[KERNEL][AOSP] mamarley's BETA Droid Incredible ICS HWA kernel - Droid Incredible Android Development

This is a kernel for designed for recent builds of drk.hd's and invisiblek's CM9 builds. It is based on the "stock" Evervolv kernel used in those ROMs, but it has more features like USB Fast Charge, HAVS, and support for the 128mHz CPU speed. It is also compiled with GCC 4.7 instead of 4.4.
This kernel is only compatible with drk.hd's a9 build, invisiblek's 5/29/2012 nightly, or later. This is because of incompatibilities between the versions of the Adreno drivers and libraries used in the ROM and kernel.
Lite kernels: The lite kernel is similar to the regular kernel, except it has "unnecessary" features such as zRam, swap, KSM and most of the CPU governors and IO schedulers removed. If you use smartassv2 and SIO and you don't use zRam or KSM, these kernels are for you.
All the releases here are classified as beta, so there may still be problems to work out. If you run into anything, please let me know.
All of the releases may be downloaded here: https://github.com/mamarley/android_kernel_htc_qsd8k/downloads
Changelog:
8/12/2012: Regular Lite
Compile with "-O3 -fno-tree-vectorization" for more speed
Quiet down the light sensor driver so it doesn't spew quite so much spam into the kmsg log
7/24/2012: Regular Lite
Pull in a few more optimization flags from the Evervolv kernel
Enable the TCPMSS iptables target because it appears to be used during tethering
7/7/2012: Regular Lite
Fix the first-touch lag issue that occurred when the phone was woken up on battery.
Package the "zram" executable with the Regular version of the kernel
7/6/2012: Regular Lite
Make USB Fast Charge setting changes take effect immediately
7/4/2012: Regular Lite
Sync changes from Evervolv kernel
Update configuration to hopefully fix the last few charger lockups
Re-enable IPv6 netfilter so that IPv6 traffic will show up in the statistics
6/29/2012: Regular Lite
Update to Linaro 12.06
Sync changes from Evervolv kernel
Add "lite" kernel with removed zRam, swap, and all governors/IO schedulers except smartassv2 and sio
6/20/2012: Regular
Allow compiling with GCC 4.7 (thanks to drewis and michaelh (http://seabright.co.nz/2012/06/11/kernel-not-booting-with-linaro-gcc/))
Compile with Linaro GCC 4.7
6/6/2012: Regular
Revert power-related changes in the OJ driver, hopefully fixing the decrease in battery life from the last version
Update configuration to hopefully fix WiFi tethering
Reduce WiFi voltage, should increase battery life slightly
6/1/2012: Regular
Update AMOLED/LCD panel driver to latest version (improves panel resume time)
Update BMA150 accelerometer driver
Update Atmel touchscreen driver
Re-enable 10-point multitouch (Should work at the kernel level, but does not work at the ROM level currently)
Update Optical Joystick driver (This one is for all you Replica Island players out there )
5/31/2012: Regular
Add the Intellidemand and Lazy CPU governors (Thanks tiny4579 for getting them to compile)
5/30/2012: Regular
Initial release based on Evervolv qsd8k ICS kernel (https://github.com/Evervolv/android_kernel_htc_qsd8k)
Add USB Fast Charge support (thanks tiny4579 and chad0989)
Add HAVS support
Add support for 128mHz CPU clock operation
Add lagfree, interactiveX, and smartass CPU schedulers. Set smartassV2 as default
Add SIO IO scheduler and set it as default
Make a few minor configuration changes to reduce binary size
Kernel Source:
https://github.com/mamarley/android_kernel_htc_qsd8k

marley, not to be rude.. but there really is no difference between this and the incredikernel hwa beta 2 kernel right?

Gorilla* said:
marley, not to be rude.. but there really is no difference between this and the incredikernel hwa beta 2 kernel right?
Click to expand...
Click to collapse
they're completely different kernel sources. This is meant to be a more stock AOSP kernel with extra features mamarley. If you were fine with the stock kernel from the most recent CM9 builds, this is similar to that with a few extra useful features.
mamarley, you should add lazy governor. It's just a thought.

tiny4579 said:
they're completely different kernel sources. This is meant to be a more stock AOSP kernel with extra features mamarley. If you were fine with the stock kernel from the most recent CM9 builds, this is similar to that with a few extra useful features.
mamarley, you should add lazy governor. It's just a thought.
Click to expand...
Click to collapse
I tried to add lazy, but I kept getting all sorts of compile errors. Here is an example, but there are many more along the same lines:
drivers/cpufreq/cpufreq_lazy.c:538:6: error: 'sampling_rate_max_old' undeclared here (not in a function)
I applied exactly the same patch that you applied to Incredikernel, so I am not sure what is going on there.

mamarley said:
I tried to add lazy, but I kept getting all sorts of compile errors. Here is an example, but there are many more along the same lines:
drivers/cpufreq/cpufreq_lazy.c:538:6: error: 'sampling_rate_max_old' undeclared here (not in a function)
I applied exactly the same patch that you applied to Incredikernel, so I am not sure what is going on there.
Click to expand...
Click to collapse
Try adding the following to include/linux/cpufreq.h:
#define cpufreq_freq_attr_ro_old(_name) \
static struct freq_attr _name##_old = \
__ATTR(_name, 0444, show_##_name##_old, NULL)
#define cpufreq_freq_attr_rw_old(_name) \
static struct freq_attr _name##_old = \
__ATTR(_name, 0644, show_##_name##_old, store_##_name##_old)
Or modify the code in cpufreq_lazy.c to work with your existing cpufreq.h.
First option is easier I think. I think those defines are not standard. Not sure how I ended up adding them or in which commit. Adding them manually would be easier here.

Thanks tiny4579, that did it.
There is a new release out that has the intellidemand and lazy governors added.

New release out with some updated drivers

Awesome ill give it a shot! Thanks!
Sent from my ADR6300 using XDA

mamarley said:
New release out with some updated drivers
Click to expand...
Click to collapse
I suggest you release the multitouch as an optional download because it will not play nice with some phones.

nfiniti9 said:
I suggest you release the multitouch as an optional download because it will not play nice with some phones.
Click to expand...
Click to collapse
Are you having problems? There is some logic in the touchscreen driver that should automatically switch to 2- or 3-point multitouch on phones that don't support 10.

mamarley said:
Are you having problems? There is some logic in the touchscreen driver that should automatically switch to 2- or 3-point multitouch on phones that don't support 10.
Click to expand...
Click to collapse
This could be the same problem I had a while back when I implemented multitouch on gingertiny-v2. Since the multitouch selection automatically reverts to 2 or 3 point there really isn't a need to set that value to 10 in the config_T9 array. This is assuming you're using the multitouch selection patch. I know for a fact setting that value to 10 broke it for some devices for me. Try reverting that and seeing.

Been using the 6/1 version for a few days now. Battery life isn't nearly as good as 5/31.

thekendog said:
Been using the 6/1 version for a few days now. Battery life isn't nearly as good as 5/31.
Click to expand...
Click to collapse
I thought I had noticed that too. Let me do some testing to figure out which driver update caused the issue.

