post the score with the same settings that I used in the first post.

Rom: shadowmod b3
CPU: 1000mhz @ 60vsel
governor: smartass
BTW, what are meant by productivity index and game index respectively ???

Wow nice games index score...
I would imagine that Productivity Index is the cpu efficiency/performance and the Games index would the related to the Graphics performance ie. higher frames per second = smoother game play.
Maybe you can try increasing your VM heap size to 40 or 48 mb and reboot to see if it increases your Productivity score.

zeppelinrox said:
Wow nice games index score...
I would imagine that Productivity Index is the cpu efficiency/performance and the Games index would the related to the Graphics performance ie. higher frames per second = smoother game play.
Maybe you can try increasing your VM heap size to 40 or 48 mb and reboot to see if it increases your Productivity score.
Click to expand...
Click to collapse
How to increase VM heap size and what is it???
any tutorials ????

Its in cm6 settings but also aopenrecovery in the froyo menu
Sent from my Milestone using Tapatalk

zeppelinrox said:
Its in cm6 settings but also aopenrecovery in the froyo menu
Sent from my Milestone using Tapatalk
Click to expand...
Click to collapse
aopenrecovery= androidiani OR??
Does it mean that the greater the size, the better is the performance??
There are two 40mb and two 48mb in the froyo menu,
What is the difference??
Sent from my A853 using XDA App

when smartbench starts analysis, it gives me force close
what's the problem ?

hearts king101 said:
when smartbench starts analysis, it gives me force close
what's the problem ?
Click to expand...
Click to collapse
Do you install it in SD card (app2sd) ??
if yes, plz move it back to internal memory
Sent from my A853 using XDA App

James91 said:
aopenrecovery= androidiani OR??
Does it mean that the greater the size, the better is the performance??
There are two 40mb and two 48mb in the froyo menu,
What is the difference??
Sent from my A853 using XDA App
Click to expand...
Click to collapse
I think the bottom 2 items don't work right.
Use the first 40mb or first 48mb options
Sent from my Milestone using Tapatalk

zeppelinrox said:
I think the bottom 2 items don't work right.
Use the first 40mb or first 48mb options
Sent from my Milestone using Tapatalk
Click to expand...
Click to collapse
I have googled the term and some said it is the Max memory that is allowed for a single app to use.
And most set their vm head size at 16 MB.
Is there any bad effect if I set it at a greater value??
Sent from my A853 using XDA App

I thought that the default value in CM6 was 24mb.
I saw your post via tapatalk so I actually tested it out at 64mb and the phone was incredibly quick and smooth.
I can't say that it affected benchmark scores much but I was surprised at how quickly Smartbench loaded up!
If you look at the OP, in the status bar I have an icon with 2 vertical yellow bars.
That's SeePU.
The first bar is CPU load/activity and the second bar is free memory.
When I was doing the test with that large VM heap size, that second bar was near the top and green - meaning lots of free ram. I guess the memory got flushed out. No need for a task killer lol.
Then I closed smartbench and there was no desktop lag and process monitor was showing over 60mb of free ram despite the fact that my memory management setting in AOR is set to mild.
I did not see any ill effects at all but keep in mind that my dalvik cache is on an ext2 partition so that may be a factor as well.
This thread that I just found pretty interesting. http://forum.cyanogenmod.com/topic/2692-vm-heap-size/
Also, as far as I can figure, an app will only use all the VM Heap memory if it needs it.
If it needs 40 mb and the vm heap is only 16 mb, that app's performance will be crap when compared to how it would perform if the VM heap was 40 or 48mb.
I guess the only downside would be that there is the potential for background apps (media player, etc...) to close because the foreground app is using up too much memory. If that's the case then I'd lower the VM heap to a more balanced level.
The lowest I'm going is my usual 48mb but I'm gonna stick with 64mb for awhile...
Edit: Hey it actually helped with linpack... I managed 15 MFLOPS

zeppelinrox said:
I thought that the default value in CM6 was 24mb.
I saw your post via tapatalk so I actually tested it out at 64mb and the phone was incredibly quick and smooth.
I can't say that it affected benchmark scores much but I was surprised at how quickly Smartbench loaded up!
If you look at the OP, in the status bar I have an icon with 2 vertical yellow bars.
That's SeePU.
The first bar is CPU load/activity and the second bar is free memory.
When I was doing the test with that large VM heap size, that second bar was near the top and green - meaning lots of free ram. I guess the memory got flushed out. No need for a task killer lol.
Then I closed smartbench and there was no desktop lag and process monitor was showing over 60mb of free ram despite the fact that my memory management setting in AOR is set to mild.
I did not see any ill effects at all but keep in mind that my dalvik cache is on an ext2 partition so that may be a factor as well.
This thread that I just found pretty interesting. http://forum.cyanogenmod.com/topic/2692-vm-heap-size/
Also, as far as I can figure, an app will only use all the VM Heap memory if it needs it.
If it needs 40 mb and the vm heap is only 16 mb, that app's performance will be crap when compared to how it would perform if the VM heap was 40 or 48mb.
I guess the only downside would be that there is the potential for background apps (media player, etc...) to close because the foreground app is using up too much memory. If that's the case then I'd lower the VM heap to a more balanced level.
The lowest I'm going is my usual 48mb but I'm gonna stick with 64mb for awhile...
Edit: Hey it actually helped with linpack... I managed 15 MFLOPS
Click to expand...
Click to collapse
Bro, I have tried to increase vm heap size to 48 mb but the score is close to the old one.
I also notice that in a post, the author did a test and he found that 12mb would result in the best performance.
Sent from my A853 using XDA App

Yeah I said near the top that it didn't effect the benchmark scores much - meaning smartbench.
But it did give me consistent high scores in linpack and I even got over 15 mflops which has never happened to me before.
I saw the post that you are referring to and they actually talk about it in that cyanogenmod forum thread that I linked. A couple of people explain VM heap very well and pretty much explain why that 12mb test didn't mean anything.
The main thing is that my phone is actually running alot smoother with a larger VM heap though

I have raised the size to 48mb
and it seems to be better
What is the Max vm head size that I can set??
Sent from my A853 using XDA App

I had it at 64 mb for a couple of days but I think that it is just too big in that when I checked running processes, some apps or services that I want to be running would not be running... so I put it back to 48mb

James91 said:
I have raised the size to 48mb
and it seems to be better
What is the Max vm head size that I can set??
Sent from my A853 using XDA App
Click to expand...
Click to collapse
hey guys
i have no idea about what U R doing
how did u raise the size to 48mb ?
what is the benefits of these changes u r making ?
plz would u explain what does vm head size mean?

