Database Develop

Emphasis on BFS

Hello everyone,
I'm the maintainer of the Liquorix kernels for Debian / Ubuntu in the linux world. By default, this kernel is setup to use BFS, plus all the other patches from zen-sources and a finely tuned configuration to reduce overhead and increase throughput / reduce latency.
I recently bought the Nexus One to replace my old phone and have been fumbling around the forums, just looking around and putting on mods like Cyanogen. Just skimming through the selection of kernels, I'm very surprised.
How come CFS is still the scheduler of choice? Many users have exclaimed in posts about the increase in speed and fluidity - very much the same reaction for users switching to my kernels on Debian. The only developer here using BFS seems to be kingslick, but I don't know which version of the kernel he's using exactly, it just says 2.6.33 even though 2.6.33.2 is out per the day of this posting.
How come there has been a slow uptake on the BFS scheduler? Is it difficult to patch, not enough feed back, instability, etc... ???

That would explain why his kernels always seem lightning quick, lol.

Where do you see King using BFS. In his Nexus Desire Rom he states in the OP he has switched back to CFS

if i remember correctly atleast when I was using the g1 bfs was causing instability issues, I dont think i've heard anything about bfs on the nexus so I cant say one way or the other why it isn't used more often.

BFS used to be good on the ARM architecture, but not anymore. Its instable and actually slows down the phone.

which version is unstable? Since the 300 series of BFS, the final stability bugs were fixed AFAIK.

damentz said:
which version is unstable? Since the 300 series of BFS, the final stability bugs were fixed AFAIK.
Click to expand...
Click to collapse
and the 300 series runs poorly on our phones

*puts up hand* whats bfs?

dogiedogie said:
*puts up hand* whats bfs?
Click to expand...
Click to collapse
BFS = Brain **** Scheduler. It controls the CPU and other processes at the kernel level

dogiedogie said:
*puts up hand* whats bfs?
Click to expand...
Click to collapse
these might help:
http://ck.kolivas.org/patches/bfs/bfs-faq.txt
http://ck.kolivas.org/patches/bfs/sched-BFS.txt

persiansown said:
and the 300 series runs poorly on our phones
Click to expand...
Click to collapse
Where can I find a forum thread discussing this? I've been only seeing good things, not bad.

http://www.google.com/search?q=BFS+site:xda-developers.com
As you can see there were many BFS roms earlier (3xx version too), but nearly all developers switched back to CFS cause of stability problems.

I'm seeing a lot of ports to 2.6.29 for BFS. If those are amateur ports, I would expect instability and unpredictable bugs.
BFS is always moving forward and the newer versions are based off of the currently stable kernel, in this case, 2.6.33 at the time of this posting.

damentz said:
I'm seeing a lot of ports to 2.6.29 for BFS. If those are amateur ports, I would expect instability and unpredictable bugs.
BFS is always moving forward and the newer versions are based off of the currently stable kernel, in this case, 2.6.33 at the time of this posting.
Click to expand...
Click to collapse
Pretty much every single Kernel developer has given BFS a go, and all of them have had the same end result. Even Cyanog and pershoot.

damentz said:
Where can I find a forum thread discussing this? I've been only seeing good things, not bad.
Click to expand...
Click to collapse
I generally talk about it in IRC

I spoke with Con about the poor performance post 200 series on these Android phones, and he said that he wasn't aware of anything changed since then.
...but also, no one had brought these bugs to his attention, so nothing got fixed.
If anyone's willing to maintain a BFS kernel for android with the newest upstream code, I can test it and try to identify what the problem might be. There's a very large potential for performance to increase on the android platform if BFS becomes the optimal choice - we cannot let this slip unnoticed!

BFS is poorly suited to android.
Reason is this:
Android has a timeout function -- when an application takes too long to respond, it issues an "application not responding" dialog. This includes BACKGROUND processes. BFS, of course, favors FOREGROUND processes, so when you have a foreground process sucking up all the CPU (because BFS is unwilling to give over the necessary CPU time to the background processes), those background processes end up taking longer than the timeout allows, which means that any time you run a CPU intensive foreground process, all the background processes will go ANR.
CFS doesn't have this problem. Background processes will still get their fair share of CPU time.
Remember that these devices are PHONES first and foremost! If the phone rings, the process handling the phone ringing MUST be able to run in a sensible amount of time.

I remember BFS on Cyanogen back with the G1. It actually did give a pretty quick feeling phone. Like others said, it had its stability issues though.
But, maybe you can talk one of the kernel devs (intersectRaven or Ivan or pershoot or someone) into maintaining a separate BFS kernel along with the ones they already do. They might turn it down as its alot of extra work.. but, you never know!
I'd be willing to test the kernels out on my phone if someone was willing to actively maintain one for testing.

lbcoder said:
BFS is poorly suited to android.
Reason is this:
Android has a timeout function -- when an application takes too long to respond, it issues an "application not responding" dialog. This includes BACKGROUND processes. BFS, of course, favors FOREGROUND processes, so when you have a foreground process sucking up all the CPU (because BFS is unwilling to give over the necessary CPU time to the background processes), those background processes end up taking longer than the timeout allows, which means that any time you run a CPU intensive foreground process, all the background processes will go ANR.
CFS doesn't have this problem. Background processes will still get their fair share of CPU time.
Remember that these devices are PHONES first and foremost! If the phone rings, the process handling the phone ringing MUST be able to run in a sensible amount of time.
Click to expand...
Click to collapse
That's dumb, CFS is the scheduler that is unfair. CFS uses a red black tree with interactivity heuristics to create a timeline for processes to run for a specific amount of time. BFS on the other hand uses deadlines to choose which process runs next, using absolutely no heuristics. The shortest deadline should run first unless a process with a lower nice value preempts it.
I would expect BFS to be more responsive with your tests, since no application should be starved from heuristics. That's why it also performs better with games and 3d applications, since the cpu time is given fairly and in steady amounts, allowing for a more enjoyable graphical experience.
BFS should also run much faster on slower cpus because it has very low execution runtime memory footprint. I remember when BFS was first starting, one of the lesser known difference between BFS and CFS were the cache misses on CPUs with low amounts of L2 cache. Of course, no one put these online since it was just fun to experiment with.
But, maybe you can talk one of the kernel devs (intersectRaven or Ivan or pershoot or someone) into maintaining a separate BFS kernel along with the ones they already do. They might turn it down as its alot of extra work.. but, you never know!
Click to expand...
Click to collapse
Yes, that would be cool. I'll also help with these kernels and get some of the most technically helpful bug reports to Con so these bugs can be fixed once and for all.