I just posted another update. The new version should fix the battery life issue and the WiFi tethering issue.

Marley, lol, I was having lock ups and freezes with incredikernel, so i flashed your 6/1 kernel, then just as I booted into the OS and realized your kernel works! You posted an update lol..
At the same time i decided to upgrade my Radio to 8.08
Now im gonna flash your 6/06 kernel
---------- Post added at 01:21 PM ---------- Previous post was at 01:18 PM ----------
The big thing I already notice about your kernel is it works!
Its responsive.. When i ask the phone to reboot it does.
With incredikernel it would hang there, at the boot menu then bootloop then come back into the OS before it executed a command.. and
Unplugging from a power source would cause a freeze..
So far so good, Im gonna check wireless tether now.
---------- Post added at 01:31 PM ---------- Previous post was at 01:21 PM ----------
so far so good with the 6/06 kernel as far as smoothness but..
WIFI TETHER IS STILL BROKEN WITH 6/06

Gorilla* said:
Marley, lol, I was having lock ups and freezes with incredikernel, so i flashed your 6/1 kernel, then just as I booted into the OS and realized your kernel works! You posted an update lol..
At the same time i decided to upgrade my Radio to 8.08
Now im gonna flash your 6/06 kernel
---------- Post added at 01:21 PM ---------- Previous post was at 01:18 PM ----------
The big thing I already notice about your kernel is it works!
Its responsive.. When i ask the phone to reboot it does.
With incredikernel it would hang there, at the boot menu then bootloop then come back into the OS before it executed a command.. and
Unplugging from a power source would cause a freeze..
So far so good, Im gonna check wireless tether now.
Click to expand...
Click to collapse
I never had any of those issues with incredikernel. I'm on marleys now and its running well. I did have one reboot on ik but I wasn't using the phone so I have no clue why it happened. Have not had any lag or reboots since using marley's. I have never had a reboot unplugging usb. Although I rarely use usb, I almost always plug it in a ac. I assume people having issues are actually plugging it into a computer usb port. And I'd be willing to bet most of those people are using laptops with a bit of software that monitors the usb for media devices.

Sorry, but I am out of ideas for the WiFi tethering. I have all of my network configuration options set the same as Incredikernel, so I don't know what is going on.

nfiniti9 said:
I never had any of those issues with incredikernel. I'm on marleys now and its running well. I did have one reboot on ik but I wasn't using the phone so I have no clue why it happened. Have not had any lag or reboots since using marley's. I have never had a reboot unplugging usb. Although I rarely use usb, I almost always plug it in a ac. I assume people having issues are actually plugging it into a computer usb port. And I'd be willing to bet most of those people are using laptops with a bit of software that monitors the usb for media devices.
Click to expand...
Click to collapse
I don't have an issue with USB and with fast charge on in incredicontrol with incredikernel. No reboots or freezes. I wonder if it doesn't like some phones. Doesn't make it a bad kernel.
Sent from my Galaxy Nexus using Tapatalk 2

mamarley said:
Sorry, but I am out of ideas for the WiFi tethering. I have all of my network configuration options set the same as Incredikernel, so I don't know what is going on.
Click to expand...
Click to collapse
WiFi hot spot works for me. As long as I remember to turn on mobile data. Don't use tether so I can't test that.
Sent from my GT-P7510 using Tapatalk 2

pmsrefugee said:
WiFi hot spot works for me. As long as I remember to turn on mobile data. Don't use tether so I can't test that.
Sent from my GT-P7510 using Tapatalk 2
Click to expand...
Click to collapse
Good. I don't know what was going on with Gorilla. I don't have a data plan (Page Plus for the win!), so I can't test either myself.

Related

[KERNEL][AOSP][2.6.38.2] - Savaged-Zen v1.0.0 [HAVS+CFS/BFS]!

ALL SOURCE CAN BE FOUND HERE
Concerns about XDA rules:
We will never charge for our forum. Our forum is read only to anyone without registering... you must register to post. Getting downloads of the advertised kernels will be listed on both XDA and forum.savage-zen.org. Savaged-zen's download site will always be publicly accessible, and nothing we have will ever be closed source or have any fees. This meets XDA's rules.
Our team consists of 5 guys. (in order of Name(real name))
NyghtGT (Anthony) Website, Graphics
Bbedward (Brandon) Kernel
Jsseidel1 (Josh) ROM/Kernel
Luisrcastillo (Luis) Kernel
mwielgosz (Mike) ROM/nonSBC Builds
Credits
Cyanogen and team --> various code bits
toastcfh --> various commits from .35 version of CM-Kernel
#teamwin --> wimax drivers in kernel and HDMI changes in kernel for full HDMI mirror (using the HDMwIn app)
erasmus --> smartass gov
intersectRaven --> HAVS
ms79723 --> SBC Mod
Zen-Kernel team --> various modifications
If we forgot anyone or you think you should be added to credits please PM us
(credits are given to the dev's where we got some code from, may no be original designer)
Please visit http://mirror.savaged-zen.com for the download.
you can give the progress of the builds @ http://buildbot.savaged-zen.org/builders one there all green you will see them on the mirrors.
Changelog is listed on the homepage: http://www.savaged-zen.org/?page_id=57
We request that you look for updates on our forum. We will post links to download there, as well as updates to the changelog.
Keep note : We will still update OP title, and answer questions on XDA, but as we all have admin rights on the forum we dont have to depend on one person to keep updated, and it will also make it easier for us to create a WIKI
our forum is located at
http://forum.savaged-zen.org
Check out our homepage at http://www.savaged-zen.org
If you like our kernel and want to support us... donate (supports the homepage payments and then beer!!)
nonSBC Versions can be found here
SBC WARNING:The Savaged-Zen team takes no responsibly for any issues to your phone or battery, including but not limited to, battery derogation or failure. SBC has been rumored to have caused issues with battery's and claims have been made that it actually causes physical damage to your phone. It is the users responsibility to review any opinions or warnings about such claims, and by flashing kernels that include the SBC mod, the user is taking the responsibility for any issues or failures to their phone. Please research this Modification, review the code, and come to your own decision before using such modification.
reservedaaaaaaaaaa
Sweeeet! Bringing the inc out of the dark ages. We'll finally be on par with the nexus one and supersonic
Sent from my ADR6300 using XDA App
I'd love to test. Just PMed you
Just pm'd you. Im so excited!
Sent from my DROOOOOID Incredible using XDA App
running GB cm7... I will test it out!
Just responded to all the PMs, once a couple people report back i'll just make it available to the public
All i heard so far was that it booted, the first build i forgot the atmel touch screen driver, but the test2 build (the one i just sent out in the masses) includes it, so hopefully most everything works.
*EDIT*
Everyone who PM'd got a build, i'm going to wait to hear back before sending out anymore so please don't PM, if the results are good i'm going to update the OP with both test builds.
Didnt work for me. It splash looped for me.
Sent from my full_inc using XDA App
flashing now. will report back
did not work for me either, I got stuck at the incredible screen, pulled battery it got past that, the screen went black and then I noticed green/red/blue bars running vertical down my screen
dang, i was too late
I got it to boot but no touch capabilities
Updated the OP
The builds there fix the touch screen issue and probably the wifi and BT issues, although as a tendency in the newest bcm4329 drivers wifi sleep policy should probably be set to never
The people who can't boot I suspect cache/dalvik arent being wiped properly (depending on recovery, usually clockworkmod is the one with the issues). Other than that all I can think of is there's some difference between incredible and incrediblec (but i don't think it necessarily matters?)
Anyway, as I don't have an incredible i'm relying on the testers
bbedward said:
Updated the OP
The builds there fix the touch screen issue and probably the wifi and BT issues, although as a tendency in the newest bcm4329 drivers wifi sleep policy should probably be set to never
The people who can't boot I suspect cache/dalvik arent being wiped properly (depending on recovery, usually clockworkmod is the one with the issues). Other than that all I can think of is there's some difference between incredible and incrediblec (but i don't think it necessarily matters?)
Anyway, as I don't have an incredible i'm relying on the testers
Click to expand...
Click to collapse
I got a newer incredible with SLCD screen
Touchscreen works now, way to go! Now for things that don't work:
- Wifi attempts to turn on, but fails
- Bluetooth also attempts to turn on, but fails
- LED notification light doesn't seem to function for alerts, charging, or anything
Other things that work:
- GPS locks on in about 5 seconds (sitting inside of a house)
- 3G data
- Voice calls
- SMS/MMS
- Still camera
- Camcorder
- Battery charge counter
neur0tk said:
I got a newer incredible with SLCD screen
Click to expand...
Click to collapse
Could be just a missing driver then, should be easily rectified, i suspect the same with LEDs, and wifi is strange unless for some reason the module isn't probing, maybe the module goes in a different place for the incredible who knows
bbedward said:
Could be just a missing driver then, should be easily rectified, i suspect the same with LEDs, and wifi is strange unless for some reason the module isn't probing, maybe the module goes in a different place for the incredible who knows
Click to expand...
Click to collapse
Want me to modprobe for the wifi? I'll need the module name.
jonesh said:
Want me to modprobe for the wifi? I'll need the module name.
Click to expand...
Click to collapse
it's bcm4329, it's built and in there, unless something funky in the makefile is changed it should work fine.
bbedward said:
it's bcm4329, it's built and in there, unless something funky in the makefile is changed it should work fine.
Click to expand...
Click to collapse
$ su
# modprobe bcm4329
modprobe: chdir(2.6.37-Savaged-Zen-INC-0.0.1-alpha+): No such file or directory
#
Something else peculiar... I'm not sure if the phone's charging. I have it plugged into a wall outlet, and while the LED is not lit as reported before, my charge percentage is dropping. That said, the battery icon is displaying the 'animating' charge. Thoughts?