Kernel and ROM Benchmarks

Users can feel free to post their benchmarks here
Here's some of my benchmarks:
- Tests were ran from [ROM][STOCK IMAGE]Android 4.1.1 (JRO03D) ROOT, Deodexed, Busybox
- Testing done with latest Quardrant Standard
- Battery, Dalvik, and Cache were all wiped after installing kernel from zip in CWM
- Kernels tested were all the latest and highest-OC'd versions available as of 8/7
- No governers, schedulers, or clock speeds were altered from default values (I plan to reupload screenshots of performance later)
[KERNEL][GPL][Linaro][OC 1.624GHz][UV][GPU+][ZRAM][SIO+V(R)] 2012-08-03 motley 1.0.12 (GPU 520, build 175)
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
Kernel[006]AP33(1.55Ghz Quad)UV,CIFS+UTF-8,PowerHAL,Linux-3.x Hybrid [Aug-04] (Ultimate Edition)
[DER KERNEL] Trinity Seven (TS-1640-ALPHA41)
Stock

Note to self get an HTC One X, man thats a fast phone

Quadrant is very inconsistent with benchmarks. I would suggest Antutu/CFBench/Nenamark/GLBenchmark/RL Benchmark as being more accurate.
Sent from my Nexus 7

If you use TKT with trinity it blows the One C X away
you can do it via the terminal but idk the code and I already had TKT
Sent from my Nexus 7 using xda app-developers app

I've never seen a trinity kernel score that low. Most people score more than double overall of what you posted. Plus your I/O scores are warped. People scoring in the 7000-11,000 range. You not benching your device properly or something or using the wrong settings. Those are the worst scores I've seen posted on a nexus 7..lol.
Main tip for benching: put device in performance mode or at least ondemand. Even on stock governor, I've seen majority score higher than you. Just take a look at majority of posted benchmarks and you will see you don't have something right set up on your device.
You making those kernels look bad. Check mines out and it wasn't even on the highest setting. This is with trinity kernel OC to 1.5ghz. Stock rom and stock gpu speed.
I've never seen Motley, Infinity kernel, faux, or even stock score as low as what you have..lol.
the last quadrant I posted was on motley kernel.
huge difference in my scores and yours. plus mines on the lower end because I only benched at 1.5ghz. if I would've went top speed, scores would be even higher. we have people now benching close to and over 7000-7100 overall on quadrant. it blows any other android device out now. so much so there's been several tech news stories on how nexus 7 is blowing away other higher end android devices in benchmarks.
p.s. my nenamark2 score is 64.7fps
also I know quadrant can be inconsistent but those scores you posted still suspiciously low for those kernels they supposedly are from.

Just did 3 more ones. Stock rom. Trinity kernel CPU speed 1.5ghz. Performance governor and deadline scheduler
I find chainfire bench to be one of the best. doesn't rely on fancy graphics testing and such. this tests more raw power. as you can see, I beat out the new Samsung galaxy S3 with its 32nm high speed chip. remember also, we still in infancy stage of nexus7 development and we blowing away all benches & devices(Android) already. pretty good for a $199-249 tablet huh? lol. this performance level also translates to real world use. not just for show.

how are you getting such high scores? mine are always around 4600.. running at 1600 with trinity. doesnt make sense that yours are so high? highest ive seen by a good ways.

You’re very professional.

demandarin said:
I've never seen a trinity kernel score that low. Most people score more than double overall of what you posted. Plus your I/O scores are warped. People scoring in the 7000-11,000 range. You not benching your device properly or something or using the wrong settings. Those are the worst scores I've seen posted on a nexus 7..lol.
Main tip for benching: put device in performance mode or at least ondemand. Even on stock governor, I've seen majority score higher than you. Just take a look at majority of posted benchmarks and you will see you don't have something right set up on your device.
Click to expand...
Click to collapse
When I benched, I just had everything set at default settings (didn't change governers, clocks, or anything).

espionage724 said:
When I benched, I just had everything set at default settings (didn't change governers, clocks, or anything).
Click to expand...
Click to collapse
when you benchmark, use highest cpu/highest cpu. if you dont set it there, you let your device decide what cpu speed it actually wants to use, therefore you have no idea what cpu speed you are benching. for all you know is that on one bench it might be benching 1100mhz, on another it might be benching less or more. btw, those benchmarks are horrible, my galaxy nexus benchmarks higher

simms22 said:
when you benchmark, use highest cpu/highest cpu. if you dont set it there, you let your device decide what cpu speed it actually wants to use, therefore you have no idea what cpu speed you are benching. for all you know is that on one bench it might be benching 1100mhz, on another it might be benching less or more. btw, those benchmarks are horrible, my galaxy necus benmarks higher
Click to expand...
Click to collapse
Would setting ROM Toolbox to use the highest CPU frequency + performance be acceptable? Or should I use another program to ensure the CPU is at highest?
Edit: After doing that:
Total, CPU, Mem, I/O, 2D, 3D
Trinity: 4142, 12791, 3737, 1387, 332, 2464
Faux: 4457, 14257, 3890, 1339, 333, 2465
Motley: 5846, 15184, 3618, 7191, 300, 2935
Tests done from Glazed JellyBean 1.12
Those scores still seem worse then what some others have said though. Is there anything else I should be doing?