Compiled a BFS kernel as per your request. It can be found on my MediaFire folder. Not really sure if BFS is active though since I only found one entry in menuconfig which is in the general section and not where I expected it to be (with CFQ and Deadline options).
*Just read more...I was confusing CFS with CFQ...

[KERNEL][AOSP][2.6.32.26] - Savage-Kernel-Inc v1 ! [HAVS+CFS+SBC!!] UPDATED 1/06!

I am proud to announce Savage-Kernel to the xda forms. Currently in beta testing we worked from the latest invisiblek kernel newly undervolted, HAVS with the SBC mod.
JUST FOR @PonsAsinorem:::::: DISCLAIMER!!!! IM NOT RESPONSIBLE FOR BRICKING YOUR PHONE... EVEN THOUGH THE KERNEL CANNOT BRICK A ROOTED PHONE UNLESS YOU FLASH A BAD RECOVERY..... DON'T MODIFY THE UPDATER-SCRIPT TO FLASH THIS TO YOUR RECOVERY!!!!!! .... ALL CHANGES TO THIS KERNEL STILL MEET THE MINIMUM SPECS OF THE INC.... UNLESS YOU OVERCLOCK... WHICH IS UNDERVOLTED... WHICH IS RUNNING AT THE STOCK VOLTAGE AND SHOULDN'T HARM YOUR PHONE ANYWAY.
SENSE BASED ROMS NOT SUPPORTED THIS IS FOR AOSP ROM'S ONLY
So far the Kernel works great in conservative governor. which all the devs prefer.
Devs:
jsseidel1 (co-creator)
luisrcastillo (co-creator)
ms79723 (SBC mod updates)
lithid-cm (contributer. Updater-script guru)
Credits:
invisiblek--> base kernel!!
ms79723 --> the SBC mod.
netarchy --> smartass gov
intersectRaven --> HAVS
(credits are given to the dev's where we got some code from, may no be original designer)
CFS Builds:
Alpha1 : Click Here to download
Stable v1 : Click Here to download
BFS Builds:
Follow the github on twitter @evokernelsrc
If you like our kernel and want to support us... donate (no obligation)

Change log
Code:
[B]Alpha 1[/B]
Undervolted to 925
HAVS implemented configured for conservative
smartass added for fans
SBC Mod added (trickle Charging)
Enabled OC to 1.19 and stable
conservative governor tweaks
HAVS tweaks
smartass governor tweaks
voltage tweaks
BFS360 (latest version) and CFS VERSION!!!
Added SBC v7
BFS tweaks to ondemand gov
tweaks to interactive gov
bad ass CFS tweaks!!!!!
aggressive swap drop
misc tweaks
[B]Stable v1[/B]
SBC v7 ported from Evo (better battery life, this is a HTC mod so doesnt matter that it came from Evo)

FAQ's
All answers are based off latest stable build
Which governor should I use?
Use Conservative for best battery, if you get any lag or choppy or shutter switch to either Smartass or interactive.
Which governors are tweaked?
Conservative and Smartass
I'm running smartass and having screenwake issuesm what do I do?
We suggest you set up a screen off profile of 245/384 to avoid the wake issue.
Have any more? Just ask!!

jsseidel1 said:
All answers are based off latest stable build
Which governor should I use?
Use Conservative for best battery, if you get any lag or choppy or shutter switch to either Smartass or interactive.
Which governors are tweaked?
Conservative and Smartass
I'm running smartass and having screenwake issuesm what do I do?
We suggest you set up a screen off profile of 245/384 to avoid the wake issue.
Have any more? Just ask!!
Click to expand...
Click to collapse
What kind of battery life are you getting with this? I know results vary, but just curious.
Sent from my ADR6300 using XDA App

excited to try this

jsseidel1 said:
All answers are based off latest stable build
Which governor should I use?
Use Conservative for best battery, if you get any lag or choppy or shutter switch to either Smartass or interactive.
Which governors are tweaked?
Conservative and Smartass
I'm running smartass and having screenwake issuesm what do I do?
We suggest you set up a screen off profile of 245/384 to avoid the wake issue.
Have any more? Just ask!!
Click to expand...
Click to collapse
How did you determine that conservative is the most power-friendly?

I would post a disclaimer about the SBC mod.
Sent from my ADR6300 using Tapatalk

Am I missing the link here or has it not been posted yet? I'm excited to test.

Guys, SBC *really* isn't good for the battery. Li-Ion batteries aren't designed to use trickle charging. IMO this will kill the batteries faster and isn't worth the extra charging.

waiting to try this... dont think it will hurt my battery in the short run... who keeps phones longer than 2 years anways?

True. I've gotten better battery life off of it, but I feel like my MIUI build recently has been getting worse due to the SBC mod.

Works on gingerbread roms?
Sent from my ADR6300 using XDA App

If it's AOSP, it should, but sometimes I've gotten weird reactions using kernels that are claimed to work on all AOSP ROMs (Kingx Kernel + MIUI = no-no. Also, MIUI + Hydra = no-no).
Point being, we should probably get this clarified. I've seen instances where people have lost touchscreen functionality (though usually that's due to a failure to wipe cache/dalvik) among other issues. Until we field test this on GB ROMs or until the OP states that it works on them, we should probably refrain from flashing it over a GB ROM. You can do it if you want, I'm in no way trying to make you turn away from it; I'm just saying take it with a grain of salt.
Plus there's the SBC issue to worry about; trickle-charging has negatively affected the battery performance of my MIUI ROM, so I'm going to switch kernels whenever I go back to it. Better charging in the short run < not needing to buy new batteries in the long run (and therefore saving money).

link??????

