Related
Hi!
This is my first kernel for Nexus S. It is based on _thalamus 3.0.29_r1 kernel.
Matt's kernel don't have OC, LiveOC or similar features.
Original changelog/features from _thalamus:
*Second stable release based on AOSP 3.0.8 source, merged up to 3.0.29 from mainline kernel.org
*PM Fast for wifi
*Deadline I/O scheduler tweaked for flash for lowest I/O latencies
*Ondemand governor tweaked to sample less at higher frequencies to reduce overhead
*Config: CFS CPU scheduler, Deadline I/O scheduler, Ondemand CPU governor, SLUB allocator, Tiny Preempt RCU, Tun/Tap, IP advanced routing and CIFS built in
*Various compiler warnings and section mismatches fixed
*Disabled a stack of debug options resulting in a slightly faster kernel.
*Built with latest Linaro toolchain (GCC 4.6-2012.03)
Click to expand...
Click to collapse
* Changelog/features for v0.8 (29/04/2012):
Stock Display Freq
Added BLD & BLN
block: limit default readahead size for small devices
BLD Enabled (3000ms)
BLN Enabled (I need test)
KSM tweaked
Added modified liblights for BLN
Changelog/features for v0.6 (29/04/2012):
Optimized compiler flags
Increased HZ to 1000
Added optimized CRC32 algorithm
Increased default readahead value to 1024KB
Changelog/features for v0.5 (29/04/2012):
Gzip
Enabled KSM for page merging
Disabled "Debug Filesystem"
Disabled "Kernel debugging"
Patched from 3.0.29 to 3.0.30
65FPs panel
Little tweak for Deadline
Added ARM RWSEM algorithm
Added SLQB memory allocator
Build with 4.6-2012.04 Linaro
I have high performance but I don't know yet about battery drain.
Remember to wipe Dalvik cache.
DOWNLOAD: http://dl.dropbox.com/u/20742004/MattsKernelv0.8.zip
SOURCES: https://bitbucket.org/MattScherbatsky/matt-nexuss_kernel
Benchmarks: zirka 2300 Quadrant Score - 3075 Antutu
Thanks to: _thalamus, franciscofranco, rcrdbrt, Ezekeel, Google, my parents.
This is for me.
Welcome to Nexus forum
Sent from my Nexus S using xda premium
Wellcome to Nexus S world, good luck with your kernel!
Updated to v0.6.
Changelog in first post.
If you can add voodoo sound, fast charge, Bld. This will be perfect for me.
Sent from my Nexus S
Welcome to nexus s man . Great to have another kernel dev
Is it possible to add the high bigmem in next release? (404MB RAM)
Sent from my Nexus S using Tapatalk 2
mixtapes08 said:
If you can add voodoo sound, fast charge, Bld. This will be perfect for me.
Sent from my Nexus S
Click to expand...
Click to collapse
Same here
Welcome Matt & hoping for improvements . Good luck .
Sent from my Crespo
Hey Matt, sup. I know you since the development in Optimus One, hope you gonna do great in Nexus.
Sent from my Nexus S using Tapatalk 2
Welcome to the nexus s community
Sent from my Nexus S using Tapatalk 2
Hoping to see bln
Sent from my Nexus S using xda premium
only thing i'm missing is bln, i'd really love to give this a try if you could incorporate this
I'll implement soon BLN, I need it.
For another things, I don't know. For now I need tests for this kernel.
Hope!
MattScherbatsky said:
I'll implement soon BLN, I need it.
For another things, I don't know. For now I need tests for this kernel.
Hope!
Click to expand...
Click to collapse
Tested, its fine everything working. Except! I cant underclock to 800mhz...
Sent from my Nexus S using Tapatalk 2
DaXmax said:
Tested, its fine everything working. Except! I cant underclock to 800mhz...
Sent from my Nexus S using Tapatalk 2
Click to expand...
Click to collapse
Wait for future releases!
Increased Hz to 1000? What is it?
Sent from my Nexus S using XDA
Welcome to the Nexus S community.
I`ve used your kernel for a couple of hours now. Like the other folks, I miss also BLN, and also Voodoo sound/colors.
Besides, the only downside till now has was the laggy 3D performance in games. I thought that it was only the game, but tested 3d in Quadrant Advance, and it was still the case.
Overall, your Kernel is snappy and stable-no random reboots till now. I`ll be waiting impatiently for your next release
v0.8 released.
Changelog in first post.
darkchaos7 said:
Welcome to the Nexus S community.
I`ve used your kernel for a couple of hours now. Like the other folks, I miss also BLN, and also Voodoo sound/colors.
Besides, the only downside till now has was the laggy 3D performance in games. I thought that it was only the game, but tested 3d in Quadrant Advance, and it was still the case.
Overall, your Kernel is snappy and stable-no random reboots till now. I`ll be waiting impatiently for your next release
Click to expand...
Click to collapse
using root explorer, go to /system/lib/egl/ and delete the libGLES_android.so file. reboot, then try quadrant again. any difference?
simms22 said:
using root explorer, go to /system/lib/egl/ and delete the libGLES_android.so file. reboot, then try quadrant again. any difference?
Click to expand...
Click to collapse
I am actually on ICS, not GB.
This kernel is built for and from codeworkx's CM9 with hardware acceleration.
Again this is for codeworkx's cm9.
One more time, this is for codeworkx's cyanogenmod 9.
Finally, this is for cm9 and cm9 only!
Don't let the "alpha" title fool you into thinking it's not stable... it is... It's only alpha because it's still in the "testing" phase.
Works on the P31XX. I don't know about other models.