Sadly I have to flash back to stock kernel to make the wifi functionally. But sure the motley kernel is pretty attractive on its performance and addons.

espionage724 said:
Would setting ROM Toolbox to use the highest CPU frequency + performance be acceptable? Or should I use another program to ensure the CPU is at highest?
Edit: After doing that:
Total, CPU, Mem, I/O, 2D, 3D
Trinity: 4142, 12791, 3737, 1387, 332, 2464
Faux: 4457, 14257, 3890, 1339, 333, 2465
Motley: 5846, 15184, 3618, 7191, 300, 2935
Tests done from Glazed JellyBean 1.12
Those scores still seem worse then what some others have said though. Is there anything else I should be doing?
Click to expand...
Click to collapse
I find system tuner to be the best for those type of settings. If you run trinity kernel though, you will get more out of it by using the trinity kernel app in playstore that lets you tweak things in their kernel that even apps like system tuner can't do. But overall system tuner is the best, IMO. Its more in depth than most apps of THST nature.
Also best to use performance governor. Or on demand. Then use deadline or sio as the scheduler.
x-magic said:
Sadly I have to flash back stock kernel to make the wifi functionally. But sure the motley kernel is pretty attractive on its performance and addons.
Click to expand...
Click to collapse
Nothing wrong with wifi on motley kernels. Which build were you running? His latest runs great and no issues. Trinity kernel wifi works perfect also.

This is my score using latest trinity build 41 and pure aosp ROM.
Sent from my Nexus 7 using xda premium

saadi703 said:
This is my score using latest trinity build 41 and pure aosp ROM.
Sent from my Nexus 7 using xda premium
Click to expand...
Click to collapse
as with others i max out at 46XX on quad if i am lucky and thats using system tuner and putting min/max sliders to 1.6 and performance gov. are you guys doing any other kernel tweaks to get these numbers

xspeed9190 said:
as with others i max out at 46XX on quad if i am lucky and thats using system tuner and putting min/max sliders to 1.6 and performance gov. are you guys doing any other kernel tweaks to get these numbers
Click to expand...
Click to collapse
I used trinity kernal build 41 with pure aosp rom. I used trinity kernel toolbox app from play store to tweak it. It is a paid app and if you don't want to buy the app then you have to use scripts but to keep things simple I used the app.
I clocked the processor to 1640 MHz, put the governor to performance which means processor would run at 1640 always. I/O scheduler to deadline and in tunable I select Fsync to faster. Good luck
Oh I forget to mention that good thing about Trinity kernel toolbox app is that you can use it with any kernel not only with trinity. The CPU temp widget is also good
Sent from my Nexus 7 using xda premium

Quadrant scores are meaningless. There are so many ways that they can be artificially inflated. I suspect that the scores in the 4000 range are the ones that truly reflect the overall performance of the device. 6000+ scores are obviously not realistic scores. Look at the IO scores, 10000 is a bogus 10x improvement over stock. No setting change anywhere is going to improve your IO performance 10x.

hecksagon said:
Quadrant scores are meaningless. There are so many ways that they can be artificially inflated. I suspect that the scores in the 4000 range are the ones that truly reflect the overall performance of the device. 6000+ scores are obviously not realistic scores. Look at the IO scores, 10000 is a bogus 10x improvement over stock. No setting change anywhere is going to improve your IO performance 10x.
Click to expand...
Click to collapse
someones feeling small..

TS-ALPHA50 + Glazed Jellybean v1.12

espionage724 said:
TS-ALPHA50 + Glazed Jellybean v1.12
Click to expand...
Click to collapse
better :victory:

[UV] Nexus 4 UV Thread - POST YOUR UNDERVOLTING RESULTS - GET A (BATTERY)LIFE :)

Hey Guys,
I read alot about OCing your Phone and Overvolting, but how about Undervolting? I thought I’ll open a new thread for discussing the UV potential of our Nexus and to show other users that they can achieve better batterylife by UVing. If you want to, please provide which chip you've got (slow, nominal, fast, faster) which ROM/kernel you are on and which voltage you’ve set on each state (use Trickster or similar apps. I recommend System Tuner Pro. You can save your stable UV settings and easily apply them again after a ROM/Kernel update).
Use this code to figure out which chip you've got (the result will either be slow - faster or numbers from 1 - 4):
1.) Install Terminal Emulator
2.) Type in su and press enter
2.) Reboot your phone
3.) Open Terminal Emulator and type:
su (press enter)
dmesg | grep PVS (press enter again)
Click to expand...
Click to collapse
I have a "FAST" chip and I’m on PA 3+ with latest _m+ kernel (motley kernel continued).
My voltages so far (StabilityTest stable):
192 MHz - 625,0 mV
304 MHz - 650,0 mV
486 MHz - 675,0 mV
594 MHz - 700,0 mV
702 MHz - 737,5 mV
810 MHz - 775,0 mV
918 MHz - 800,0 mV
1,02 GHz - 825,0 mV
1,13 GHz - 850,5 mV
1,24 GHz - 875,0 mV
1,29 GHz - 900,0 mV
1,35 GHz - 910,0 mV
1,40 GHz - 925,0 mV
1,45 GHz - 940,0 mV
1,51 GHz - 950,0 mV
Attention! Some users reported glitches/artifacts when going under a certain Voltage-level. So don't just test your settings with StabilityTest, take some pictures and or videos as well
Here is a post I consider to be very useful. Please read it:
kelah said:
I think google/LG messed up with voltages in more ways it looks on the first glance.
Put their values on the graph and You'll notice that the line is linear. But it should be somewhat parabolic.
It's easy to demonstrate on the graph.
Blue line - stock google voltages.
Red line - this I get when I set UV -150 (I can't go lower it).
Violet line - my stable setup.
As you can see my stable setup is parabolic. And in the center difference is big - 225mV at 1188mHz and 200mV at 1026mHz. For other values it ranges from 25 to 35.
Green line consist of values crashes device immediatly - I set them and right after I tap on "apply" my lovely N4 rebooting. The point is what between them and stable values 50mV. If difference abot 5-25 device works but it reboot for sure in half of hour or something like that. This is the easy way to determine how much UV you can - just find values of instant crash and rise them about 40-50mV. These values are easy to find - no need to run long tests. Would be better to do nandroid before - many repeatedly sporadic reboots can corrupt some of your data.
My surrent setup:
384 - 700 (-150)
486 - 700 (-175)
594 - 720 (-180)
702 - 740 (-185)
1026 - 825 (-200)
1188 - 875 (-225)
1350 - 940 (-185)
1512 - 1000 (-150)
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
Click to expand...
Click to collapse
Greetz