*sigh*...
Ok, clearly I need to start citing the comments I make. Here you go:
http://forum.xda-developers.com/showthread.php?t=885476&page=25
The first post on this last page pretty much sums it up (actually, read 23-25 to sum up the main issues, it also has links to OTHER posts talking about the issues), but if you want other chronological examples of touchscreen failure or other issues you can read examples from pages 2, 5, and some places past page 15. SBC is bad news on its own, and adding overclocking to it just makes it worse. Again, you can use the kernel if you want, I'm not gonna stop you, I'm just saying that there have BEEN problems with it on other AOSP ROMs. MIUI seems to take pretty well to SBC in terms of battery, as I've seen; but, it's been getting worse as of late. UltimateDroid had a touchscreen failure (page 5) in v1; I don't believe that's been fixed yet, but feel free to correct me. MIUI had a touchscreen failure in v2 (though I think there might have been a bad flash involved, because I flashed SBC to my MIUI 12.24 at the time and it worked fine), cited page 23. Need I go on? Lol.
Point is (I've said that a lot, I know) that it does damage, multiple AOSP ROMs are getting touchscreen failure, and there's more harm than good that comes out of it.

Read: http://www.ziggy471.com/2011/01/02/overcharging-batteries/

That's a good link also. It's actually in the thread I posted, but its equally as important if not more important than my post. Good job man XD
Sent from my HTC Incredible running Carbonite Red 1.5.

Ok you all need to relax. SBC does not over charge your battery. The trickle charging is essentially the battery staying at 100% while on the charger. Lithium ion batteries have tech built into them to protect from over charging and such. SBC just means when you take it off the charger it will really be at 100%. You can't over charge aanyway. This kernel allows you to overclock but it doesn't force it. You need to choose overclocking yourself in setcpu. ... and 90% of the kernels in this forum have SBC they just either don't say it or they call it the battery fix. You can see it in their gits. On the evo forums we just call it SBC. Ill be posting the kernel today. If you don't want to use it. Then don't. It was just requested that I port my evo kernel over to the Inc. So I am.
Sent from my PC36100 using XDA App

gmichaelow said:
How did you determine that conservative is the most power-friendly?
Click to expand...
Click to collapse
Conservative by nature is more power friendly as it only ramps up the processor as requested then brings it right back down to your min.
The way we have set our Governor is to optimize the Conservative Governor to get the most speed stability and battery life out of your phone.
I just got a config from one of your rom dev's to build the kernel so this should be posted by tomorrow.

I for one appreciate what you are doing. There is nothing wrong with having a variety of kernels to choose from. If you disagree with what this dev is doing, then use another kernel. We don't need to be running them off.
Sent from my ADR6300 using XDA App

[Kernel][3.0.84][CK3] Mangekyou-Kernel V3 [03-07-2013]

​
In this thread you may find my customized kernel 3.0.x for JB 4.2.2 rom (CM10.1 and derivatives).
I hope it will be useful, and ensures better performance and battery life.
Of course, your feedback, as always, are welcome and certainly valuable.
Main Features:
-- Kernel 3.0.86
-- Compiled with Linaro 4.7.4 modded by christopher83
-- Sources synched with Cm10.1 branch
-- Built with Cortex-A8 and NEON VFP optimization flags
-- Kernel is very Stable
-- 374 mb of free RAM
-- Init.d support
-- Battery Life eXtender support ( thanks ezekeel and ryuinferno)
-- Optimized AES and SHA1 routines for ARM cpu architecture
-- Optimized ARM RWSEM (read/write semaphore) algorithm
-- Optimized asynchronous I/O latency on solid state disks
-- Optimized string and memcopy libs
--Fast Random generator (frandom) support
-- optimized LPDDR RAM timings (~10% throughput increase)
-- Dynamic management of dirty page writebacks
-- Added optimized CRC32 algorithm.
-- Voku and Dorimanx Tweak for LowMemoryKiller
-- Disable Gentle Fair Sleep for better perfomance
-- Asynchronous I/O latency
-- Lower swappiness 600----->45
-- Add -O3 optimization for modules
-- CK3 con Kolivas Tweaks
Avaible Cpu Governors
-- PegasusQ
-- ZenX
-- Nightmare
-- Ktoonservative
-- Sakuractive
-- HYPER
-- LagFree
-- AbyssPlug
-- Wheatley
-- HotPlug
-- Conservative
-- Perfomance
-- UserSpace
-- PowerSave
-- Ondemand
-- Interactive
Avaible I/O schedular
-- VR
-- Sio
-- Row
-- CFQ
-- DeadLine
-- Noop
Demon-Kernel MediaFire Folder
Sources : My sources of DemonProjects and other projects
CREDITS:
BIG THANKS FOR CHRISTOPHER FOR HIS HELP !!
Mnazim
fuss
Codeworkx for all
Ryuinferno
DOOMSDAY94
Diablo55 for helped my
Christopher83 for his toolchain and his source
voku to vokulmk
cyanogen team =D
and other =))

ChangeLog :
Code:
[B]Mangekyou-Kernel V2 03-07-2013[/B]
adjustes some file
fix name of kernel in defconfig
add [B]CK3[/B] Con Kolivas Tweaks !
[B]Mangekyou-Kernel 02-07-2013[/B]
- Change name of kernel ( because i use mangekyou image in my github and i like this name :P )
- Update LMK
- Voku and Dorimanx Tweak for LowMemoryKiller
- Disable Gentle Fair Sleep for better perfomance with good battery status
- Asynchronous I/O latency to a solid-state disk greatly increased
- Lower swappiness 600----->45
- Add -O3 optimization for modules
[B]Demon-Kernel 01-07-2013[/B]:
-Initial Release

wow first rep........
is it inspired from vampire diaries character........:silly:
just kidding.......
only on 4.2.2 aosp roms???????

for the moment...yes

biagio7xD said:
for the moment...yes
Click to expand...
Click to collapse
whokay then its time to flash new rom after a neat's release.........:silly:
will give feebback by tom may be with screenies.......
will flash on mokee......