FEATURES/CHANGES
Overclockable to 1.38GHz
Underclockable to 150MHz
GPU overclocked to 384MHz (highest possible without bottlenecking and battery waste)
Tuned OnDemand Governor (recommended)
Interactive Governor with time view concept (learns from your use and adjust parameters accordingly)
Wheatley Governor (optimized use of C4 state)
Hotplug Governor (turns of one of the cpus to save power when load is very low)
Pegasusq Governor (A better version of Hotplug, but experimental and not designed for our devices)
SIO scheduler for speed/performance with flash storage
Linaro Toolchain
Linaro optimizations and patches
Linaro makefile compiler flags
Makefile Optimizations to boost performance
Optimized CRC32
Optimized ARM RWSEM algorithm
Tiler Memory increased to 32MB (linaro)
Other Tiler patches/tweaks to improve 2D gpu performance
SLQB allocator
lib/string use glibc version for speed
lib/memcopy use glibc version for speed
lib/sha1: use the git implementation of SHA-1 and removed ARM sha1 routines
I/O backport from 3.2 linux kernel for less dirty throttling
Tweak CPUIDLE to enter deep sleep faster thus saving battery
LowMemoryKiller backported from 3.4 kernel
Disable GENTLE_FAIR_SLEEPERS for improved performance
Allow Deep Sleep to occur when media decoder is active allowing the cpu to shut off when playing music or video.
Allow MPU logic to power down when in deep sleep to save power.
Increased default readahead value to 1024KB for faster sdcard reads
A lot of stuff I probably forgot I did.
Also includes a ridiculous amount of bugfixes, patches, and tweaks that can found on my github commit log.
***Read Me***
DO NOT, UNDER ANY CIRCUMSTANCE, USE SETCPU PROFILES OF ANY KIND OR ANYTHING SIMILAR WITH THIS DEVICE. Things like screen off max frequency, temperature throttling, and others are already built in to the OMAP architecture. Using ANY profiles will ONLY lead to problems, instabilities, and battery drain.
***************
Let me know how it "feels" especially by default without overclocking or underclocking. Also I'd be interested in how the battery treats you. Thanks!
DOWNLOAD
a1 - http://goo.im/devs/Metallice/Tab2/MetalliKernel_CM9_a1.zip
a2 - http://goo.im/devs/Metallice/Tab2/MetalliKernel_CM9_a2.zip - Changes to: Wifi, GPU, Low Memory Killer, others
a3 - http://goo.im/devs/Metallice/Tab2/MetalliKernel_CM9_a3.zip - Changes to: ondemand, interactive, add wheatley, add pegasusq, deep idle, compiler flags, SDcard read, others
a4 - http://goo.im/devs/Metallice/Tab2/MetalliKernel_CM9_a4.zip - Changes to: Compiler flags, CPUidle, RCU, others.
a5 - http://goo.im/devs/Metallice/Tab2/MetalliKernel_CM9_a5.zip - reverts and bugfixes, use 10.1 ramdisk in 10.1 version
a6 - http://goo.im/devs/Metallice/Tab2/MetalliKernel_CM9_a6.zip - CLASSIFIED change, NEED feedback on speed, slowness, smoothness, and lag compared to stock and a5. Reverted some revets. Reverted changes to RCU to fix SODs. Graphics. For 10.1 fixed overclocking.
If you would like to overclock but your device can't handle it... and are willing to suffer through extreme battery drain to try this - Link
Thanks go out to - CodeworkX, Ezekeel, Francisco Franco, Imoseyon, Faux123, Fugumod, Sonicxml, Texas Instruments, Samsung, Google, and all the others I probably forgot about. Sorry!
Source
https://github.com/Metallice
______________________________________________________________________________________________________________________
So, I'm a broke college student without a job this summer - instead taking extra classes and studying like crazy for the MCAT. If you like my work and want to thank me, feel free to buy a coffee or an energy drink so I can continue to stay up late studying and working on Android stuff (or help me pay for med school ). I don't want to have to choose! Thanks and I appreciate your support!
I like this format, just like Trinity kernel. Much less placebo for added things. Like, "zOmg, linaro made my phone fly! Thanks for including it, it rocks!!!!!" With the things being changed kept hidden, I'd LOVE for you to release a couple builds, with different compile dates, without changing a thing, but telling people you did. Watch the placebo roll in, ha
Thank god you released this. Seriously the only thing this ROM was missing was a little OC. You guys are really making this purchase (Getting the GTab 2) a really good idea!
Sent from my GT-P3113 using Tapatalk 2
Will this kernel work on p3100 becaise all i see around here all the custom roms and kernels are for p3113 so please just confirm the models on whicj it would be flashable.
Thanx
Sent from my GT-P3100 using Tapatalk 2
tapan15in said:
Will this kernel work on p3100 becaise all i see around here all the custom roms and kernels are for p3113 so please just confirm the models on whicj it would be flashable.
Thanx
Sent from my GT-P3100 using Tapatalk 2
Click to expand...
Click to collapse
I don't know. I have no way of testing.
Sent from my GT-P3113 using xda premium
Oh yeah. You da man!
Just flashed... will give updates on how it is running.. so far so good!!
Sent from my GT-P3113 using xda app-developers app
cannt wait for this, thank u so much
I will tell my feeling about this after I flashed it
Sent from my GT-P3110 using Tapatalk 2
Runs smooth and stable at 1380mhz on my 3113. Thank you for your hard work! On a related note, what governor do you recommend? In your experience, does using hotplug actually save battery?
yes it is working with 3100. great job.
Flashed this on my P3110, and it bricked it NOT.
Thanks for this. Waiting to see what more you will add to the kernel.
Cheers.
I've got a question. Does the tab 2 not have a CPU temp Senser?
Sent from my GT-P3113 using xda premium
Been using this since last night and so far no RR or SOD's. Everything is running smooth without any errors, glitches, etc. Great work, as always, Metallice!