[Kernel][Request] - HTC Stock Sense kernel with GPU+ 5pt multi touch?

Would any dev or anyone who knows a bit about compiling be able to do this? I've tried some other kernels, and generally love some of the features and performance they bring, but none of them has been 100% stable for me. Stock kernel is 100% stable, but lacks the 5 point multitouch which I do use for emulators, and GPU+ which makes a noticeable difference in 3D graphics smoothness.
The closest I found was HeyItsLou's #8 (I might have the number wrong) which runs great, has the multitouch, but no GPU+ Lou's has some other good features that seem to help smoothness and battery life too.
Maybe Lou or someone could throw GPU+ into that kernel? Or just take the stock kernel and add GPU+ and multitouch? Would anyone else be interested in this?
Um what about the. 32 incredikernel?
Sent from my ADR6300 using XDA Premium App
That's what I'm running now but it will randomly peg the CPU (system panel shows Android System as maxed out) and slow the phone to a crawl, requiring a reboot. And no, I'm not running SetCPU. That is the ONLY problem I have with that kernel, but I've never had that happen on the stock kernel. I don't want to deal with the random lockups anymore.
Ziggy's kernel doesn't have the lock ups, but doesn't run as smooth either, and I notice a lot of stutters while gaming.
I don't know why that would be. I compiled it for someone a few days ago and he loves how it runs.
Sent from my ADR6300 using XDA Premium App
what is 5pt multi touch?????
jdog94 said:
what is 5pt multi touch?????
Click to expand...
Click to collapse
Just that, you are able to use five fingers (or "points") simultaneously.
From my understanding, multitouch is affected by the touch panel you have, and also the way it is flashed. There was a compatible recovery floating around a while ago, but it's no longer maintained.
Yeah not all phones support 5pt
Sent from my ADR6300 using XDA Premium App
Phydo - I've actually been running your undervolted version of Chad's kernel for the last few days. I'm still having the same issue. In Chad's thread, there were actually a few people having the exact same issue. It was thought that SetCPU was somehow to blame, but again, I don't even have that installed.
From a performance standpoint, Chad's kernel is the one I've had the best results with and keep going back to, but this issue with the CPU getting stuck at 100% load happens regularly, every 24-48 hours.
bast525 said:
Phydo - I've actually been running your undervolted version of Chad's kernel for the last few days. I'm still having the same issue. In Chad's thread, there were actually a few people having the exact same issue. It was thought that SetCPU was somehow to blame, but again, I don't even have that installed.
From a performance standpoint, Chad's kernel is the one I've had the best results with and keep going back to, but this issue with the CPU getting stuck at 100% load happens regularly, every 24-48 hours.
Click to expand...
Click to collapse
I'd be willing to bet your problem is the ProcessStats.java bug documented in Issue 9733. It is exacerbated by kernels that add extra frequencies. Some ROMs work around the issue by making the time_in_state file unreadable, but SetCPU undoes this workaround every time you open it. 24-48 hours seems about right for the time it was taking for me to run into this problem. I'm now running with an init.d script that redoes the chmod 600 every 5 minutes just in case I open SetCPU and the problem has gone away.
Aweaver33 - would this be something I could easily incorporate into my own ROM? I'm running a mostly stock ROM that I customized by adding an earlier Evo framework and taking out some stuff I didnt want. Would I be able to flash your init.d script? Or since I don't run setcpu, could you explain or point me somewhere I could figure out how to make time_in_state unreadable?
EDIT: nevermind.... aweaver you are awesome for posting that link. I will try to chmod the time_in_state file as outline in that thread and see if I can make it 48 hours without it making the cpu.
Do you know how specifically this is kernel related? I've had this issue with Chad's kernel, and the old KingKlick kernels. I have NOT had the issue on the very old Hydra kernel or any of the stock kernels. I guess its just the number of frequencies they enable? Man if chmod'ing that file is all it took to fix that bug.... you are my friggin hero dude. I've been mostly using Chad's kernel since December and love it but MAN I hated the phone locking up every two days!
bast525 said:
Do you know how specifically this is kernel related? I've had this issue with Chad's kernel, and the old KingKlick kernels. I have NOT had the issue on the very old Hydra kernel or any of the stock kernels. I guess its just the number of frequencies they enable?
Click to expand...
Click to collapse
Yeah, it's mostly just the number of frequencies they enable. Also a potential, but smaller, contributor would be governor settings that result in a more even distribution of frequencies, which would result the overall file size getting larger more quickly. I first noticed this issue when I was doing stability testing that intentially created an even distribution of frequencies and made it less than a day before triggering the bug.
Well, I'm at 36 hours of running now, so far no issues. I did open the time in state file and noted it is still updating, though the permissions were correctly changed.
My only problem now is the leaked GB roms that popped up today... I want to flash one but I REALLY want to see if after six months of dealing with this very annoying bug, if it is finally fixed. I gonna TRY to resist flashing til tomorrow.
Any way to make the change in permissions to the time in state file stick? It undoes the changes on reboot
bast525 said:
Any way to make the change in permissions to the time in state file stick? It undoes the changes on reboot
Click to expand...
Click to collapse
That's where the init.d script comes in. Put the following in a file in /etc/init.d/ and set permissions to 755:
Code:
#!/system/bin/sh
while [ 1 ]; do
chmod 400 /sys/devices/system/cpu/cpu0/cpufreq/stats/time_in_state
sleep 300
done
If there is no /etc/init.d folder, and I create and throw that file in there will it work?
EDIT: nevermind... tried it, didn't work. Created the folder, put the file I'm there, rebooted. Let phone sit five mins, looked at permissions for time in state file, they were still at default. Does it need to have an extension? I just named it "fix".
Sorry, it looks like the /etc/init.d/ directory is not standard Android. The following is the last line in my /bootcomplete.inc.rc file, which looks like what enables running the scripts in the /etc/init.d directory:
Code:
/system/xbin/busybox run-parts /system/etc/init.d
I'm not sure if that line was automatically added when I installed busybox, or if something else put it there.
It's cool, I found an alternate method. There's a very small app in the market simply called "Autostart", to use it you create a script, name it autostart.sh, drop it in a folder you create in /data/, and it runs at boot same as if it were in init.d. I took your script and modified it a bit to make it not repeat every 300 seconds. I don't run SetCPU so I only need it to run once at start up.
I still haven't let the phone run for 48 hours straight to verify this fixed it, but still man thank you for the info, it is greatly appreciated. After reading that Google bug thread you linked, I have very high hopes that this is the fix I've been looking for for six months now.
Thanks again man and I'll report in a day or two as I have no more reasons to reboot for a while.
48 hour mark just passed and still running like a champ