I like the sharingan icon so much
Mangekyo sharingan of Itachi
Sent from my GT-I9100G using xda premium

Project Nexus kernel is dead?

Whats the different between this n nexus kernel??
Sent from my GT-I9100G using xda app-developers app

cyelanius said:
Whats the different between this n nexus kernel??
Sent from my GT-I9100G using xda app-developers app
Click to expand...
Click to collapse
nexus kernel have more features and nexus is a perfomance kernel.....not battery friendly...this kernel is more stable and more battery friendly...this is better than nexus
EDIT: release version 20130702.... Mangekyou kernel

I had just installed your damon kernel and was ready to thank you, but saw that there was already a new version. Whooow, if you're speaking of fast, this is fast. So thanks man, and do so on I would say. Doing well. BTW, downloading Mangekyou kernel now!
Verstuurd van mijn GT-I9100G met Tapatalk

wackowizzard said:
I had just installed your damon kernel and was ready to thank you, but saw that there was already a new version. Whooow, if you're speaking of fast, this is fast. So thanks man, and do so on I would say. Doing well. BTW, downloading Mangekyou kernel now!
Verstuurd van mijn GT-I9100G met Tapatalk
Click to expand...
Click to collapse
It is just a rename, no changes yet...as seen from his GitHub...expect if he did not update it before releasing as per real world law...
Hi, at first I wanted to congratulate you on releasing your kernel, but you disappointed me again:
Misconception:
-- Init.d support [all AOSP kernel supports this, it is a feature in stock but not AOSP, already told you before]
-- Add -O3 optimization for modules [not done anywhere]
Below are the features from Slim kernel:
-- Optimized AES and SHA1 routines for ARM cpu architecture
-- Optimized ARM RWSEM (read/write semaphore) algorithm
-- Optimized asynchronous I/O latency on solid state disks
-- Optimized string and memcopy libs
-- Voku and Dorimanx Tweak for LowMemoryKiller
-- Disable Gentle Fair Sleep for better perfomance
-- Asynchronous I/O latency
-- Lower swappiness 600----->45 [it's 60]
What you did:
Compiled with Linaro 4.7.4
optimized LPDDR RAM timings (~10% throughput increase)
add Frandom v2
Added NEON compilation flag
mm: Dynamic management of dirty page writebacks
Battery Life eXtender (BLX)
add New Governor From Nexus Kernel
Added optimized CRC32 algorithm
So this kernel is very based on Slim kernel...and why didn't you state it and credit @mnazim properly?
And, removal of sources of the old Nexus kernel means you cannot share it in public anymore...didn't you learn anything since the Semaphore kernel incident?

Ryuinferno said:
It is just a rename, no changes yet...as seen from his GitHub...expect if he did not update it before releasing as per real world law...
Hi, at first I wanted to congratulate you on releasing your kernel, but you disappointed me again:
Misconception:
-- Init.d support [all AOSP kernel supports this, it is a feature in stock but not AOSP, already told you before]
-- Add -O3 optimization for modules [not done anywhere]
Below are the features from Slim kernel:
-- Optimized AES and SHA1 routines for ARM cpu architecture
-- Optimized ARM RWSEM (read/write semaphore) algorithm
-- Optimized asynchronous I/O latency on solid state disks
-- Optimized string and memcopy libs
-- Voku and Dorimanx Tweak for LowMemoryKiller
-- Disable Gentle Fair Sleep for better perfomance
-- Asynchronous I/O latency
-- Lower swappiness 600----->45 [it's 60]
What you did:
Compiled with Linaro 4.7.4
optimized LPDDR RAM timings (~10% throughput increase)
add Frandom v2
Added NEON compilation flag
mm: Dynamic management of dirty page writebacks
Battery Life eXtender (BLX)
add New Governor From Nexus Kernel
Added optimized CRC32 algorithm
So this kernel is very based on Slim kernel...and why didn't you state it and credit @mnazim properly?
And, removal of sources of the old Nexus kernel means you cannot share it in public anymore...didn't you learn anything since the Semaphore kernel incident?
Click to expand...
Click to collapse
this is based on cyanogen source....nexus is based on slim....not this

biagio7xD said:
this is based on cyanogen source....nexus is based on slim....not this
Click to expand...
Click to collapse
You said Nexus was based on CyanogenMod before this...your sources on GitHub is not supporting your speech...your old repo is gone too...

Real law means intellectual property right? Well then i have to state my opinions. One, this kernel is not earning any money or use ad like someone. Second, original creator of such code is also not earning money frm his creation so i can only see a moral intellectual property rights which is to demand this thread to credits those people properly. Third intellectual property rights is not so easy to be claim as you think. There are few principle such as improvising or this op did expanded his labour work n time to compile and the extent of his own work to actually produce this piece of kernel.
Sent from my GT-I9100G using xda app-developers app

cyelanius said:
Real law means intellectual property right? Well then i have to state my opinions. One, this kernel is not earning any money or use ad like someone. Second, original creator of such code is also not earning money frm his creation so i can only see a moral intellectual property rights which is to demand this thread to credits those people properly. Third intellectual property rights is not so easy to be claim as you think. There are few principle such as improvising or this op did expanded his labour work n time to compile and the extent of his own work to actually produce this piece of kernel.
Sent from my GT-I9100G using xda app-developers app
Click to expand...
Click to collapse
Nope...real law as in GPL...it is different from what you are saying...I did not say he broke it just yet (athough he did before), I was just saying he has to follow GPL, that's all...faux123 got penalised a few times for not adhering to GPL...
Even there is no such law (which is not the case), can't an innocent "user" like me know what's being modified and done to my device? :angel:

Ryuinferno said:
Nope...real law as in GPL...it is different from what you are saying...I did not say he broke it just yet (athough he did before), I was just saying he has to follow GPL, that's all...faux123 got penalised a few times for not adhering to GPL...
Even there is no such law (which is not the case), can't an innocent "user" like me know what's being modified and done to my device? :angel:
Click to expand...
Click to collapse
:thumbup::thumbup: on the GPL point but well xda is at own risk afterall for what you flash isnt it?
Sent from my GT-I9100G using xda app-developers app

cyelanius said:
:thumbup::thumbup: on the GPL point but well xda is at own risk afterall for what you flash isnt it?
Sent from my GT-I9100G using xda app-developers app
Click to expand...
Click to collapse
On your own risk, but what's the point of sharing if people do not know what's going on?