---------- Post added at 11:37 AM ---------- Previous post was at 11:37 AM ----------
Kidney Poker said:
I've got a question. Does the tab 2 not have a CPU temp Senser?
Sent from my GT-P3113 using xda premium
Click to expand...
Click to collapse
I don't know, but wander why you would ask?
miketoasty said:
Been using this since last night and so far no RR or SOD's. Everything is running smooth without any errors, glitches, etc. Great work, as always, Metallice!
---------- Post added at 11:37 AM ---------- Previous post was at 11:37 AM ----------
I don't know, but wander why you would ask?
Click to expand...
Click to collapse
CPU controllers have a temp read out. I like having one. I know most apps pull the info from kernel. Was trying to find out if we have the hardware built in and the kernel isn't passing info or if the hardware is just not there.
Sent from my GT-P3113 using xda premium
Maybe it's just my tablet, but I can't get any clock speed greater than stock to work. Tried it on this kernel and one I did myself.
Kidney Poker said:
I've got a question. Does the tab 2 not have a CPU temp Senser?
Sent from my GT-P3113 using xda premium
Click to expand...
Click to collapse
Yeah the omap does. Not sure if I turned it on or not. If its on you can get the temp from dmesg I believe.
Sent from my Galaxy Nexus using xda premium
imnuts said:
Maybe it's just my tablet, but I can't get any clock speed greater than stock to work. Tried it on this kernel and one I did myself.
Click to expand...
Click to collapse
I can try adding a 1.1ghz slot?
Sent from my Galaxy Nexus using xda premium
Kidney Poker said:
CPU controllers have a temp read out. I like having one. I know most apps pull the info from kernel. Was trying to find out if we have the hardware built in and the kernel isn't passing info or if the hardware is just not there.
Sent from my GT-P3113 using xda premium
Click to expand...
Click to collapse
Most apps actually pull their info from the battery temp and are unreliable.
Sent from my Galaxy Nexus using xda premium
nphil said:
Runs smooth and stable at 1380mhz on my 3113. Thank you for your hard work! On a related note, what governor do you recommend? In your experience, does using hotplug actually save battery?
Click to expand...
Click to collapse
OnDemand. I don't bother with hot plug as I don't see any benefit. If you do a lot of non interactive activities like reading or music it may be helpful, but I would try the 200mhz min before I try hot plug.
Sent from my Galaxy Nexus using xda premium
Have you flashed on cm9 or stock rom??
Sent from my GT-P3100 using Tapatalk 2
[GUIDE] *19/8 UPDATE!* 4.3 RESULTS! N7 'Best' Kernel (Performance & Battery Tests)
A lot of people have been posting in these forums, "Which kernel is the best"?
I was intrigued and decided to work this out for myself and share it with the rest of the community.
This is not meant to hurt the makers of these kernels.
But FIRST, it is not a fact that there is a 'best' kernel. 'Best' is an opinion and everyone has a different opinion.
On the Nexus 7, I personally believe there is no 'bad' kernel and none of these kernels I have tested perform badly or lag heaps.
People still like to push their N7's as hard as they can though.
The Main 2 Features people look for in a kernel are:
Performance & Battery Life
Some people prefer performance over Battery, or vice versa, or an all-round kernel.
NEXUS 7
8GB
Paranoid Android AOSPAL 3.97
KERNELS
Kernel[022]JB4.3(1.60Ghz Quad)UV,CIFS+UTF-8,PowerHAL,Linux-3.x Hybrid]
by faux123
Ultimate Edition 22
[KERNEL][GPL][N7] franco.Kernel - r75
by franciscofranco
r75
[KERNEL] [ANDROID 4.3] LeanKernel: Minimalistic Kernel [V2.0, 8/2/13]
by azoller1
Version 2
[Kernel][4.2.2] M-Kernel - a61[WiFi/3G]
by Metallice
a61
[KERNEL] [3.1.10] [JB 4.3/4.2.2] [Purified Kernel r130] [Linaro] [07-08-2013]
by eng.stk
r130
[KERNEL] [GROUPER/TILAPIA] [4.2.2/4.3] OXYDO [3.2.49] [Linaro]
by sgt. meow
v7
[KERNEL][Grouper][JB 4.3] 3.1.12-XPerience-04 [12/08/13]
by TeamMex
v4
[KERNEL] Bricked-Kernel Grouper/Tilapia
by show-p1984
4.3 Stock Kernel
I will test another kernel if someone requests me too. I will not post other's results because they most likely have a different setup and therefore different results making it inaccurate.
NEW! 4.3 RESULTS
19/8 UPDATE
PERFORMANCE
Performance was tested with Antutu. Tested three times and averaged.
1.3 RESULTS
Kernels set to 1.3GHz max and used the default governor for the test.
1. M-Kernel: 13489
2. Franco: 13440
3. LeanKernel: 13298
4. Bricked: 13194
5. Oxydo: 13179
6. Xperience: 13139
7. Faux: 13122
8. Purified: 13102
9. Stock: 12903
MAX RESULTS
Kernels set to highest clockspeed, performance governor.
1. Bricked (1.6): 15339
2. Oxydo (1.6): 14938
3. M-Kernel (1.7): 14463
4. LeanKernel (1.5): 14343
5. Faux (1.7): 14155
6. Xperience (1.6):14147
7. Franco (1.3): 13440
8. Purified (1.3): 13102
9, Stock (1.3): 12903
BATTERY LIFE
Default Governor, 100% Brightness
Watching HD 720p Video on a Loop (MX Player Software Decoding) in a 10% range. The reason for this is not to kill my battery lifetime.