Re: Nexus 4 UV Thread
I am on the new pa ROM
Faux latest kernel
Min 384
Max 1300
UV at -150
WiFi and data use.
Screen brightness just below half .
Now on 56% battery left and been off charge for
1 day 2 hours. ..... Gimme a hell yehh....
Sent from my pie'd nexus 4

Re: [UV] Nexus 4 UV Thread - POST YOUR UNDERVOLTING RESULTS
CM10.1
Sent from my Nexus 4 using Tapatalk 2

Re: [UV] Nexus 4 UV Thread - POST YOUR UNDERVOLTING RESULTS
Bigxie AOKP PUB build 2/15 (4.2.2)
Frankien intelli-plug kernel
Fast bin
Sent from my Nexus 4 using xda premium

Re: [UV] Nexus 4 UV Thread - POST YOUR UNDERVOLTING RESULTS
revamper said:
CM10.1
Sent from my Nexus 4 using Tapatalk 2
Click to expand...
Click to collapse
+1
Sent from my Nexus 4 using xda premium

How do you test for different frequency stages? Stability test always runs at maximum CPU speed.

Re: [UV] Nexus 4 UV Thread - POST YOUR UNDERVOLTING RESULTS
CM 10.1 with franco
Sent from my Nexus 4 using xda app-developers app

384Mhz min
1024Mhz max
-150mV
Faster binned CPU
Completely stable, could probably go to at least -175mV stable... But it's okay as it is, phone is still snappy(can't even tell the difference tbh) and the battery will last me 1,5 - 2 days, which was something unseen on the iPhone 4 I had before..

How do you test for different frequency stages? Stability test always runs at maximum CPU speed.
Click to expand...
Click to collapse
You just have to choose "Scaling Stability Test (Root)" in the main menu. There you can choose which frequencies will be tested.

I'm really interested in undervolting but want to find out which chip I have. I tried doing the terminal emulator technique above but the sheer amount of info it pulls seems incomprehensible to me! Could someone point me in the right direction of how I find out what kind of chip I have got in here so I can start playing around with UVing? Just for reference I'm on the latest PA 3.0 ROM with Faux Kernel
Cheers

NightflyUK said:
I'm really interested in undervolting but want to find out which chip I have. I tried doing the terminal emulator technique above but the sheer amount of info it pulls seems incomprehensible to me! Could someone point me in the right direction of how I find out what kind of chip I have got in here so I can start playing around with UVing? Just for reference I'm on the latest PA 3.0 ROM with Faux Kernel
Cheers
Click to expand...
Click to collapse
http://forum.xda-developers.com/showthread.php?t=2052496
You should be able to UV to at least -100mV, no matter what chip you have..

I might be being daft but when I run that command nothing comes up (can't post PIC) I'm definitely rooted as I'm running custom ROM and kernel and using terminal editor. It just comes up
1|[email protected]:/#

Okay i did those steps:
1. Installed Terminal Emulator
2. SU and enter
3. Reboot
4. SU command and command from the OP
5. The last word describes your chip (could be also numbers from 1 - 4 I think)

Herminator19640927 said:
Okay i did those steps:
1. Installed Terminal Emulator
2. SU and enter
3. Reboot
4. SU command and command from the OP
5. The last word describes your chip (could be also numbers from 1 - 4 I think)
Click to expand...
Click to collapse
Cheers man that sorted it its giving me a result as 1 so at which end of the scale is that? Also when I run the acpuclk its telling me it can support 15 frequencies

NightflyUK said:
Cheers man that sorted it its giving me a result as 1 so at which end of the scale is that? Also when I run the acpuclk its telling me it can support 15 frequencies
Click to expand...
Click to collapse
I think 1 means slow or nominal, not quite sure. Read it somewhere... just look around in the dev section (I think it was a post by Harsh).
Greetz

NightflyUK said:
I might be being daft but when I run that command nothing comes up (can't post PIC) I'm definitely rooted as I'm running custom ROM and kernel and using terminal editor. It just comes up
1|[email protected]:/#
Click to expand...
Click to collapse
I had to reboot after installing terminal emulator in order to get the command to work.

Herminator19640927 said:
I think 1 means slow or nominal, not quite sure. Read it somewhere... just look around in the dev section (I think it was a post by Harsh).
Greetz
Click to expand...
Click to collapse
Thanks. I just tried it but it always stay at 1016Mhz. I'm using Franco's latest Kernel. Governor is set to ondemand, but I wont change frequencies.

Re: [UV] Nexus 4 UV Thread - POST YOUR UNDERVOLTING RESULTS
Faster, cm10.1 original, francos kernel r72
Sent from my Nexus 4 using xda app-developers app

digitalfrost said:
Thanks. I just tried it but it always stay at 1016Mhz. I'm using Franco's latest Kernel. Governor is set to ondemand, but I wont change frequencies.
Click to expand...
Click to collapse
Strange... for me everything's working fine. StabilityTest sets my governor to userspace and changes through all frequencies in the specified delay.

Stock 4.2.2, franco r92.
Stability test, long term usage, etc... is stable.

I need some quadrant benchmarks from everyone.

This will help to give me a better idea of what to aim for as in terms of performance, what roms people are using and more. Please run and attach 3 screenies.
Please follow this example when posting.
Rom: AOIP 1.5
Android Version: 4.2.2
Kernel:3.4.63 Whitescreen Fix 4.2
Optional Tweak Scripts and/or Apps: Ramexpander, V6 Supercharger, PerfomanceX

Rom: [ROM][4.1.2] Cyanogenmod 10 unofficial Galaxy Exhibit 4G 04/15/2013
Android Version: 4.1.2
Kernel:2.6.35 OC Kernel
Optional Tweak: CPU: 1.6 Ghz Max, 245 Mhz Min, Interactive Governor, I/O Scheduler: SIO, Greenify, Viper4Android, RAM Manager Pro using "Balance" profile.
Quadrant Benchmark Score:
Total: 2869
CPU: 3470
Mem: 1422
I/O: 7504
2D: 472
3D: 1462

Rom: Vanir 4.3.1.102713
Android Version: 4.3.1
Kernel: 3.4.67 JB_OC_UV_360MB+
Optional Tweaks: Max CPU: 1401 MHz, MIn CPU: 1401, CPU gov: SMARTASSV2, I/O: NOOP.
Quadrant score:
Total: 2371
CPU: 3036
MEM: 1387
I/O: 5662
2D: 298
3D: 1472