And what is the point with such restriction that will reduce development in xda which was meant to promote n encourage people to develop and introduce ways to put our device to better use? Op since last time in s plus thread got good reputation in his works and even now in s2. I hope this wil not stop people from giving their time n energy working on their works.
Sent from my GT-I9100G using xda app-developers app

cyelanius said:
And what is the point with such restriction that will reduce development in xda which was meant to promote n encourage people to develop and introduce ways to put our device to better use? Op since last time in s plus thread got good reputation in his works and even now in s2. I hope this wil not stop people from giving their time n energy working on their works.
Sent from my GT-I9100G using xda app-developers app
Click to expand...
Click to collapse
*facepalm*...I'm not gonna argue anymore...the fact is, things can be done, but get your path right... That's all I want to convey...

cyelanius said:
And what is the point with such restriction that will reduce development in xda which was meant to promote n encourage people to develop and introduce ways to put our device to better use? Op since last time in s plus thread got good reputation in his works and even now in s2. I hope this wil not stop people from giving their time n energy working on their works.
Sent from my GT-I9100G using xda app-developers app
Click to expand...
Click to collapse
I do not care if the op has a good reputation. The op isn't going to buy me a new device if I brick mine due to this kernel now would he? A kernel is far more important to how a device functions rather than a ROM. If there's little to no CORRECT information given, how would I know what I'm flashing is safe. You may tell me to just skip this and not flash this. But what about other people who sees the stuff in the op but isn't getting what they came for. Complying to GPL is a lot more than what you may think it is. It also helps learning and experienced people know exactly what they're getting.
Sent from my Nexus 7

I9300 (Int) 4.4 Custom Kernels for Omni Nighties

First things first: this is not a "best kernel thread", so please avoid asking or stating in that direction. This is all about spotting what custom kernels are compatible with Omni (for now) Nighties, and what has been your experience with them so far.
In my case, I've tried both Yank's (1.6g-beta2) and Boeffla (beta 1a) and they are working flawlessly. At the moment I'm using Boeffla just cause I find it easier to toy with different configurations and Yank's can only be tweaked from aroma. And in my personal experience, I've found that using ART instead of dalvik, it feels like I'm having a smoother experience (with a bit more battery drain, I must admit). But like I said, this is only how those kernels work on my device and that doesn't mean it will happen exactly the same way in your's and how you'd feel about them.
Any other kernels have been tested and are working in Omni 4.4 nighties?

fermasia said:
First things first: this is not a "best kernel thread", so please avoid asking or stating in that direction. This is all about spotting what custom kernels are compatible with Omni (for now) Nighties, and what has been your experience with them so far.
In my case, I've tried both Yank's (1.6g-beta2) and Boeffla (beta 1a) and they are working flawlessly. At the moment I'm using Boeffla just cause I find it easier to toy with different configurations and Yank's can only be tweaked from aroma. And in my personal experience, I've found that using ART instead of dalvik, it feels like I'm having a smoother experience (with a bit more battery drain, I must admit). But like I said, this is only how those kernels work on my device and that doesn't mean it will happen exactly the same way in your's and how you'd feel about them.
Any other kernels have been tested and are working in Omni 4.4 nighties?
Click to expand...
Click to collapse
I follow you, I would install dual boot with a custom stock 4.3 (Android Revolution HD) and an android 4.4 rom (OmniROM).
I could create dual boot between 4.1/4.2/4.3/4.4(only CM11) roms with excellent Googy-Max Kernel... but nothing to do with OmniROM 4.4 due to recovery or kernel incompatibility.
---------- Post added at 11:11 PM ---------- Previous post was at 11:03 PM ----------
Ah, I found Devil2 Kernel http://forum.xda-developers.com/showthread.php?t=2528624 that supports Omni 4.4... and dual booting

Never been a huge fan of dualbooting but since I'm planning to stay with Omni until we reach a estable build... I wouldn't mind giving stock 4.3 a spin.
Has anyone tried out Devil and Omni already?
Sent from my GT-I9300 using XDA Premium 4 mobile app

Just as an FYI with regards to custom kernels.
Just like with CM - Anyone can submit code to a kernel repository through gerrit for review. So if you're running something other than the official kernel, ask yourself:
1) What is the alternate kernel maintainer hiding? If their change is so good, why shouldn't it be the default, and why haven't they submitted their change to gerrit for review?
2) If they did submit a patch it was rejected - ask yourself why? Code review is public so you can find the reasons given for any patch being negatively reviewed (assuming the kernel maintainer even bothered to submit it).

Entropy512 said:
Just as an FYI with regards to custom kernels.
Just like with CM - Anyone can submit code to a kernel repository through gerrit for review. So if you're running something other than the official kernel, ask yourself:
1) What is the alternate kernel maintainer hiding? If their change is so good, why shouldn't it be the default, and why haven't they submitted their change to gerrit for review?
2) If they did submit a patch it was rejected - ask yourself why? Code review is public so you can find the reasons given for any patch being negatively reviewed (assuming the kernel maintainer even bothered to submit it).
Click to expand...
Click to collapse
I´m not much more than a resourceful noob, I must admit. But even when your line of thoughts seems very logical, I wonder if there could be any other reasons for kernel maintainers like Lord Boeffla or Yank555 to remain aside from main ROM projects. Both of them have been around for long now, and I've experienced with their work in my device a lot, with no harm at all (I can't guarantee results for other people, just talking about my experience).
I guess custom kernels give you the chance to adjust your kernel to your own needs. Some people would be ok about sacrificing some performance to obtain a better battery result, or would like to have the freedom to choose the governor that best suits you, for instance. Isn't customization and freedom the whole thing about custom ROMS and GPL?
As for me, I go for customs kernels seeking a better balance between performance and battery life. The way I use my phone, I won't last a full day with original kernels. But I must admit that if I had an app to tweak Omni's original kernel myself, I'd do it (I don't got the knowledge to dev my own kernel, thought).