Max Freq at stock speed (1.3GHz)
1. M-Kernel:7 hours 5 minutes
2. Oxydo: 7 hours
3. LeanKernel: 6 hours 20 minutes
4. Franco: 5 hours 50 minutes
5. Purified: 5 hours 35 minutes
6. Xperience: 5 hours 30 minutes
7. Faux: 5 hours 10 minutes
8. Bricked: 5 hours 5 minutes
9. Stock: 5 hours
Max Freq set as high as it can go.
1. LeanKernel (1.5): 6 hours
2. Franco (1.3): 5 hours 50 minutes
3. Purified (1.3): 5 hours 35 minutes
4. M-Kernel (1.7):5 hours 25 minutes
5. Oxydo (1.6): 5 hours 10 minutes
6. Stock (1.3): 5 hours
7. Bricked (1.6): 4 hours 55 minutes
8. Xperience (1.6): 4 hours 45 minutes
9. Faux (1.6): 4 hours 40 minutes
OVERALL
Best Performance:
[KERNEL] Bricked-Kernel Grouper/Tilapia
by show-p1984
[Kernel][4.2.2] M-Kernel - a61[WiFi/3G]
by Metallice
a61
Best Battery Life:
[KERNEL] [ANDROID 4.3] LeanKernel: Minimalistic Kernel [V2.0, 8/2/13]
by azoller1
Version 2
Best Overall:
[Kernel][4.2.2] M-Kernel - a61[WiFi/3G]
by Metallice
a61
[URL="http://forum.xda-developers.com/showthread.php?t=1950858"][KERNEL] [ANDROID 4.3] LeanKernel: Minimalistic Kernel [V2.0, 8/2/13]
by azoller1
Version 2
[/URL]
4.3 Results
Nice! This should be a sticky to help reduce unnecessary threads about kernels. My only suggestion would be to run Quadrant 3 times and take the average number. Only because 3 runs will have different results every time
Sent from my Atrix 4.1.1
Best battery life can't really be determined since everyone uses their tablet differently i wouldn't even waste my time.
Sent From My Toro+ via White Tapatalk
I am testing several kernels here. The fastest for me so far is trinity, but the battery is draining really fast. I then tried the 1400 MHz trinity and things got a little better with a good performance in quadrant.
I am testing motley now and seems good but a little slower than trinity. Home screen doesn't look as buttery as in trinity. I will keep testing a little longer.
Sent from my Nexus 7 using xda premium
---------- Post added at 06:19 PM ---------- Previous post was at 05:54 PM ----------
Crap...i just had a reboot while using motley and started playing dead trigger. I returned to trinity 111 and it is definitely smoother than the others for me. Maybe it is because it is the first one i flashed and now when i try others, even wiping, something remains and the other kernels don't work right. Who knows....
Sent from my Nexus 7 using xda premium
ÜBER™ said:
Best battery life can't really be determined since everyone uses their tablet differently i wouldn't even waste my time.
Sent From My Toro+ via White Tapatalk
Click to expand...
Click to collapse
I was going to do a HD video playback test. See how long it lasts on each.
Sent from my Nexus 7 using xda premium
Can't wait for the battery life results!
Sent from my Nexus 7 using xda premium
I posted about the so called battery life screen shot thread in there thread.
the only way to really test battery life is to start with ful ly charged battery each day. 1 TEST PER DAY or charge cycle.
you will have to get a mathematical equation figured out first .it would have to be something like
cpu cycles x screen brightness x memory usage and s on divided by then another equation to determined battery pull watts im guessing
you would have to keep the first exact on each test device to determine the battery life ..
ok while typing this i have decided its way over my head and quite honestly not sure it can be done ..
HELP My brain now hurts ..
The best way to test CPU performance would be to make sure that all kernels are at the same exact core frequency max, and all cores enabled, on performance governor. Performance governor alone doesn't enable all cores, and benchmarks sometimes don't trigger all 4 cores as on either (it can, but not always, and that can lead to totally different numbers on different kernels).
Some kernels have higher max frequencies then others, some use the new CPU quiet driver and some don't (although this wouldn't really matter if all cores are forced on for testing).
Some kernels have higher GPU clocks then others, some have higher refresh rates then others, some have both higher refresh rates and clocks, etc.
Personally, the best way to judge a kernel is by what features are implemented in it for example, is it linaro compiled? Does it use the old hotplug method for CPU cores, or the new quiet driver? Updated WiFi driver? fsync toggle? Was it built with RNDIS drivers?
Quadrant isn't a stable means of benchmarking (run it 3 times in a row and watch those numbers change), but generally speaking, if the same kernel is compiled with linaro, vs no linaro, the linaro build would be faster (don't quote me on this). Not all those kernels are linaro that I'm aware of.
Also, Jay's buttery smooth tweaks, his fsync doesn't work on franco.Kernel. That kernel specifically uses a different command to toggle fsync, and unless that method was used, any I/O results would be void (basically comparing fsync kernels to non-fsync, where fsync would obviously dominate). And there is even the possibility the ROM you were using may interfere with CPU clocks and voltages unknowingly.
So with all that said, trying to find the "best" kernel, is going to take a bit more effort then just benching with Quadrant
And for my own personal opinion (no hard-testing done) I like franco.Kernel Stock clocks and voltage, fsync toggle, new CPU quiet driver and WiFi driver, works awesomely faux's kernel would be my second preferred kernel, only because lack of fsync control, and getting rid of the CPU quiet driver.
espionage724 said:
The best way to test CPU performance would be to make sure that all kernels are at the same exact core frequency max, and all cores enabled, on performance governor. Performance governor alone doesn't enable all cores, and benchmarks sometimes don't trigger all 4 cores as on either (it can, but not always, and that can lead to totally different numbers on different kernels).
Some kernels have higher max frequencies then others, some use the new CPU quiet driver and some don't (although this wouldn't really matter if all cores are forced on for testing).