[MOD][KERNEL] Lazy - yet another CPUfreq governor

My last effort of improving the ondemand governor by implementing a dynamic sampling rate did not work out as I hoped. Thus I did switch back to ondemand and, with a sampling_rate of 40000 and an up_threshold of 80, XviD playback with MobyPlayer was fine and also FLAC playback with MortPlayer was ok. However, as Morfic and others have noted the GUI feels a bit sluggish with these settings and suggested to try a sampling_rate of 15000. I tried this before, however with a lower rate, playback with MobyPlayer has serious problems with cracks and stuttering sound. Exactly this was the motivation to try and implement this dynamic sampling rate in the first place.
So the question is, why does a lower sampling rate lead to these problems? Since a smaller sampling rate means a greater flexibility and responsiveness, missing the point in time when the frequency should be increase cannot be the cause. Also the polling function which determines the frequency change is relatively light-weight, thus the extra CPU load for calling this function more often should also not be the problem. The only reason I could think of was that maybe switching the CPU frequency at these fast rates might be the root of the problem.
Thus I took the ondemand governor and implemented an additional parameter 'min_timeinstate' defining a minimum time the CPU will stay in a certain frequency state before it will be allowed to switch frequencies again. This way one can have a fine grained polling by setting the sampling_rate to a low value without running into problems with these fast frequency changes.
I did some extensive testing with a sampling_rate of 10000, min_timeinstate of 40000 and up_threshold of 90 and FLAC, mp3 and video playback all seem to work flawlessly. So it seems the root of the problem was indeed that the CPU does not handle fast frequency changes well.
I tested several apps and games and so far I have not found anything that this governor cannot handle. Thus I dare to say that it seems to be the superior choice over ondemand.
Test it, torture it, try to break it - I am looking forward to your feedback.
Changes to the source: http://www.pastie.org/2586936
BUGFIX:
Fixed compilation error with 'lazy' chosen as default CPUfreq governor.
Bug fix: http://www.pastie.org/2587040
BUGFIX #2:
Added sysfs parameter 'screenoff_maxfreq' which when enabled will cause lazy to always select the maximum frequency while the screen is off.
Bugfix: http://www.pastie.org/2775900
No further patches will be published here. I have set up a git repo for all my tweaks. Each mod has its own branch to keep the tweaks cleanly separated and one can simply pull the latest patches from the corresponding branch.
https://github.com/Ezekeel/GLaDOS-nexus-s/tree/lazy
now this sounds interesting
Did you try building with lazy as default?
EDIT: builds fine as option, but not default. Too eager to try it to go fix code.
Oh, you are right. That does not work. I will fix it.
Bug fix: http://www.pastie.org/2587040
do we have to wait for this to be implemented in netarchy etc to use it .. ?
theallian said:
do we have to wait for this to be implemented in netarchy etc to use it .. ?
Click to expand...
Click to collapse
Yes, unless you can compile a kernel yourself you will have to wait. I do believe that there's a testing Trinity TEUV kernel by Morfic in his thread with the new governor as an option if you want it now.
theallian said:
do we have to wait for this to be implemented in netarchy etc to use it .. ?
Click to expand...
Click to collapse
Check trinity thread, there are 2 kernels with it and cooler colors, closer to voodoo punchy (if not exactly punchy)
Sent from my Nexus S using Tapatalk
awesome work ezekeel!
its running great here with the teuv kernel and the lazy governor. no complaints as of yet.
youll have to wait for it to be compiled into netarchy, or compile it yourself. you can try it already compiled into the trinity teuv kernel found here http://forum.xda-developers.com/showpost.php?p=17839805&postcount=4304
Yep it's working awesome so far. Using morfic's teuv with lazy gov.
Code:
diff --git a/include/linux/cpufreq.h b/include/linux/cpufreq.h
index 2c00088..b03d1e1 100644
--- a/include/linux/cpufreq.h
+++ b/include/linux/cpufreq.h
@@ -357,6 +357,9 @@ extern struct cpufreq_governor cpufreq_gov_performance;
#endif
#ifdef CONFIG_CPU_FREQ_DEFAULT_GOV_PERFORMANCE
#define CPUFREQ_DEFAULT_GOVERNOR (&cpufreq_gov_performance)
+#elif defined(CONFIG_CPU_FREQ_DEFAULT_GOV_LAZY)
+extern struct cpufreq_governor cpufreq_gov_lazy;
+#define CPUFREQ_DEFAULT_GOVERNOR (&cpufreq_gov_lazy)
#elif defined(CONFIG_CPU_FREQ_DEFAULT_GOV_POWERSAVE)
extern struct cpufreq_governor cpufreq_gov_powersave;
#define CPUFREQ_DEFAULT_GOVERNOR (&cpufreq_gov_powersave)
seems to do the trick
Hi Ezekeel, an observation in the trinity thread mentioned that some games lagged really badly, i think it could be with too slow changes, due to your time delay feature. I'm not sure if it's feasible, but maybe a solution to this could be a dynamic time delay, longer for the lower frequencies (suited for music while screen off) and shorter for the higher frequency (for these intensive games).
Now I'm not a kernel guy, although I do wish to learn at some stage in the near future. This is just the product of some quick thinking, I have no idea if it is a workable theory or if it would just cause overhead and create more lag. Please let me know what you think though!
Sent from my Nexus S
Been running the TEUV kernel with the lazy governor on since last night and here are a few of my observations:
1) Amazing battery performance. During heavy use with pretty much everything on and the screen at maximum brightness, I lost a mere 6% over a 30 minute period. Turn the screen down and back to normal use? 1% over 2 hours. Not to mention the fact that for the first time in a very long time my phone charged to 99% and when pulled off the charger stayed there for a good amount of time.
2) Upon booting, things stutter a bit. Most likely due to the fact that a lot of things are trying to start fighting for resources and since its a UV kernel, there isn't as much power to go around. Its not horrible like using my MT3G was but its noticeable. Also, there is, again just after boot, about a .25 or .5 second delay between pressing icons and the subsequent action happening.