Entropy512 said:
Just as an FYI with regards to custom kernels.
Just like with CM - Anyone can submit code to a kernel repository through gerrit for review. So if you're running something other than the official kernel, ask yourself:
1) What is the alternate kernel maintainer hiding? If their change is so good, why shouldn't it be the default, and why haven't they submitted their change to gerrit for review?
2) If they did submit a patch it was rejected - ask yourself why? Code review is public so you can find the reasons given for any patch being negatively reviewed (assuming the kernel maintainer even bothered to submit it).
Click to expand...
Click to collapse
Would you like to impelement CPU and GPU overclock?
Would you like to implement CPU and GPU undervolt?
Would you like to implement WIP governors like zzmoove?
I aksed only three questions from at least a few more, and I'm wondering if you could answer "yes" to at least one. Making custom kernel or custom rom is a great thing because we can push device to the limits, while doing so shouldn't be suggested neither implemented in any official kernel.

JustArchi said:
Would you like to impelement CPU and GPU overclock?
Would you like to implement CPU and GPU undervolt?
Would you like to implement WIP governors like zzmoove?
I aksed only three questions from at least a few more, and I'm wondering if you could answer "yes" to at least one. Making custom kernel or custom rom is a great thing because we can push device to the limits, while doing so shouldn't be suggested neither implemented in any official kernel.
Click to expand...
Click to collapse
Voltage control/overclock - The stability issues these cause isn't worth it. Especially undervolting - at best it saves 3-4% on CPU core power savings, which is only 1-2% of system power even when the cores are at full tilt.
Alternate governors - These will have to, of course, go through review as to whether they have a valid reason for existing. 90% of alternate governors out there achieve nothing that can't be achieved merely by tuning the defaults of existing ones. Many of them are 100% marketing bull**** - Lionheart being the most blatant example. (Lionheart is nothing more than conservative with the minimum polling rate limit properly fixed...)

Entropy512 said:
Voltage control/overclock - The stability issues these cause isn't worth it. Especially undervolting - at best it saves 3-4% on CPU core power savings, which is only 1-2% of system power even when the cores are at full tilt.
Alternate governors - These will have to, of course, go through review as to whether they have a valid reason for existing. 90% of alternate governors out there achieve nothing that can't be achieved merely by tuning the defaults of existing ones. Many of them are 100% marketing bull**** - Lionheart being the most blatant example. (Lionheart is nothing more than conservative with the minimum polling rate limit properly fixed...)
Click to expand...
Click to collapse
Of course you're right but if we're talking about gaming then in long-run uV can actually benefit users a bit more that 3-4%.
The most interesting thing is OC. It's always nice to have a bit more power than usual if you're using it only for a short time. This way you get real performance boost when it's needed, for example when CPU is doing a longer task. Nobody gets hurt if CPU will work at let's say 1.6 GHz instead of 1.4 for a few seconds.
I understand what you're trying to say and I totally agree with you. Most of the changes could be and should be contributed to the base but still most of the devs (including me) wouldn't contribute non-universal device-specific changes.

[Kernel] Vindicator [AOE3] [Strict-Aliasing/UberTC/Loop-Nest] [-R3-]

Vindicator Kernel
Intro:
This is the first Samsung kernel I've ever developed. I just started developing on android a few months ago so I'm brand new to this. I don't really consider myself a developer, I'm just good at figuring things out, all credit goes to the real hard workers coming up with original ideas like Franco, Imoseyon, Ktoonsez, etc. I hope you like it and I won't be offended if you hate it. All ideas are welcome and I'll try to help as best I can. If you wish to use this kernel in a rom , feel free! Although I would love it if you let me know just for my own curiosity.​
Goal:​Keep kernel as lean as possible while adding certain, desirable features, upstreaming source, and optimizing for maximum performance and battery life. The default settings should be optimal for a balance of battery and performance without a need to tweak.​
Disclaimer:​If your phone blows up, its not my fault. If your phone somehow gains perpetual battery life, I'll take full responsibility.​
Features:​-AOE3 Source Based
-ONLY TESTED ON TMOBILE VARIANT
-VControl Kernel Control Script(Loosely Based On lkconfig by imoseyon)
-Compiled With Uber ToolChain 4.9
-Compiled With Strict-Aliasing, A53/57 Tweaks, And Loop Nest Optimizations
-Various Other Build Optimizations
-WLAN_RX & WLAN_CTRL Wakelocks Reduced
-Optional GPU OC To 772Mhz
-A53 UnderClockable To 200Mhz, Overclockable To 1.6Ghz
-A57 UnderClockable To 200Mhz, OverClockable To 2.5Ghz
-Slub Allocator Updated To Linux 3.18
-GPU Slightly UnderVolted
-Small CPU Task Packing
-Nvidia Power Efficiency Patches
-GPU Voltage Control
-LED Fade Control
-Deep sleep fix
-Optimized Power Coefficients
-FSync Toggle
-Memory Leak Patches
-Sleep/Suspend Patches
-3X Faster Integer Square Root Operations
-Interactive Wont Scale Up Below 5% Load
-HRTimer Optimizations
-Avoid load Average Moire
-Scheduler Optimizations
-Memcopy and String Optimizations
-Use Compile Time Constants For Jiffies Conversions When Possible(Reduces CPU Overhead)
-Workqueue Updated To Linux 3.18
-Optional Power Efficient Workqueues (Expanded to more modules than normal as well)
-Moto QuickWakeup Driver
-Frandom (Efficient Random Number Generator)
-Cruft Free
-Slimmed Down
-Tons Of Debugging and Tracers Removed
-Many Misc. Optimizations
-FIOPS, CFQ, NOOP & BFQ IO Schedulers
-Tons Of IO Fixes/Performance Boosts
-Voltage Control
-Entropy Optimizations/Tweaks
-Optimized RWSEM
-Many Parts Pulled From Upstream
-Init.D Support
-Installs SU
-Potentially Something I Forgot...
Downloads:
[AFH] Kernel: https://www.androidfilehost.com/?w=files&flid=32684
Credits:
Imoseyon
Ktoonsez
AndreiLux
G.lewarne
Samsung
Flar2
anarkia
hybridmax
Franco
Faux123
Linaro
Code Aurora Foundation
Cl3Kener
Google
Motorola
If I Missed You Lemme Know!​
XDA:DevDB Information
Vindicator Kernel, Kernel for the T-Mobile Samsung Galaxy S6
Contributors
Xileforce
Source Code: https://github.com/XileForce/Vindicator-S6
Kernel Special Features:
Version Information
Status: Stable
Current Stable Version: R3
Created 2015-05-30
Last Updated 2015-05-30