Some kernels have higher GPU clocks then others, some have higher refresh rates then others, some have both higher refresh rated,
Personally, the best way to judge a kernel is by what features are implemented in it for example, is it linaro compiled? Does it use the old hotplug method for CPU cores, or the new quiet driver? Updated WiFi driver? fsync toggle? Was it built with RNDIS drivers?
Quadrant isn't a stable means of benchmarking (run it 3 times in a row and watch those numbers change), but generally speaking, if the same kernel is compiled with linaro, vs no linaro, the linaro build would be faster (don't quote me on this). Not all those kernels are linaro that I'm aware of.
Also, Jay's buttery smooth tweaks, his fsync doesn't work on franco.Kernel. That kernel specifically uses a different command to toggle fsync, and unless that method was used, any I/O results would be void (basically comparing fsync kernels to non-fsync, where fsync would obviously dominate). And there is even the possibility the ROM you were using may interfere with CPU clocks and voltages unknowingly.
So with all that said, trying to find the "best" kernel, is going to take a bit more effort then just benching with Quadrant
And for my own personal opinion (no hard-testing done) I like franco.Kernel Stock clocks and voltage, fsync toggle, new CPU quiet driver and WiFi driver, works awesomely faux's kernel would be my second preferred kernel, only because lack of fsync control, and getting rid of the CPU quiet driver.
Click to expand...
Click to collapse
Yeah. I know. I'm going to add a other features section. If someone else could do it that would be good. If not, I'll do it later.
Sent from my Nexus 7 using xda premium
bcvictory said:
I was going to do a HD video playback test. See how long it lasts on each.
Sent from my Nexus 7 using xda premium
Click to expand...
Click to collapse
Who JUST watches video... would be as effective as leaving it on a white screen for hours. Like I said there is no point and would just give false hope.
Sent From My Toro+ via White Tapatalk
Used Motleys 520 at first, probably still my favorite. Used trinity and both of clemsyns latest. Clemsyns elite 1.8 got too hot for my liking, I'm using the 1.5 UV right now and batteries not too great, but its very smooth. Trinity is smooth as well but Motleys 520 seemed to both perform better and give better battery life at comparable clock speeds to the rest.
So Motleys will be my daily, but I refuse to stop flashing!
this is very helpfully
thanx
Hey Guys! I have been running some battery tests over the last week. I'm finishing the results now. Show you guys soon!
Battery Tests are in the OP! Check them out, give your own opinion!
bcvictory said:
Battery Tests are in the OP! Check them out, give your own opinion!
Click to expand...
Click to collapse
Interesting... not what I expected.
They varied so much did you use all the same governor? Stock is interactive. All the same I/O? That is quite a difference a shocking one at that.
Sent From My N7 via White Tapatalk
ÜBER™ said:
Interesting... not what I expected.
They varied so much did you use all the same governor? Stock is interactive. All the same I/O? That is quite a difference a shocking one at that.
Sent From My N7 via White Tapatalk
Click to expand...
Click to collapse
Yeah. I used interactive on all of them. Set the minimum at the lowest it could go and the max as high as it could go on each kernel.
bcvictory said:
Yeah. I used interactive on all of them. Set the minimum at the lowest it could go and the max as high as it could go on each kernel.
Click to expand...
Click to collapse
-_- that's not fair.... nobody uses the absolute high... redo... at stock frequency.
You can't tell which is best if they are not under the same circumstances. Obviously Franco is top because his only goes to 1.3 which is stock. While others are over clocked... thats unreliable.
Sent From My Toro+ via White Tapatalk
ÜBER™ said:
-_- that's not fair.... nobody uses the absolute high... redo... at stock frequency.
Sent From My Toro+ via White Tapatalk
Click to expand...
Click to collapse
All of them at 1.3GHz Max? Whats the point of overclocking then?
bcvictory said:
All of them at 1.3GHz Max? Whats the point of overclocking then?
Click to expand...
Click to collapse
For benchmarks... not battery performance.
So your telling me the benchmarks where at different frequencies too? Wow have you never done a science expierenment? You don't test under different variables... that's inconsistent.
That's like frying an egg at 100F and comparing it to frying it at 120F. Obviously the 120F will fry faster. Cmon man.
Sent From My N7 via White Tapatalk
ÜBER™ said:
For benchmarks... not battery performance.
Sent From My N7 via White Tapatalk
Click to expand...
Click to collapse
Yeah. I guess... I was going to try Motley at 1.3 for fun anyway. See how I do...
This is the BOJ1 version of NN-kernel. It's going to go in a little different direction than the previous kernel. And I'm not very good with Git. So may I present to you.....
NN-J-Kernel
This will work on the T710 and 810 Wifi only models. I built this for my BNN Nook, but it works with other 710's and 810's.
You flash this at your own risk. There's the discaimer. :silly:
Install instructions:
You MUST HAVE ROOT and TWRP installed to flash this!!
It's built on Lollipop 5.1.1 - I have no idea what will happen if you flash this on anything else!!
REMOVE SYNAPSE "set on boot" BEFORE INSTALLING!!!
Flash kernel then wipe cache and dalvic
Reboot and profit!!
Changelog
V1.3
GPU Control
UV now possible as well as select frequency
Samsung has the Max listed as 600Mhz. I'm pretty sure its 700 as thats what I've been using this at. Just an FYI
V1.2
Charge Control for battery - It's the best I can do for a "Faster Charge". Be careful. This can shorten the life of your battery. Or worse, ruin it if you change from default. You have been warned.