Cm10

Jorge_007 said:
Perfectly done. Thank you both so much and anyone else that contributes after this post.
The reason being for this info is because I'm getting ready to release first initial AOIP CE very soon. It won't be that special YET. It's just basic AOIP 1.5 8/10 with aroma installer and few other things that will be mentioned when released. Wanting to know how well it compares up to whats already out there before making additional default settings changes. ION, SWAP feature, and other of my own additions will be released at a later time. With that said I do think everyone will still enjoy first release and where I'm headed with our phones.
Click to expand...
Click to collapse

TripFX said:
Jorge_007 said:
Perfectly done. Thank you both so much and anyone else that contributes after this post.
The reason being for this info is because I'm getting ready to release first initial AOIP CE very soon. It won't be that special YET. It's just basic AOIP 1.5 8/10 with aroma installer and few other things that will be mentioned when released. Wanting to know how well it compares up to whats already out there before making additional default settings changes. ION, SWAP feature, and other of my own additions will be released at a later time. With that said I do think everyone will still enjoy first release and where I'm headed with our phones.
Click to expand...
Click to collapse
Nice, man! Looking forward to it! :victory:
Click to expand...
Click to collapse

OK well here's mine. Score on this test is near 2700 which is what I've been averaging lately. I'm running AOIP 1.5 with Cam's latest OC kernel set to 245 min, 1.4 max. Governor is "on demand" and scheduler is "SIO" (which works best with that Governor). Memory tweaks are RAM Expander with 1gb swap and 100 swappiness (a very high setting but useful for a low performance device like this one), Zram enabled at 26% (useful now I have additional memory, no more freezing when using it) and finally KSM enabled. Seeder running at maximum and V6 mainly to maintain minfrees as Ram Manager was hit or miss. Try for 512hp aggressive setting.
A lot of work but worth it so I can run apps that are resource heavy such as the launcher from the Nexus 5. It has Google Now integrated into it.
It is worth noting that all of our high scores are mainly to do with the I/O score we get, but memory and CPU are not so hot!
Sent from my Samsung Galaxy Exhibit running AOIP.

ArtfulDodger said:
OK well here's mine. Score on this test is near 2700 which is what I've been averaging lately. I'm running AOIP 1.5 with Cam's latest OC kernel set to 245 min, 1.4 max. Governor is "on demand" and scheduler is "SIO" (which works best with that Governor). Memory tweaks are RAM Expander with 1gb swap and 100 swappiness (a very high setting but useful for a low performance device like this one), Zram enabled at 26% (useful now I have additional memory, no more freezing when using it) and finally KSM enabled. Seeder running at maximum and V6 mainly to maintain minfrees as Ram Manager was hit or miss. Try for 512hp aggressive setting.
A lot of work but worth it so I can run apps that are resource heavy such as the Touchwiz launcher from the Nexus 5. It has Google Now integrated into it.
It is worth noting that all of our high scores are mainly to do with the I/O score we get, but memory and CPU are not so hot!
Sent from my Samsung Galaxy Exhibit running AOIP.
Click to expand...
Click to collapse
not in the format i prefer but it works. I/O isn't the only thing that changes but for the most part you are right. However if you look at the screenies you guys posted so far and the one i posted in aoip http://forum.xda-developers.com/attachment.php?attachmentid=2330114&d=1381961624 you can tell there is other things that do change

TripFX said:
not in the format i prefer but it works. I/O isn't the only thing that changes but for the most part you are right. However if you look at the screenies you guys posted so far and the one i posted in aoip http://forum.xda-developers.com/attachment.php?attachmentid=2330114&d=1381961624 you can tell there is other things that do change
Click to expand...
Click to collapse
Sorry I know the others did line by by line but there is so much to explain with these tweaks (plus you can see each separate score in the picture). 2900 is the highest score I have gotten too, and it looks like we all are getting similar scores using a 4.2 ROM. I really didn't think 4.3 (Vanir) was near as friendly to our phone.
BTW interesting to note the launcher seems to be just an extension of Google Now and is not Touchwiz at all. Running applications list and V6 no longer detect a launcher running. Articles I've read seem to confirm that as Google really wants Gnow utilized on as many devices as possible. Memory consumption is the same as Nova anyway.

So I rebooted my phone and gave it a few more bench runs (with same settings) here are the results.
Hope this helps!

Doobski said:
So I rebooted my phone and gave it a few more bench runs (with same settings) here are the results.
Hope this helps!
Click to expand...
Click to collapse
Since I started here one thing we've learned is that benchmarks only give you an idea how your phone is doing but is not the definitive answer on how your device runs. If everything is running smoothly with no lag, freezing or random reboots than things are good is what it boils down to.