3) After that initial slowness, it runs extremely smoothly. Like buttah, as mike myers would say.
All in all, I'm quite fond of this governor. I've tried UV kernels before on my phone and it tends to dislike them. Like random reboots and random bootloop, dislike them. Perhaps its the rom, perhaps its morfic's coding, or perhaps its just the governor. Whatever it is the collaboration is absolutely phenomenal. Kudos.
Jonathon Grigg said:
Hi Ezekeel, an observation in the trinity thread mentioned that some games lagged really badly, i think it could be with too slow changes, due to your time delay feature. I'm not sure if it's feasible, but maybe a solution to this could be a dynamic time delay, longer for the lower frequencies (suited for music while screen off) and shorter for the higher frequency (for these intensive games).
Now I'm not a kernel guy, although I do wish to learn at some stage in the near future. This is just the product of some quick thinking, I have no idea if it is a workable theory or if it would just cause overhead and create more lag. Please let me know what you think though!
Sent from my Nexus S
Click to expand...
Click to collapse
I just downloaded and tried the two games which were mentioned in the Trinity thread, Fruit Ninja and Air Attack, and I am experiencing no problems at all with the reference kernel from my sig set to 100/800. So until more people report such problems I am inclined to believe that these are not directly related to the governor.
kenvan19 said:
2) Upon booting, things stutter a bit. Most likely due to the fact that a lot of things are trying to start fighting for resources and since its a UV kernel, there isn't as much power to go around. Its not horrible like using my MT3G was but its noticeable. Also, there is, again just after boot, about a .25 or .5 second delay between pressing icons and the subsequent action happening.
Click to expand...
Click to collapse
I think that is normal. The GUI starts up before all programs and settings are properly loaded. For example if I open the app drawer directly after boot, only the stock apps are visible and after some seconds all my apps appear one after another. It takes about 10s after the GUI has started until everything is setup.
Ezekeel said:
I just downloaded and tried the two games which were mentioned in the Trinity thread, Fruit Ninja and Air Attack, and I am experiencing no problems at all with the reference kernel from my sig set to 100/800. So until more people report such problems I am inclined to believe that these are not directly related to the governor.
I think that is normal. The GUI starts up before all programs and settings are properly loaded. For example if I open the app drawer directly after boot, only the stock apps are visible and after some seconds all my apps appear one after another. It takes about 10s after the GUI has started until everything is setup.
Click to expand...
Click to collapse
I personally think your new governor is working great. Only time will tell if its better on the battery. I really hope so. I'm hoping for at least another hour of use. is it possible?
Sent from my Nexus S using xda premium
dsellers2 said:
I personally think your new governor is working great. Only time will tell if its better on the battery. I really hope so. I'm hoping for at least another hour of use. is it possible?
Sent from my Nexus S using xda premium
Click to expand...
Click to collapse
I expect the battery runtime to be similar to ondemand. On the other hand, I have made the experience that modifying governors is a delicate business and seemingly small changes can have an unexpectedly large impact. So as you said, only time will tell.
If the sampling rate is 10000 and the min_timeinstate is 15000, will the governor not consider load changes until min time is over, but at the normal sampling intervals? Which would mean an essential 20000 in this state?
I was looking through the patch on pastie right now, but opera closed twice, so I figured I ask you instead.
Sent from my A501 using Tapatalk
morfic said:
If the sampling rate is 10000 and the min_timeinstate is 15000, will the governor not consider load changes until min time is over, but at the normal sampling intervals? Which would mean an essential 20000 in this state?
I was looking through the patch on pastie right now, but opera closed twice, so I figured I ask you instead.
Sent from my A501 using Tapatalk
Click to expand...
Click to collapse
When the frequency is changed the next sampling will be performed min_timeinstate µs later, after that sampling with the normal rate is resumed. So in your example, the first sampling after a frequency change will be performed after 15000 µs, after that every 10000 µs.
Hi EZkeel your governor implemented the latest TEUV kernel has helped improved my battery life by close to 20!, was getting around 10hrs of battery with heavy use. About 3+hrs of screen time, display set to auto. Was able to run for 13hrs today.
One day of testing so far but I haven't seen my phone charge to 99% or last over 13hrs until I put the latest TEUV kernel with lazy in. Will continue testing throughout the week. No compatibility issues with the apps so far. Was running on demand governor previously and with the previous version of TEUV
Thanks for your work on the governor development.
robin_77 said:
Hi EZkeel your governor implemented the latest TEUV kernel has helped improved my battery life by close to 20!, was getting around 10hrs of battery with heavy use. About 3+hrs of screen time, display set to auto. Was able to run for 13hrs today.
One day of testing so far but I haven't seen my phone charge to 99% or last over 13hrs until I put the latest TEUV kernel with lazy in. Will continue testing throughout the week. No compatibility issues with the apps so far. Was running on demand governor previously and with the previous version of TEUV
Thanks for your work on the governor development.
Click to expand...
Click to collapse
Looking at your screen shots vs what mine is and my 2nd most is gallery. Very confused as to why this is
Sent from my Nexus S 4G using xda premium
you may have a corrupt media file , so everytime your device syncs it gets stuck. clear your files.
turn off device, remove battery for 2mins.; try again
possibly flag the directories you don't want scanned with .nomedia file (there's an app to do this)
anyways excellent job Ezekeel, 2nd day and lets see what happens with battery life.