Maybe some other minor stuff
Also see/read this post
_______________
V1.1
Hotplug Big cores - They are the only ones that can be
OC/UC Big and Little - 2.1Ghz-500Mhz, 1.6Ghz-200Mhz
UV all CPU's
650Mhz step added to GPU
FIOPS Scheduler
Booting freq to 1.4Ghz
Lots of Optimizations
Gentle Fair Sleepers
Synapse stuff added
HMP controls - Big little cores
Memory controls
_______________
V1
OJ1 source
Synapse tuner - it's the only one that seems to work. Finally, after much effort.
Selinux Permissive
KNOX removed
Init.d support
Enabled OTG
Updated Frandom
Govs enabled- Powersave, Conservative, Ondemand - removed Userspace
Some optimizations
_____________________________
This is a Tablet, so I won't be adding things like I would for a phone.
Use "Synapse" from the Playstore for your Kernel tuner
______________________________
Downloads:
NNJ-Kernel-V1.3-710
NNJ-Kernel-V1.3-810
NNJ-Kernel-V1.2-710
NNJ-Kernel-V1.2-810
NNJ-Kernel-V1.1-710
NNJ-Kernel-V1.1-810
NNJ-Kernel-V1-710
NNJ-Kernel-V1-810
Thanks to:
ashyx
BigBot96
UpInTheAir
djmax81
And other peoples gits I can't remember
XDA:DevDB Information
NN-J-Kernel-LP-T710/810, Kernel for the Samsung Galaxy Tab S2
Contributors
engine95
Source Code: https://github.com/engine95/Exynos5433-BOJ1-gts28-210wifi.git
Kernel Special Features:
Version Information
Status: Stable
Stable Release Date: 2016-02-24
Created 2016-02-24
Last Updated 2016-02-24
And yes, Schedulers are fixed in this.
Flashed, running perfectly so far....
Thank you for your work.
Gesendet von meinem SM-T810 mit Tapatalk
I flashed this kernel (oj1-based) and the NN kernel (oh7-based) and let antutu benchmark the performance of both kernels (both default settings). This kernel wins outstanding!
Is this the "newer" one yes ?
Awesome! Thanks, it's nice to see some development on thus tablet. Otherwise there is nothing to flash
Sent from my LG-D855 using Tapatalk
halloarno said:
I flashed this kernel (oj1-based) and the NN kernel (oh7-based) and let antutu benchmark the performance of both kernels (both default settings). This kernel wins outstanding!
Click to expand...
Click to collapse
I did the same and the OJ Kernel is slower for me. Did 2 tests. 69-70k in old Kernel while i getting 62-65 with OJ1. Any idea ? How much do you get?
I tried setting max core to 2.x ghz but its not working, is it fixed to 1.9 ghz=no oc enabled so far?
Very strange, worked before i think. Just have Synapse installed and also wiped caches after flashing.
Any idea?
IngoPan said:
I tried setting max core to 2.x ghz but its not working, is it fixed to 1.9 ghz=no oc enabled so far?
Very strange, worked before i think. Just have Synapse installed and also wiped caches after flashing.
Any idea?
Click to expand...
Click to collapse
This kernel has no OC function, at least until now.
Actually, right now, it has no OC or extras. This is just the start. It's in the changelog.
Antutu scores, for me are 66-67000.
OC/UC and UV will be coming shortly. It's done. I just can't find out why Big 1500 ghz and Little 1300ghz are favored. Messed with a lot of stuff to no avail. Yet
V1.1 is up!!
SEE/READ changelog as there is a ton of stuff in it now. Both Kernel and Synapse. Yea for Hotplug!!!
Google "HMP" so you can understand it better. Big/Little core control. Little cores are ALWAYS online.
Think I found a way to get us a Fast Charge. Gotta finish and test it. So far it's looking good.
engine95 said:
Think I found a way to get us a Fast Charge. Gotta finish and test it. So far it's looking good.
Click to expand...
Click to collapse
Incredible. Fast charge on our tab s2? I hope so?
Great job! Thanks! Flashed and works perfectly, new synapse settings seems cool. UC seems to work fine.
Shiuldnwebuse the DVFS app as mentioned in synapse?
Sent from my SM-T810 using Tapatalk
whynottoday said:
Incredible. Fast charge on our tab s2? I hope so
Click to expand...
Click to collapse
Gugus38 said:
Great job! Thanks! Flashed and works perfectly, new synapse settings seems cool. UC seems to work fine.
Shiuldnwebuse the DVFS app as mentioned in synapse?
Sent from my SM-T810 using Tapatalk
Click to expand...
Click to collapse
Thanks. I have no idea if that app helps as I'm new to Samsung. If you try it, make a backup first.
Everything works fine without it though.
Can you say Fast Charge?!!!!!
I had to port Hotplug and Fast charge, but I think we're good to go. I still need to put finishing touches and build for both 710/810. But it should be out tomorrow IF my Tab doesn't meltdown while charging tonight. :silly:
Edit: Also, remember that as with everything computer, Fast Charge will be limited by Thermals. Just like CPU/GPU.
engine95 said:
Thanks. I have no idea if that app helps as I'm new to Samsung. If you try it, make a backup first.
Everything works fine without it though.
Can you say Fast Charge?!!!!!
I had to port Hotplug and Fast charge, but I think we're good to go. I still need to put finishing touches and build for both 710/810. But it should be out tomorrow IF my Tab doesn't meltdown while charging tonight. :silly:
Edit: Also, remember that as with everything computer, Fast Charge will be limited by Thermals. Just like CPU/GPU.
Click to expand...
Click to collapse
It wont really be fast charge as our tab doesn't support it
Reuben_skelz92 said:
It wont really be fast charge as our tab doesn't support it
Click to expand...
Click to collapse
Interesting. Guess we'll find out. I have been able to vary the voltages while charging. Up and down. But as I said, thermals will probably limit it.