[WIP][5.0/5.1][CM12/12.1] Custom Droid 4/RAZR OC Kernel and SDSwap

This is a custom kernel based off of the CM12.0 Kernel that was created by stargo for Lollipop. Below is the list of features this kernel adds to the Stock CM12.x:
Over-clocking to 1.3 and 1.4Ghz
Additional Governors - InteractiveX, KToonservative, Lionheart, SmartassH3 and ZZMove)
Additional Schedulers - ROW, SIO, SIO Plus, VR and Zen
IntelliPlug driver - Hotplugging while sleeping for all governors (leave governor Hotplug disabled as this overrides)
I've included my SDSwap in separate zip as well here for folks who would like to swap sdcard0 and sdcard1 around (makes external SD card primary). This also includes a modified mount.exfat which fixes crashes caused by using large SD cards (>32GB). See links at the bottom.
Speeds up to 1.4Ghz. I have no issues with 1.4, but mileage may vary per device. Try the LV (Low Voltage) first. If this does not work, try the HV. If neither work, you may just have to use one that maxes out at 1.3Ghz.
Welcome RAZR users.
Changes:
Update: Sync with upstream: just some fixes for MM, OMM and SMP
Update: 10-04-15: Added SDSwap zip file.
Update: 09-22-15
Includes new BFQ I/O scheduler. This is supposed to improve some of the laginess. This is the new default I/O scheduler.
Includes a modified libthermal-manager.cfg. This appears to fix the issues with the CPU getting stuck at 1Ghz when using the "Interactive" governor.
Shortened name of files so it's easier to see the full file name from recovery.
Update: 07-25-15 - List of changes:
Includes an additional 410 kernel patches, bringing it much closer to 3.0.101
Ktoonsertative sampling_rate now defaults to 30000 (that's for you @septfox )
Update: 07-14-15 - Added an addtitional 75 upstream patches from 3.0.101 in general kernel, mmc, ext3, fat, drivers bass, power management and hwmon.
Update: 07-12-15 - Added an additional 14 mem manger patches bringing the total to 116. (All of these are now in the mainline CM12.x kernel/nightlies after 07-15)
Update: 07-09-15 - List of changes:
Fixed missing ZRam settings during build
102 patches included to the kernel memory manager from the mainline Linux kernel up to 3.0.101 (These are not in D4 CM12.x stock.)
Sync to upstream. This enables a 256MB ZRam with LZ4 compression (nightly for 07-08 and up)
New SDSwap to accompany here. (Note: sorted by date). This is ONLY required if you want to Swap SD Cards 0 and 1 and is needed for stock OR my kernel.
***Repeat: NOT REQUIRED for my kernel unless you want swap.***
Update: 07-08-15 - List of changes: Pulled. Missing ZRAM in config during build. Oopsie Links are dead for now, new build on the way with more patches.
Update: 06-18-15 - List of changes:
Sync with upstream. Mostly binder fixes in the IPC layer. You can see all the changes involved here.
Added ULV (Ultra-Low Voltage) option for all 1.4Ghz enabled kernels. These kernels run the 1.4Ghz frequency at 1.409v instead of 1.415 (LV) or 1.439 (HV). I noticed some throttling at times on 1.415 due to heat and this seemed to help. Obviously, not all devices are going to like this so if it doesn't boot, just use the normal voltage you have been using.
Update: 05-17-15 - Sync with upstream. Fix CPU Frequency policy from getting stuck when user & kernel min/max don't overlap.
Update: 05-15-15 - Sync with upstream. This includes the new CPU-BOOST driver @stargo added. This will be enabled with tonight's (05-16) nightly. You can use this with older builds but the driver will do nothing.
Update: 04-22-15 - List of changes:
Added IntelliPlug 3.8 support. This allows hotplugging independent of the governor chosen. Also, it currently will only shut down cpu1 when the screen is off. No SODs so far in testing and I notices a nice boost in battery life. I was initially using version 4.0, but this too caused SOD. Thanks to @dtrail1, I've used a modified 3.8 that seems to make these Omap4's happy.
KToonServative governor adjustments: sampling_rate_min is now 15,000 (I'm looking at you, @Septfox), and hotplugging is force disabled as the setting the default flag does not work correctly. You cannot enable even if you wanted to for now.
Update: 04-15-15 - List of changes:
Added Lionheart governor
Set default profile of ZZMove to 1, disabled Hotplugging
Sync to upstream: Use -O2 instead of -Os for faster zram and disk performance.
Update: 04-04-15 - List of changes
sync to upstream: disable memory resource controller for cgroups
Update: 03-26-15 - List of changes
Reduced 1.4Ghz LV voltage to 1.415 and increased 1.4Ghz HV to 1.439
Removed modified Power HAL since changes were merged into official CM device tree
Update: 03-20-15 - Updated all non fullscale 14 kernels. 1300 was left in but should not have been.
Update: 03-19-15 - List of changes
Added links for individual PowerHALs for maserati, spyder and umts spyder. See in downloads section below.
Added the following I/O schedulers: ROW, SIO, SIO Plus, VR and Zen
Added the following governors: InteractiveX, Ktoonservative, SmartassH3 and ZZMoove.
Updated all kernels with the new schedulers, governors and Power HALs. All kernels also contain the modded power HAL for Maserati, Spyder and UMTS Spyder.
All downloads now hosted at new location. Please let me know if any issues are encountered.
Update: 03-17-15 - List of changes
New download location thanks to @stargo : http://droid.cs.fau.de/jjb666/Kernels/
Work has begun on adding some additional governors, currently SmartAssH3 and InteractiveX are running. A test kernel is listed below. It's currently set for Full Scale 1.4Ghz and the default governor is still interactive.
Update: 03-09-15 - List of changes:
Included modified power hal module to better set some of the scale settings and add some logging.
Added options for full range of frequencies avail: 300Mhz, 400Mhz, 500Mhz, 600Mhz, 700Mhz, 800Mhz, 900Mhz, 1.0Ghz, 1.1Hgz, 1.2Ghz, 1.3Ghz and 1.4Ghz
Added Low Voltage and High voltage version of all 1.4Ghz enabled kernels (LV = Low Voltage 1.42, HV = High Voltage of 1.435)
Update: 03-04-15 - Added additional options (downloads at bottom):
Run 'N Gun - Has the following frequencies enabled: 300, 1Ghz, 1.2Ghz and 1.4Ghz
Run 'N Gun Lite - Has the following frequencies enabled: 300, 1Ghz, 1.2Ghz and 1.3Ghz
Update: 03-02-15
Max to 1.4Ghz.
Available speeds 300Mhz, 600Mhz, 800Mhz, 1.2Ghz and 1.4Ghz
02-26-15
Initial release.
Available speeds 300Mhz, 600Mhz, 800Mhz, 1.2Ghz and 1.3Ghz
Downloads:
12-05-15 CM12 Run 'N Gun Kernel ULV (Ultra-Low Voltage) (300Mhz, 1Ghz, 1.2Ghz and 1.4Ghz)
12-05-15 CM12 Run 'N Gun Kernel LV (Low Voltage) (300Mhz, 1Ghz, 1.2Ghz and 1.4Ghz)
12-05-15 CM12 Run 'N Gun Kernel HV (High Voltage) (300Mhz, 1Ghz, 1.2Ghz and 1.4Ghz)
12-05-15 CM12 Run 'N Gun Lite Kernel (300Mhz, 1Ghz, 1.2Ghz and 1.3 Ghz)
12-05-15 CM12 OC 1.3Ghz Kernel (300Mhz, 600Mhz, 800Mhz, 1Ghz, 1.2Ghz and 1.3 Ghz)
12-05-15 CM12 OC 1.4Ghz Kernel ULV (Ultra-Low Voltage) (300Mhz, 600Mhz, 800Mhz, 1Ghz, 1.2Ghz and 1.4 Ghz)
12-05-15 CM12 OC 1.4Ghz Kernel LV (Low Voltage) (300Mhz, 600Mhz, 800Mhz, 1Ghz, 1.2Ghz and 1.4 Ghz)
12-05-15 CM12 OC 1.4Ghz Kernel HV (High Voltage) (300Mhz, 600Mhz, 800Mhz, 1Ghz, 1.2Ghz and 1.4 Ghz)
12-05-15 CM12 OC Full Scale Lite 1.3Ghz Kernel (300Mhz, 400Mhz, 500Mhz, 600Mhz, 700Mhz, 800Mhz, 900Mhz, 1Ghz, 1.2Ghz 1.2Ghz and 1.3 Ghz)
12-05-15 CM12 OC Full Scale 1.4Ghz Kernel ULV (Ultra-Low Voltage) (300Mhz, 400Mhz, 500Mhz, 600Mhz, 700Mhz, 800Mhz, 900Mhz, 1Ghz, 1.1Ghz, 1.2Ghz, 1.3 Ghz and 1.4Ghz)
12-05-15 CM12 OC Full Scale 1.4Ghz Kernel LV (Low Voltage) (300Mhz, 400Mhz, 500Mhz, 600Mhz, 700Mhz, 800Mhz, 900Mhz, 1Ghz, 1.1Ghz, 1.2Ghz, 1.3 Ghz and 1.4Ghz)
12-05-15 CM12 OC Full Scale 1.4Ghz Kernel HV (High Voltage) (300Mhz, 400Mhz, 500Mhz, 600Mhz, 700Mhz, 800Mhz, 900Mhz, 1Ghz, 1.1Ghz, 1.2Ghz, 1.3 Ghz and 1.4Ghz)
12-05-15 CM12 Custom NON-OC RunNGun Kernel (300Mhz, 1Ghz and 1.2Ghz)
SD Card Swapper zip
10-05-15 SD Card Swapper
Older/Archived kernels can be downloaded here if needed: http://droid.cs.fau.de/jjb666/Kernels
Power HALs to correct speed setting issues (No longer needed as of 3/23 nightlies and 3/26 kernels)
power.maserati.so
power.spyder.so
power.umts_spyder.so
Disclaimer: You did it, not me.
XDA:DevDB Information
Custom 1.3Ghz and 1.4Ghz OC Kernel for Droid 4/RAZR on CM12, Kernel for the Motorola Droid 4
Contributors
joojoobee666, stargo, Hashcode
Source Code: https://github.com/JooJooBee666/android_kernel_motorola_omap4-common
Kernel Special Features: Overclock to 1.3Ghz and 1.4Ghz
Version Information
Status: Testing
Created 2015-02-27
Last Updated 2015-12-29