2.6.35.7 vs 2.6.32.9 kernel

Okay here's the thing,
my last most stable ROM I was running on was CacingKalung's MIUI. And that had a wakelock issue too. I'm pretty sick of it, so I decided to try Edowar's CM7 and maybe even Bumblebee rom.
when I flashed to Edo's CM7, all was good, no wakelocks, apn thing I had it fixed by using yashwant's APN fix.. until I ran Temple Run. The frames were skipping really bad, in the store part the Back button couldnt even been seen until scrolled up. The vines kept disappearing, and the gameplay was even worse, laggy. Mind you I updated the game already.
I never had this issue when I was on 2.6.32.9. So tell me, with me switching back to .6.32.9 now, what do I have to lose? Please share your thoughts and opinion!
PS: Thinking of going back to Maclite rom, least it's stable.
Bro its too early to comment on difference between these two kernel.
for .32.9 kernel various custom kernel are available while for .35.7 only TJ's custom version is available and I think it need more support from kernel developer.
till now if we think for .32.9 kernel, mOCk is best for gamer.
I have found that the best gaming combo is maclite with bumblebee kernel for mortals(non devs like me) and newer gpu drivers as far as benchmarks go. It also gives a superfast rom. What I have found so far is that eventhough there is a scarcity of custom kernels for .35 the situation will not improve any better than the performance of tj with a score of 27.9. I have already achieved 27.7 at 1.4 GHz with the. 32 kernel and 27.2 with stock speed in nenamark 2.2. Still this is just my opinion and it is very likely that your device may hit tjs score.
Sent from my MacLite-FB0 using XDA
Between .35 and .32... seem that .35 hasn't been hacked with wakelock trick while wifi is on. This hack is needed to prevent our device to meet Sleep of Death when it goes to sleep while wifi is on.
Sent from my CSL-MI410 @CKMIUI 2.4.13 using Tapatalk
CacingKalung said:
Between .35 and .32... seem that .35 hasn't been hacked with wakelock trick while wifi is on. This hack is needed to prevent our device to meet Sleep of Death when it goes to sleep while wifi is on.
Sent from my CSL-MI410 @CKMIUI 2.4.13 using Tapatalk
Click to expand...
Click to collapse
oh i thought the SoD had nothing to do with kernel.. lol well thank God im back to. 32..
am still getting pretty bad wakelocks every now and then x.x
but honestly i tried STOCK sharp rom to play temple run and even on that the frames were choppy, graphics n sum sprites were missing. ugh.
anyone with good batt life on either roms can u guys share your rom,kernel n config? my phone lasts about 8-10 hrs a day only if heavily used non stop.
cheezzzz said:
Okay here's the thing,
my last most stable ROM I was running on was CacingKalung's MIUI. And that had a wakelock issue too. I'm pretty sick of it, so I decided to try Edowar's CM7 and maybe even Bumblebee rom.
when I flashed to Edo's CM7, all was good, no wakelocks, apn thing I had it fixed by using yashwant's APN fix.. until I ran Temple Run. The frames were skipping really bad, in the store part the Back button couldnt even been seen until scrolled up. The vines kept disappearing, and the gameplay was even worse, laggy. Mind you I updated the game already.
I never had this issue when I was on 2.6.32.9. So tell me, with me switching back to .6.32.9 now, what do I have to lose? Please share your thoughts and opinion!
PS: Thinking of going back to Maclite rom, least it's stable.
Click to expand...
Click to collapse
I'm on latest edowar's CM7, got lag issue when playing Temple run on 35.7 kernel, but then after I set the I/O governer to Deadline, and CPU gorvener to smartassv2 or interactive X, it seems better.
From my experience with this rom, the gpu lib is not very good to play HD games (got so much lag when playing Asphalt 6, but not dungeon defender 3), or very much like fast moving graphic game like temple run. I've test to install the older gpu lib from 32.9 kernel (just the gpu lib, not the kernel itself), it gave me back low nenamark2 value, but the lag is gone like 32.9 kernel cm7, however, it seems making the fast locking gps not to be that fast anymore (wonder why, still need to play2 around again).
About the wifi, it really piss me off as it will brings the phone to sleep of death each time it goes to sleep although I set the wifi sleep policy to 'never'. however, the workaround would be to install Advanced Wifi Lock from the market for now.
Originally Posted by CacingKalung
Between .35 and .32... seem that .35 hasn't been hacked with wakelock trick while wifi is on. This hack is needed to prevent our device to meet Sleep of Death when it goes to sleep while wifi is on.
Sent from my CSL-MI410 @CKMIUI 2.4.13 using Tapatalk
Click to expand...
Click to collapse
oh i thought the SoD had nothing to do with kernel.. lol well thank God im back to. 32..
am still getting pretty bad wakelocks every now and then x.x
Click to expand...
Click to collapse
Because of the wakelock hack, when wifi is on, our device will never go to sleep. So it's very recomended to turn on wifi only when we need it and turn off if we've done. so our device will able to enter sleep mode when in idle to save power.
Sent from my CSL-MI410 @CKMIUI 2.4.13 using Tapatalk
CacingKalung said:
Because of the wakelock hack, when wifi is on, our device will never go to sleep. So it's very recomended to turn on wifi only when we need it and turn off if we've done. so our device will able to enter sleep mode when in idle to save power.
Sent from my CSL-MI410 @CKMIUI 2.4.13 using Tapatalk
Click to expand...
Click to collapse
CK bro, Mokee v0.3 still has the SOD issue though it uses mock kernel(0.32). Any ideas on how to fix this?
whyzee said:
CK bro, Mokee v0.3 still has the SOD issue though it uses mock kernel(0.32). Any ideas on how to fix this?
Click to expand...
Click to collapse
Are you sure? While wifi was on?
As far as I know, mOCk and SD and also last TJ's miui kernel have been patched with wakelock if wifi is on.
Or maybe you can use wifi library and driver from miui just to make sure.
If still remain then I don't have an idea for that problem.
Sent from my CSL-MI410 @CKMIUI 2.4.13 using Tapatalk
thanks for the info bro. I am certain that it happened. I'll try again and see.
EDIT
I can't replicate the problem. Though the lockscreen turns on, I'm able to get back without any problems. So I think may be its gone now. thanks anyway.....
UPDATE
fixed sod after using different wifi lib files
natsuke said:
I'm on latest edowar's CM7, got lag issue when playing Temple run on 35.7 kernel, but then after I set the I/O governer to Deadline, and CPU gorvener to smartassv2 or interactive X, it seems better.
From my experience with this rom, the gpu lib is not very good to play HD games (got so much lag when playing Asphalt 6, but not dungeon defender 3), or very much like fast moving graphic game like temple run. I've test to install the older gpu lib from 32.9 kernel (just the gpu lib, not the kernel itself), it gave me back low nenamark2 value, but the lag is gone like 32.9 kernel cm7, however, it seems making the fast locking gps not to be that fast anymore (wonder why, still need to play2 around again).
About the wifi, it really piss me off as it will brings the phone to sleep of death each time it goes to sleep although I set the wifi sleep policy to 'never'. however, the workaround would be to install Advanced Wifi Lock from the market for now.
Click to expand...
Click to collapse
finally someone else who feels the same wayabout the gpu lib!
thanks for the wifi suggestion bro!