Edit. Let's call it "Charge Control" then.
engine95 said:
Interesting. Guess we'll find out. I have been able to vary the voltages while charging. Up and down. But as I said, thermals will probably limit it.
Edit. Let's call it "Charge Control" then.
Click to expand...
Click to collapse
Sorry me being an idiot. Will be cool to see our tabs charge a little.bit quicker
Thank you very much for the kernel!
Running 1.1 with
A57 cores at 1300/500 and UV 50mV
A53 cores at 1000/200 and UV 25mV
2 A57 cores online
FIOPS scheduler, 2048 readahead
Gentle Fair Sleepers enabled
It's running snappy! Will watch battery life
P. S.: Why doesn't Synapse have live monitoring where one can see which frequency each core has at the moment? Just like other kernel tuners?
hasenbein196601 said:
Thank you very much for the kernel!
Running 1.1 with
A57 cores at 1300/500 and UV 50mV
A53 cores at 1000/200 and UV 25mV
2 A57 cores online
FIOPS scheduler, 2048 readahead
Gentle Fair Sleepers enabled
It's running snappy! Will watch battery life
P. S.: Why doesn't Synapse have live monitoring where one can see which frequency each core has at the moment? Just like other kernel tuners?
Click to expand...
Click to collapse
They are there. On the core screen at the top. Just wiggle the screen and you'll see it change. Big cores don't move as much until there is demand for them.
V1.2 is in the OP! With "Charge Control" See/read changelog for warnings.
This was a big port, and everything seems to be working. Just watch your tab for temps just to be sure it can handle the voltages. We should be ok, as it's built in and not on it's own. Meaning that all of Samsungs limits are still there. So Thermals should hold it from melting down. My Tab has been plugged in all night and is just fine.
This has all kinds of cable configurations, so it should cover about everything out there. I only use AC and USB but left all the others in for all of you.
One more time, we should be OK using Charge Control, but BE WARNED that it may ruin your battery and shorten battery life span. You use this at your own risk.
This is a home of White Wolf Kernel for OneUI and Los/AOSP based roms.
Would like to thank to all devs that work has been used to create this kernel.
Al Credits belong to devs listed below
@farovitus - ELS
@AndreiLux
@Lord Boeffla
@franciscofranco
@ Despair Factor
@ak
@flar2
@arter97
@TheCrazyLex
@Sultanxda
@eng.stk
@osm0sis
@neobuddy89
@Tkkg1994
@frap129
@jcadduono
@yank-lu.555
@dorimanx
@acuicultor
@Alucard24
@ZeroInfinity
@morogoku
@nathanchance
@mostafaz
@joshuous
@faux123
@djb77
@Eamo5
@remilia15
@jesec
@synt4x93
@DEVILOPS 007
@abtekk
@osm0sis - AnyKernel
and all others devs that make our lives easier.
Special thanx for @urgali , @Sungfive and @Duncan1982 .
@Cybertron™ for thread and signature images.
@Saber for his superb thread with all detailed kernel stuff.
Short Feature list:
Always based on latest Linux Stable
Governors:
Schedutil, ShedutiliX
Custom Frequencies:
Gpu - 598MHz
Additional drivers:
- Power Efficient Workqueues
- apk launch and touch boost - from Sultanxda
- Wake Gestures from flar2
Flashing instructions
1. Flash in recovery no black magic
WARNING
Always flash on stock kernel.
Compatibility list:
Kernels have been tested with ROMs below:
- official Ultimate Pie ROM all credits to @rifek4
- official TeamExyKings LightROM all credits to @yash92duster.
- LOS Kernel have been tested with unofficial LOS 17
Telegram Support Group
https://t.me/joinchat/DNkhGFMc-o5FP9qSwz-4og
Downloads:
https://mega.nz/folder/IkdUVQ5D#9mgIsnTS6GnBEAHZSx7QAA
TWRP With f2fs support
https://www.androidfilehost.com/?w=files&flid=294117
Kernel Source
https://github.com/yarpiin/White-Wolf-Uni-LOS
https://github.com/yarpiin/White-Wolf-Uni-OneUI
Twrp Source
https://github.com/yarpiin/android_device_samsung
https://github.com/yarpiin/android_kernel_samsung_universal9810
If You like my work - hit thanx button or buy me a beer
https://www.paypal.me/yarpiin
XDA:DevDB Information
White Wolf Kernel, Kernel for the Samsung Galaxy S9
Contributors
yarpiin
Source Code: https://github.com/yarpiin/White-Wolf-SGS9-TW
Kernel Special Features:
Version Information
Status: Testing
Created 2018-11-13
Last Updated 2020-11-11
Changelog:
OneUI Based kernels
https://github.com/yarpiin/White-Wolf-Uni-OneUI/blob/master/changelog.txt
LOS 17
https://github.com/yarpiin/White-Wolf-Uni-LOS/blob/master/changelog.txt
Set of rules for flashaholics by @flar2 .
Great guide how to avoid problems when flashing .
https://elementalx.org/ten-rules-for-flashaholics/
Please read and follow.
TWRP Usage Guide
Make sure your kernel support f2fs file system
1.Flash patched TWRP by yarpiin
2.MAKE BACKUP( Problems with restoring backups are solved now, but just in case)
3.Go to Wipe > Advanced wipe select Data and Change File System to F2FS, you can do the same to Cache ( Not recommended )
4.Restore backup of your data
5.Reboot & enjoy
Troubleshoot
It might happen that package extract error will occure when flashing the rom (or any updater-script based zip) . In this case just restart instalator and all should be fine.
Follow these steps or you will loose your data.
reserved
Glad to see you here!
ExtentedByte said:
Glad to see you here!
Click to expand...
Click to collapse
Had massive problems with rhread opening . Xda became so slow and tricky.