Changelog:
R1:
-Initial Release
R2:
-CPU Task Packing For A53 Cores(Read In FAQ)
-Numerous Patches To HMP From Linaro(This controls how and when tasks are moved from big to LITTLE CPUs)
-CPUIdle Improvements
-GPS Fix(Thanks To Glewarne)
-Increase hash table size on futexs for better performance
-Tweaked Interactive Governor Values A Bit
-Additional Compile Time Optimizations
-Fix VControl Set On Boot
-Add Toggle For Task Packing To VControl(Currently Not Working :/ )
R3:
-Updated To Linux 3.10.67(From .61)
-Made binder mutex instead of rtmutex(less stutter)
-Reduce packing limit. (This will load the CPUs less before spreading load. Actually improves battery and performance)
-Some upstream cpuidle improvements
-Reduce preempt latency
-Optimize scheduler modules with maximum compiler optimizations
-Increase AES interleave to 4x(for encrypted people this showed a performance improvement of 37% on an A57 processor which is our Atlas cores)
-Use high priority workqueue for devfreq which is related to the GPU
-Scheduler considers cost of balancing and doesn't overestimate average anymore
-Small micro optimizations
-Improve Low Memory Killer routines patch from neobuddy89
-Add rbtree optimizations to LMK. (Significantly improves LMK performance)

FAQ's
1. What Is VControl And How Do I Use It?:
VControl is a kernel tweaking "app" built into the ramdisk. It is accessed via terminal emulator. To use it simply download terminal emulator, Type "su" without quotes and hit enter. Grant terminal emulator super user permission and then type "VControl" without quotes. Please note that it is case sensitive. This will bring up a text menu and you navigate by simply entering the option number you want and hitting enter. It is currently a WIP. FSync control and Power Efficient Workqueues don't currently set on boot properly More will be added to VControl over time.
2. You Broke My Phone! Fix It!:
It will be quite hard for me to figure out what is going on if you don't provide details and a log of what happened. To get a log use any file manager with root support and navigate to proc/last_kmsg.
3. What Kernel Tweaking App Should I Use?:
To Tweak the kernel, currently the only app that works well is KTweaker. To get KTweaker simply flash ktoonsez kernel which will install the app, then flash this. You will have the app unless you wipe data. You can also use VControl for a limited number of things
4. What Do Power Efficient Workqueues Do?:
Typically when a task is created, it is bound to the cpu it was created on. This is great for performance due to maintaining cache locality. However its not the best for battery life. Power efficient workqueues attempts to schedule tasks on already awake cpus. A quick note about awake and sleeping cpus. A sleeping cpu doesnt mean the cpu is turned off or not hotplugged, nor does this mean it has to be deep sleeping. A core can be online and still be sleeping which means it uses very little power. This is why we are able to run with all 8 cores on all the time. Most of the time the majority of the cpus are sleeping. With power efficient workqueues, jobs will attempt to schedule on awake cpus which means we dont have to wake up the cpus that are sleeping already. This means we lose cache locality so we suffer a small (very small) performance hit but we potentially save noticeable amounts of battery. Enable for battery, disable if you plan on benchmarking or just want the extra power. They are off by default.
5. What Is Strict Aliasing And Why Should I Care?:
One of the big things Linaro does with improving Android's performance is fixing violations of what's known as "the strict aliasing rule." A pointer is said to alias another pointer when they both refer to the same location of memory. This is OK and not an uncommon thing to do. The strict aliasing rule is that pointers of different types should never refer to the same location of memory (aka alias each other). Strict aliasing allows a compiler to make some assumptions when compiling and optimizing code that it otherwise couldn't. It's actually one of the biggest optimizations you can do and it is fairly hard to add which is why most people don't. Here's a nice read up on it http://cellperformance.beyond3d.com/articles/2006/06/understanding-strict-aliasing.html#benefits
6. What Does Moto QuickWakeup Do?:
QuickWakeup is a driver developed by motorola which increases battery life when the screen is off. It allows for only the modules which need to be woken up during sleep to wake up, thus conserving power.
7. What does Small CPU Task Packing Mean?:
Task packing Is a process done by the scheduler. What it tries to do is pack all tasks onto as few cores as possible. In this case, it only packs tasks for the low power(small) cores as we want to maintain maximum performance and spread the load out on the bigger cores. The advantage of packing tasks is that if we envision a scenario where there is a 40% load and that FETs spread out across the 4 small CPUs at 10% load each all 4 small CPUs are awake and running at a low frequency. With task packing all 40% would be shoved onto 1 core and the other 3 would sleep. This provides energy savings. However there becomes a point where the frequency required to process the extra tasks becomes high enough that it would use less power to run two cores at half the speed rather than 1 core at a very high speed. In this scenario the scheduler will automatically let tasks spill over onto adjacent CPUs in order to maximize power efficiency. One down side to this is a small performance hit due to having to pull the other cores out of a sleep state in the event of needing immediate power. However I would guess most people couldn't notice a difference.
More will be added as they come up and I think of them.