[Kernels] N4N-2DQI6/I5-[12-19-17]710,715,810,815,817

N4N (Neked 4 Nougat) I6/I5 Kernel for 710,715,810,815,817​There's just nothing like good ol' stock. Everything just works. Even after 2 years they're still selling these Tabs. And they can still compete with newer ones. So on with the show!!!!
You flash this at your own risk. I'm am not responsible for anything that happens!!! There's the discaimer.
MAKE A FULL BACKUP IN TWRP!!!!!
These kernels modify your ROM. So you WILL NOT be able to FLASH/RESTORE the stock kernel only.
A kernel tuner (Synapse) is enabled for V2.0 only - Get Synapse from the Playstore
See post #2 for Synapse info and other stuffs​
Install instructions:
You MUST HAVE TWRP installed to flash this!!
MAKE A FULL BACKUP IN TWRP!!!!! YOU HAVE BEEN WARNED!!!!!!!!!
Flash Kernel- it'll wipe cache, dalvic automatically,
Flash SU v2.82, https://s3-us-west-2.amazonaws.com/supersu/download/zip/SuperSU-v2.82-201705271822.zip
Wipe cache and dalvic when SU asks,
Reboot
Profit!!
So whats in this thing?!
V2.0 Synapse enabled (You should use a tuner (Synapse recommended) for this one as that's all that was added) But will run fine without
All below plus
Under volting
Overclock-big 2.2Ghz LITTLE 1.6Ghz
Underclock-big 500Mhz LITTLE 200Mhz
HMP tweaks and tunables
KSM with tweaks
Enabled all stock govs
V1.3 All below plus
Patched to 108EOL
V1.2 All below plus - Pulled for random issues-Use V1, V1.1 or V1.3
Patched to 3.10.105
817 only has this version
V1.1 All below plus
Patched to 3.10.61
V1
4.9 UBERTC Toolchain
Screen Mirroring built in - Seems to only works with Samsung enabled TV's
CIFS
OTG
TCP's-Cubic is default. Reno, Westwood+ and HTCP added
CDROM support
XPAD game controller support
A few tweaks
Bugs
If you have Synapse installed , make all settings to stock and uninstall. It'll probably still give you a "System UI error". Just move on, and reboot after a couple minutes.
Source
https://github.com/engine95
Thanks to
@andyroidking - for the 715,815, and 810 boot.img's
Donations. Never expected, but much appreciated
@sebmue
@andyroidking
Downloads
V2.0 Synapse enabled and recommended
710-I6 - https://www.androidfilehost.com/?fid=673791459329065685
810-I6 - https://www.androidfilehost.com/?fid=962021903579496837
715-I6 - https://www.androidfilehost.com/?fid=745849072291693169
815-I5 - https://www.androidfilehost.com/?fid=889964283620770516
817-3CQE4 - https://www.androidfilehost.com/?fid=962021903579499361
V1.3
710-I6 - https://www.androidfilehost.com/?fid=889964283620768631
810-I6 - https://www.androidfilehost.com/?fid=745849072291691184
715-I6 - https://www.androidfilehost.com/?fid=962021903579494937
815-I5 - https://www.androidfilehost.com/?fid=962021903579494940
817-3CQE4 - https://www.androidfilehost.com/?fid=889964283620772774
V1.2 Only for 817
817-3CQE4 - https://www.androidfilehost.com/?fid=745849072291694892
V1.1
710-I6 - https://www.androidfilehost.com/?fid=962021903579492239
810-I6 - https://www.androidfilehost.com/?fid=745849072291688529
715-I6 - https://www.androidfilehost.com/?fid=817906626617946405
815-I5 - https://www.androidfilehost.com/?fid=889964283620765858
817w-3CQE4 - https://www.androidfilehost.com/?fid=673791459329067028
V1
710-I6 - https://www.androidfilehost.com/?fid=889964283620763792
810-I6 - https://www.androidfilehost.com/?fid=817906626617944296
715-I6 - https://www.androidfilehost.com/?fid=745849072291686448
815-I5 - https://www.androidfilehost.com/?fid=889964283620763794
I'm just not a fan of the DevDB format anymore. It's difficult to update and modify. So this is what you get.
SYNAPSE Kernel Tuner - Maybe some others stuffs too.....​
Well, because Synapse is built in now, here's some needed info on it:
Synapse is an awesome and custom built kernel tuner. It can do so much more than a lot of other tuners. And because we have Exynos processors, it's about the only one we can use. But it can create a lot of problems getting to work. Hence this post. :silly:
I decided to make this built in instead of a flashable add on. It's actually easier and less invasive to do it this way.
UCI, is what Synapse uses to work. It's built into the ramdisk. I've added a lot of tunables and may add more later.
Now for the fixes and problems you WILL run into:
If you have issues, make sure to check out these threads. As they will probably have better info about Synapse issues. https://forum.xda-developers.com/showthread.php?t=2736986 and https://forum.xda-developers.com/showthread.php?t=3558888
No UCI support detected These are the famous words you WILL run into.
When will this show up?
When Synapse is first installed
On reboot
When you've pushed your Tablet too far
The fixes:
On first install from the Playstore. You will need to go to Settings>Apps>Synapse>Storage. Then select "Clear Cache", THEN "Clear Data. In that order!!
You may have to backout to Settings and do it one or two more times. You want the numbers for cache and data at "0".
On reboot: This usually works unless you've messed up the settings or flash another kernel
On reboot select "Menu" and "Close all".
Open Synapse and hit "CANCEL" to have your last settings applied. (I know it's not normal but it works)
Close synapse, and reopen it to make sure of settings. (If "apply or cancel" don't show up at the top, you're good to go.)
Worst case scenario:
You will need to go to Settings>Apps>Synapse. Select "Force stop". Select "Storage". Then select "Clear Cache", THEN "Clear Data. In that order!!
You may have to backout to Settings and do it one or two more times. You want the numbers for cache and data at "0".
If you follow the above, Synapse will work. It hasn't failed me in over 50 tests, probably more. :silly:
810 good to go. Have to wait til later this week to try mirroring, gave my daughter my sammy tv..
Edit: confirmation #5H0834012J5874316 - Thanks for your hard work!
andyroidking said:
810 good to go. Have to wait til later this week to try mirroring, gave my daughter my sammy tv..
Edit: confirmation #5H0834012J5874316 - Thanks for your hard work!
Click to expand...
Click to collapse
Thank you for the donation!
Hopefully the 810 looks better on the screen. The 710 is a weird screen size and doesn't fill it. But it still comes in handy.
My "Neked Maker Rom" has it built into the build.prop. But i found I could add it to the kernel. It'll still work with both installed, which I have.
engine95 said:
Thank you for the donation!
Hopefully the 810 looks better on the screen. The 710 is a weird screen size and doesn't fill it. But it still comes in handy.
My "Neked Maker Rom" has it built into the build.prop. But i found I could add it to the kernel. It'll still work with both installed, which I have.
Click to expand...
Click to collapse
I posted a couple days ago but apparently it didn't send. Screen is fine! Everything seems good. It is charging good again too, probably not kernel related but take the credit will ya...
andyroidking said:
I posted a couple days ago but apparently it didn't send. Screen is fine! Everything seems good. It is charging good again too, probably not kernel related but take the credit will ya...
Click to expand...
Click to collapse
Ok.
I'm wondering if it is Synapse with the battery. Guess we'll see.
Haven't spent a lot of time with the kernels lately. Trying some different things, but keep coming back to what I've had in the past.
EX kernel managers builder was interviewed on XDA recently. flar2 . Had some interesting things to say about SELinux. Guess that might be why his tuner doesn't have an option for permissive. Just food for thought for us all.