Sent from my Galaxy S9+ using XDA Labs
yarpiin said:
Had massive problems with rhread opening . Xda became so slow and tricky.
Sent from my Galaxy S9+ using XDA Labs
Click to expand...
Click to collapse
Been there lol. Great to see more development for the device. Welcome and congratulations on release
rifek4 said:
Wow, great to see new kernel! Keep it up, gonna follow you and try it!
Clean duplicaded threads in S9+ Review and S9+ ROMs&kernels section. Just keep one thread here so in S9 / S9+ Cross section where are all kernels and ROMs
Yeah, it's since years like that... You must be used to that Now you will be know Always after click "send thread" go to check section where you opening thread because your page seems to be stuck while adding but in real the thread is opened there and added correctly...
P.S.
Polacy przejmują S9 sekcje xD
Click to expand...
Click to collapse
Duplicated threads are @Xda Fault . Didnt had a clue that they have been created as I had bad gateway and xda server couldnt phrase request messages qhen created . It has been reported to xda .
I will get them removed
Btw im not newbie at xda
Sent from my Galaxy S9+ using XDA Labs
wow didnt had idea that so many multiple threads has been created have requested removal
Sent from my Galaxy S9+ using XDA Labs
Deleted
Great performance!
rifek4 said:
I didn't tell that you're newbie I'm just letting you know that adding threads working like that since 1 year or something... I released S7 ROM then and was same like is now, nothing changed since then
Click to expand...
Click to collapse
no offence m8 :laugh: never had any problems with any of my 4 threads
Sent from my Galaxy S9+ using XDA Labs
Thanks! Using it right now, keep it up.
rifek4 said:
So you're really lucky bcz 99% users has it... And you also had it if you made XXX dupliaced themes
Click to expand...
Click to collapse
Looks like last of my threads been started 2 years ago , but never faced any issue with it.
Ohh well **** happens
Sent from my Galaxy S9+ using XDA Labs
Better battery life with this kernel. Standby drain is much better, even with gaming with pubg with high fps
Hi there, just had an opportunity to delve through the source. Very impressive work and well done, without a doubt this will influence future releases of Endurance! Would also like to quickly address something I noted, as my kernel was heavily influenced by Andrei's work.
Your commit utilising Andrei's EAS values are outdated and following the BRE5 source drop have since been updated in the exynos9810.dtsi.
Have a good day, and thanks for giving me some more stuff to play with for future releases!
Eamo5 said:
Hi there, just had an opportunity to delve through the source. Very impressive work and well done, without a doubt this will influence future releases of Endurance! Would also like to quickly address something I noted, as my kernel was heavily influenced by Andrei's work.
Your commit utilising Andrei's EAS values are outdated and following the BRE5 source drop have since been updated in the exynos9810.dtsi.
Have a good day, and thanks for giving me some more stuff to play with for future releases!
Click to expand...
Click to collapse
Thank you for that . Will revert it then. Yea shame that Andrei stop working on it . Have read his article. This is my first exynos device so it was a nice point of start.
One thing that i noticed withs flars OC that frequencies are going to max value 2.7MHz just at ceratin ocxasions and i think that not all cores at the same time . I decided to use that commit to find out how it does work. When disable Exynos hotplug and OC big xluster to anything above 2324MHz device crashes under the load . So to make it bir snappier i jave kept Andreis max frequencies but also raise min as 850Mhz is the highest one with same voltage as the stock lowest.
Sent from my Galaxy S9+ using XDA Labs
yarpiin said:
Thank you for that . Will revert it then. Yea shame that Andrei stop working on it . Have read his article. This is my first exynos device so it was a nice point of start.
One thing that i noticed withs flars OC that frequencies are going to max value 2.7MHz just at ceratin ocxasions and i think that not all cores at the same time . I decided to use that commit to find out how it does work. When disable Exynos hotplug and OC big xluster to anything above 2324MHz device crashes under the load . So to make it bir snappier i jave kept Andreis max frequencies but also raise min as 850Mhz is the highest one with same voltage as the stock lowest.
Click to expand...
Click to collapse
I agree, truly great work accomplished by Andrei. it is a shame, albeit his merging in of the 16ms PELT was IMO the best thing to happen to this device yet.
Yep that's the nature of hotplugging. I eventually decided to maintain hotplugging and the max quad freq of 1794MHz to conserve battery. I've found hotplugging just to give insanely good Geekbench scores and not much more, likely due to the very very rare usage of those frequencies in realistic workloads. No noticeable regression in performance using 1794MHz quad in regards to user experience when using 16ms PELT.
I use 2002MHz on my OC build as I find it the best balance of performance and battery for the device. 2314 is too much of a battery regression with quad frequencies for me to justify the performance boost.
Eamo5 said:
I agree, truly great work accomplished by Andrei. it is a shame, albeit his merging in of the 16ms PELT was IMO the best thing to happen to this device yet.
Yep that's the nature of hotplugging. I eventually decided to maintain hotplugging and the max quad freq of 1794MHz to conserve battery. I've found hotplugging just to give insanely good Geekbench scores and not much more, likely due to the very very rare usage of those frequencies in realistic workloads. No noticeable regression in performance using 1794MHz quad in regards to user experience when using 16ms PELT.
I use 2002MHz on my OC build as I find it the best balance of performance and battery for the device. 2314 is too much of a battery regression with quad frequencies for me to justify the performance boost.
Click to expand...
Click to collapse
I will agree but after weeks of testing I havent found 2314MHz has a massive impact in the battery
Nature of hotplug in exynos is totaly different than snapdragon and one that I jsed to . Trus hotplugging is actualy offlining not needed cpus in my meaning.
Sent from my Galaxy S9+ using XDA Labs