Annndd one more

Please note that this is only tested on the smg920T model. I have no idea if it will work on others

Works well thank you!

guaneet said:
Works well thank you!
Click to expand...
Click to collapse
Happy to hear that sir! You are the only other person to have tried it so far so that's good news. BTW. Idk what AVS# you are but I'm an 11 and I have stably undervolted by 75 for the A53 cores and 50 for the A57s. Maybe that will help someone.

Xileforce said:
Happy to hear that sir! You are the only other person to have tried it so far so that's good news. BTW. Idk what AVS# you are but I'm an 11 and I have stably undervolted by 75 for the A53 cores and 50 for the A57s. Maybe that will help someone.
Click to expand...
Click to collapse
I like your ideology behind this kernel. Totally reminds me of imosyen nice to see another dev here! I tend to keep my volts as is. I always have problems
Would also like to say that this kernel is best one for my phone so far! Haven't noticed anything but stability. Thanks

guaneet said:
I like your ideology behind this kernel. Totally reminds me of imosyen nice to see another dev here! I tend to keep my volts as is. I always have problems
Click to expand...
Click to collapse
Haha. My first kernel I ever made was for nexus 6 based on his lean kernel so that's what I learned on. In fact I believe I've got almost all the stuff removed that he has removed back on nexus 6 plus even more that I've found. I'm a firm believer that less is very often more. Expect more trimming down in the following builds.
Sent from my SM-G920T using XDA Free mobile app

Xileforce said:
Haha. My first kernel I ever made was for nexus 6 based on his lean kernel so that's what I learned on. In fact I believe I've got almost all the stuff removed that he has removed back on nexus 6 plus even more that I've found. I'm a firm believer that less is very often more. Expect more trimming down in the following builds.
Sent from my SM-G920T using XDA Free mobile app
Click to expand...
Click to collapse
Well now I know I'll be in good hands ?
I've been following him since the note 3 days.
Can't wait to see where you take this project!

guaneet said:
Well now I know I'll be in good hands ?
I've been following him since the note 3 days.
Can't wait to see where to take this protect!
Click to expand...
Click to collapse
Well I hope you like it! I'll leave you alone now lol.
Sent from my SM-G920T using XDA Free mobile app

Downloading and testing now. Not really related, but I was wondering if power save overrides cpu clock/gov/anything else? The reason I ask, is that I typically underclock or undervolt a bit to save some battery but keep in normal high performance state with PS off. But if PS mode does this anyway, I shouldn't bother should I?
Wow, that 3rd post (FAQs) was super-informative!
Sent from my SM-G920T using XDA Free mobile app
Edit: Everything seems to be running perfectly fine so far. Charging up so I can see how the battery life is. I'm currently on OZOP's ROM.

welcome to the lonely S6 kernel devs club!
ive sent you a pm

benjmiester said:
Downloading and testing now. Not really related, but I was wondering if power save overrides cpu clock/gov/anything else? The reason I ask, is that I typically underclock or undervolt a bit to save some battery but keep in normal high performance state with PS off. But if PS mode does this anyway, I shouldn't bother should I?
Wow, that 3rd post (FAQs) was super-informative!
Sent from my SM-G920T using XDA Free mobile app
Edit: Everything seems to be running perfectly fine so far. Charging up so I can see how the battery life is. I'm currently on OZOP's ROM.
Click to expand...
Click to collapse
To be honest I dont know specifically what power save mode does. I've only had the phone for like 6 days. Ik ps limits frame rate and stuff like that so You will save additional power beyond Running in normal mode, but I'm not sure how it affects cores and frequencies currently. This phone is a whole different beast than my nexus 6...I would like to add more, lower frequency steps to save more power. Glad to hear you are liking it so far though! And I plan on adding a configuration guide of sorts in my 4th post with more detail on the controls available . I also plan on getting synapse working properly as that will enable significantly easier kernel tweaking.
Sent from my SM-G920T using XDA Free mobile app

g.lewarne said:
welcome to the lonely S6 kernel devs club!
ive sent you a pm
Click to expand...
Click to collapse
Thank you! Your build script and zip file were a massive help to me. This is completely different from building for a nexus. And I'm not too worried yet the device is little over a month old. I'm hopeful we will see more development start to come. I tried to use your new script setup but I couldn't get it booting for some reason. I'll have to revisit it. I'm sure I didn't something wrong. I'm a Noob.
Sent from my SM-G920T using XDA Free mobile app

Oh em gee! Im switching to tmo!!!!!! Lol j/k...Glad to see you here Xile! If anyone here likes bleeding edge performance and stability, Vindicator is the kernel for you!!!

Custodian said:
Oh em gee! Im switching to tmo!!!!!! Lol j/k...Glad to see you here Xile! If anyone here likes bleeding edge performance and stability, Vindicator is the kernel for you!!!
Click to expand...
Click to collapse
Lol once I wrap my head around all these different variants I'll try and get a build working for the edge. Took me days just to get call audio working
Sent from my SM-G920T using XDA Free mobile app

Xileforce said:
Lol once I wrap my head around all these different variants I'll try and get a build working for the edge. Took me days just to get call audio working
Sent from my SM-G920T using XDA Free mobile app
Click to expand...
Click to collapse
I know there's a unified method that ktoonz and a few others use/developed, reach out to him for assistance! However, just out of curiosity I flashed this baby on top of sac and ram deodexed rom for the s6 edge sprint variation and it works lol....only issue ive noticed so far is the "storage message" keeps popping up.

Custodian said:
I know there's a unified method that ktoonz and a few others use/developed, reach out to him for assistance! However, just out of curiosity I flashed this baby on top of sac and ram deodexed rom for the s6 edge sprint variation and it works lol....only issue ive noticed so far is the "storage message" keeps popping up.
Click to expand...
Click to collapse
O_O that's suprising...ik there's some differences...hmm...well then. And I am going to try and get their method working soon. I tried to get it working earlier and got it compiling correctly but couldn't get it to boot sadly I will revisit it.
Sent from my SM-G920T using XDA Free mobile app

Going to give your kernel a run today and see how it goes! Always love to see more options around here.
I'm curious about the memory leak(s) you say you've patched. I wasn't really aware of any on the kernel side, can you give a bit of explanation here?