engine95 said:
Ok.
I'm wondering if it is Synapse with the battery. Guess we'll see.
Haven't spent a lot of time with the kernels lately. Trying some different things, but keep coming back to what I've had in the past.
EX kernel managers builder was interviewed on XDA recently. flar2 . Had some interesting things to say about SELinux. Guess that might be why his tuner doesn't have an option for permissive. Just food for thought for us all.
Click to expand...
Click to collapse
I'll read through that later... his kernel is what I use on my nexus9. His dev app is pretty good too, playstore- DevCheck Hardware and System Info. Smart guy.
Just a little update
I have the 710 and 810 kernels done. I should get 715 and 815 shortly. Then I'll get them uploaded
Just patched to 3.10.61. No additions til I get them as far as I can.
The kernel is EOL at 3.10.108
V1.1 for all 4 is in OP Just patched to 3.10.61
I've changed the way I'm doing it so it's faster. And it's easier for others to track.
I'm hoping to get all the way to 3.10.108 for the next version. Then the addons.
We shall see. :fingers-crossed:
Thank you ! Amazing and smooth as expected ??
810 v1.1 A-Okay
V1.2 is in OP!!!!! Just patched to 3.10.105, which is Feb 2017
Not quite sure if I'll go to 108, or do some addons. I'm kinda burned out on the patches.
More than likely I'll patch to 108, but I am dreading it as just the 3 more patches has over 750 fixes.
And seeing as how our Tablets AND our Kernel is at EOL, I might as well finish it out.
Regardless, enjoy.
engine95 said:
V1.2 is in OP!!!!! Just patched to 3.10.105, which is Feb 2017
Not quite sure if I'll go to 108, or do some addons. I'm kinda burned out on the patches.
More than likely I'll patch to 108, but I am dreading it as just the 3 more patches has over 750 fixes.
And seeing as how our Tablets AND our Kernel is at EOL, I might as well finish it out.
Regardless, enjoy.
Click to expand...
Click to collapse
Thank you, thank you. Don't burn out!
---------- Post added at 01:50 PM ---------- Previous post was at 12:56 PM ----------
andyroidking said:
810 v1.1 A-Okay
Click to expand...
Click to collapse
1.2 good to go of course.
Okay, so I don't know what is missing or changed in the kernel but all of a sudden I am getting horrible battery life I can almost watch it drop sometimes. If you're thinking of updating more maybe putting back some of the tweaks that you had? First?
Edit- it literally dropped a percent when I typed the above.
andyroidking said:
Okay, so I don't know what is missing or changed in the kernel but all of a sudden I am getting horrible battery life I can almost watch it drop sometimes. If you're thinking of updating more maybe putting back some of the tweaks that you had? First?
Edit- it literally dropped a percent when I typed the above.
Click to expand...
Click to collapse
Now that's interesting. I used my Tab last night and all was well.
It could be a rogue app. Or maybe you need to shut it off for a minute. Then boot it up again. There were a lot of changes in the kernel and resetting doesn't always fix things. The battery might need to relearn too.
Side note:
Ya know how when climbing you see a peak and you just have to keep going til you can see over it? Yeah, well, I've just about finished the patches to 108. It'll be a couple days. I just barely got it built and need to see how it does.
Here's some pics of battery for me. And I've installed the kernel a few times too. So reboots are included.
Edit
There's an old trick I still do for "Ghosts in the computer".
Power the Tab off, wait for bout 15 sec, then just press the vol down, vol up and power buttons. Just a press. Do it a few times then boot it up.
The old belief is that there is still an electrical charge that can keep things in memory. By doing this, it'll discharge any current that is lurking.
It may be nothing, but I feel better doing it.
engine95 said:
Now that's interesting. I used my Tab last night and all was well.
It could be a rogue app. Or maybe you need to shut it off for a minute. Then boot it up again. There were a lot of changes in the kernel and resetting doesn't always fix things. The battery might need to relearn too.
Side note:
Ya know how when climbing you see a peak and you just have to keep going til you can see over it? Yeah, well, I've just about finished the patches to 108. It'll be a couple days. I just barely got it built and need to see how it does.
Here's some pics of battery for me. And I've installed the kernel a few times too. So reboots are included.
Edit
There's an old trick I still do for "Ghosts in the computer".
Power the Tab off, wait for bout 15 sec, then just press the vol down, vol up and power buttons. Just a press. Do it a few times then boot it up.
The old belief is that there is still an electrical charge that can keep things in memory. By doing this, it'll discharge any current that is lurking.
It may be nothing, but I feel better doing it.
Click to expand...
Click to collapse
I really wish it was ghosts or Rogue app... rebooted 1.2 browsed Reddit for 17 minutes lost 7%. Flashed 1.1 browsed Reddit and as I'm typing this only 2% in 22 minutes.... We will see what 1.3 brings when it is cooked.
HTML:
andyroidking said:
I really wish it was ghosts or Rogue app... rebooted 1.2 browsed Reddit for 17 minutes lost 7%. Flashed 1.1 browsed Reddit and as I'm typing this only 2% in 22 minutes.... We will see what 1.3 brings when it is cooked.
Click to expand...
Click to collapse
So with 1.2 I was getting 3 to 4% at idle an hour battery drain. Installed 1.1 and overnight 8 hours 1%. Could just be something with my tab and that kernel not liking each other, lol.
andyroidking said:
So with 1.2 I was getting 3 to 4% at idle an hour battery drain. Installed 1.1 and overnight 8 hours 1%. Could just be something with my tab and that kernel not liking each other, lol.
Click to expand...
Click to collapse
I think I found the issue. It was very random and hard to find, but I'm pretty sure I have it fixed now. It'll come out when I finish/ release 108.
For now, all the kernels run fine, so if there's no issues with what's installed, then just keep running it till I get 108 up. Like I said, it's a random thing with 105. Hit or miss. I'll pull v1.2(105)when I get v1.3(108) up.
Thanks for making me look into it. Lol
Thank you @engine95 for your amazing effort. You made my tablet super smooth. And really I was about to sell it, but you let me keep it for another year . Thanks ??????
V1.3 is in OP for all 4 kernels Patched to 3.10.108EOL
I've pulled V1.2 as it had random issues and/or isn't needed now with the fixed V1.3
Just a little FYI
Stock 3.10.9, vs Patched 3.10.108 - Stability vs Security
V1 is the mostly stock 3.10.9 kernel. It will be the most stable of all the kernels as Samsung has tailored it to our Tablet
The rest
These should be the most secure as it has all the Linux patches to date. Thing is, Samsung hasn't made adjustments to optimize the kernel to anything past 3.10.9
Patching it can cause some problems. Usually WiFi, Bluetooth, Battery, USB.
Patching it may make more current products useful. Routers, Game Pads, Hard drives etc....
So from here, I'll wait and see if we have more issues with the patched kernels before moving on.
My options?
Keep going with the patched kernel to build on
or
Go back to 3.10.9 for the stability, and leave the patched kernel where its at.
Let me know how V1.3 does so I can start with the addons on one or the other.
Thanks all and enjoy. :good:

Categories

Resources