Antutu seems to think it's alright.
Overall score rose by ~800 (15400 -> 16200)
Multitask rose by ~200
Runtime rose by ~100
CPU Integer and Float-point rose by ~60
Single-thread Integer and Float-point rose by ~100
Not bad at all for a 100mhz bump, about what I would expect.
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}

If this is similar to the JBX kernel's OMAP4 overclocking functions, you should be able to use trickstermod to overclock the MPU (basically, like overclocking a PC's front side bus).
Stock MPU is 100, CPU multiplier is 13 for 1.3ghz. If you increase MPU to 108, you should get 1404mhz out of the CPU (13x108=1404). I ran 1391mhz for about 5 months straight without any stability issues, I think the CPU voltage was at 1.360v to achieve this stable.
Does this kernel also allow OCing the GPU to 384mhz? I would just try it if I still had my D4 (replaced with photon Q, sold to sister).

Elanzer said:
If this is similar to the JBX kernel's OMAP4 overclocking functions, you should be able to use trickstermod to overclock the MPU (basically, like overclocking a PC's front side bus).
Stock MPU is 100, CPU multiplier is 13 for 1.3ghz. If you increase MPU to 108, you should get 1404mhz out of the CPU (13x108=1404). I ran 1391mhz for about 5 months straight without any stability issues, I think the CPU voltage was at 1.360v to achieve this stable.
Does this kernel also allow OCing the GPU to 384mhz? I would just try it if I still had my D4 (replaced with photon Q, sold to sister).
Click to expand...
Click to collapse
At this point, no GPU OC. It's just running at the stock 304. I planned on testing MPU @ 1.4 and possibly adding some intermediate speeds. GPU possible too, but I had to start somewhere.

Can you commit this in official CM repository? Just to make sure that custom kernel is always compatible with last version of CM.

nicolay.n said:
Can you commit this in official CM repository? Just to make sure that custom kernel is always compatible with last version of CM.
Click to expand...
Click to collapse
It doesn't work that way. If he has this on his own private gh and tracks omap4-common commits it will always be compatible. Btw our board has a max freq of 1.5Ghz. I used to run my bionic on 1.490 and it was fast but got too hot. 1.35/1.4 is a more stable freq to work with.
Sent from my XT894 using XDA Free mobile app

The problem with 1.4ghz is, it seems to be where things go pear-shaped unless phone-specific amounts of voltage are pushed. A one-size-fits-all approach for voltage will probably work, but then you're inevitably going to overvolt some devices that don't need it.
MPU and voltage control, along with an extra multiplier at the bottom end (to compensate for the higher MPU), would be ideal. Probably a fair bit of work, though.
GPU overclocking is apparently not as productive as you would think for the increase in frequency and power/heat (I've heard that the bottlenecks are elsewhere in the system, and it seems like the increase in benchmark scores under KK didn't strike me as impressive). Might get more out of it overall by just focusing on increasing the CPU speed.
My 2c, anyway. Whatever you decide to do, it improves what we have, and for that you have my appreciation.

Septfox said:
The problem with 1.4ghz is, it seems to be where things go pear-shaped unless phone-specific amounts of voltage are pushed. A one-size-fits-all approach for voltage will probably work, but then you're inevitably going to overvolt some devices that don't need it.
MPU and voltage control, along with an extra multiplier at the bottom end (to compensate for the higher MPU), would be ideal. Probably a fair bit of work, though.
GPU overclocking is apparently not as productive as you would think for the increase in frequency and power/heat (I've heard that the bottlenecks are elsewhere in the system, and it seems like the increase in benchmark scores under KK didn't strike me as impressive). Might get more out of it overall by just focusing on increasing the CPU speed.
My 2c, anyway. Whatever you decide to do, it improves what we have, and for that you have my appreciation.
Click to expand...
Click to collapse
Yeah, well this is all in test phase anyhow. I just didn't see any other CM12 compatible kernels with OC around at this point and figured it would be a nice gap to fill. I'm still planning on additional testing with voltage/clocks and maybe eventually adding different governors, etc. However, I'm still fairly wet behind the ears on Linux kernels in general so don't expect a lot of updates quickly. Though I plan on keeping this and future variations in line with stargo's cm12 kernel so there's always an alternative to stock.
nicolay.n said:
Can you commit this in official CM repository? Just to make sure that custom kernel is always compatible with last version of CM.
Click to expand...
Click to collapse
Probably never going to happen, these mods would never be approved for official CM. Again, I do plan on keeping this updated to work with D4/Moto Omap4 CM12 as long as I have my crappy D4.

joojoobee666 said:
as long as I have my crappy D4.
Click to expand...
Click to collapse
Take that back! Lol the droid4 is awesome. I love the hardware keyboard. I would say that I wish they'd make a droid5 but with the turbo.....god knows they'd make it unhackable..
Sent from my XT894 using XDA Free mobile app

Oggie7797 said:
Take that back! Lol the droid4 is awesome. I love the hardware keyboard. I would say that I wish they'd make a droid5 but with the turbo.....god knows they'd make it unhackable..
Sent from my XT894 using XDA Free mobile app
Click to expand...
Click to collapse
Hardware QWERTY is the ONLY reason I'm still here

joojoobee666 said:
Yeah, well this is all in test phase anyhow. I just didn't see any other CM12 compatible kernels with OC around at this point and figured it would be a nice gap to fill. I'm still planning on additional testing with voltage/clocks and maybe eventually adding different governors, etc. However, I'm still fairly wet behind the ears on Linux kernels in general so don't expect a lot of updates quickly.
Click to expand...
Click to collapse
Don't get me wrong, I wasn't trying to criticize what you've done or how you're going about it. Sorry if I came off sounding like...er...something unpleasant, I don't know. Wasn't my intention.
-
Thus far, no problems with the OCed kernel. The extra bit of UNFH is especially visually noticeable in the Store, where scrolling has smoothed a bit more, but the phone overall feels a bit faster. Good stuff.
Can't really say anything in the way of extra battery drain yet.

joojoobee666 said:
Hardware QWERTY is the ONLY reason I'm still here
Click to expand...
Click to collapse
Right lol if $700 would ever appear in my wallet id go get me a nexus6 id overlook no keyboard then.
Sent from my XT894 using XDA Free mobile app

Septfox said:
Don't get me wrong, I wasn't trying to criticize what you've done or how you're going about it. Sorry if I came off sounding like...er...something unpleasant, I don't know. Wasn't my intention.
-
Thus far, no problems with the OCed kernel. The extra bit of UNFH is especially visually noticeable in the Store, where scrolling has smoothed a bit more, but the phone overall feels a bit faster. Good stuff.
Can't really say anything in the way of extra battery drain yet.
Click to expand...
Click to collapse
You got yourself wrong. I didn't take it offensively. at all

Work in to CM11?
Sent from my XT894 using XDA Free mobile app

maisdoiscorregos said:
Work in to CM11?
Sent from my XT894 using XDA Free mobile app
Click to expand...
Click to collapse
Idk why it wouldn't....flash it and if it fails to boot simply wipe /system and flash cm11+gapps again. It won't hurt anything or brick you to the point you need an sbf
Sent from my XT894 using XDA Free mobile app

#joojoobee666, could carry a CM11 kernel with OC?
Sent from my XT894 using XDA Free mobile app

maisdoiscorregos said:
#joojoobee666, could carry a CM11 kernel with OC?
Sent from my XT894 using XDA Free mobile app
Click to expand...
Click to collapse
First off, to tag on xda use a @ and secondly, not to my knowledge and making one would be a little counterproductive imho since lollipop is kind of the new thing. Could always test flash this on cm11 literally the worst thing that could happen is it won't boot and that's a basic fix: wipe system and reflash cm11 and gapps. No data will be lost.
Sent from my XT894 using XDA Free mobile app

Oggie7797 said:
First off, to tag on xda use a @ and secondly, not to my knowledge and making one would be a little counterproductive imho since lollipop is kind of the new thing. Could always test flash this on cm11 literally the worst thing that could happen is it won't boot and that's a basic fix: wipe system and reflash cm11 and gapps. No data will be lost.
Sent from my XT894 using XDA Free mobile app
Click to expand...
Click to collapse
Excuse me, I have 84 apps installed on CM11, if I install the same 84 apps on CM12, I have a very slow, making it impossible to use, can barely answer a call. I agree that CM12 is evolution, but as he is extremely slow exporadicamente, CM11 use in stock slot, and CM12 in slot 1, have a kernel overclocked to CM11, would be good for multiple users of CM11 as daily rom.

maisdoiscorregos said:
Excuse me, I have 84 apps installed on CM11, if I install the same 84 apps on CM12, I have a very slow, making it impossible to use, can barely answer a call. I agree that CM12 is evolution, but as he is extremely slow exporadicamente, CM11 use in stock slot, and CM12 in slot 1, have a kernel overclocked to CM11, would be good for multiple users of CM11 as daily rom.
Click to expand...
Click to collapse
You must not be reading what I'm saying.....make a nandroid in twrp. Then flash this kernel. If it won't boot just restore the nandroid. Test to see if it works. If it does then that'll be cool
Sent from my XT894 using XDA Free mobile app

Oggie7797 said:
You must not be reading what I'm saying.....make a nandroid in twrp. Then flash this kernel. If it won't boot just restore the nandroid. Test to see if it works. If it does then that'll be cool
Sent from my XT894 using XDA Free mobile app
Click to expand...
Click to collapse
I already tested and does not work