[KERNEL][VZW][4.4.3] ★☆ Kangaroo Kernel ☆★ 04 OCT | v068

Search This thread
By:
Advanced…
poondog

poondog

Inactive Recognized Contributor
Dec 11, 2011
14,998
12,953
Melbourne
cesium said:
Heh, that I did, 116 is working great so far. Speaking of default settings, do you recommend leaving fiops i/o sched? Am curious cuz of the improvements listed for bfq and deadline in the changelog. I like fiops becaues it's made for nand, but always up for trying something new
Click to expand...
Click to collapse

yeah thats why its default. I have good results for it, but deadline and bfq are good also :)
 
poondog

poondog

Inactive Recognized Contributor
Dec 11, 2011
14,998
12,953
Melbourne
UPDATE: 5th Aug yay

Kangaroo Kernel v122 4.4.2:
.thermal: bring in new thermal driver. Seems to have better control over freqs compared to before. Butchered it. Now we have some weird roo thermal beast. See below. (myfluxi, me)
.introduce powersaver cap for limiting cpu max freq (me, see below)
.patch to upstream 3.4.99 (kernel.org)
.2w: rewrote this is some places. Add support for bi-directional sweep2wake and sweep2sleep (thx @Savoca for a few pointers and the idea, flar for the original code, see below)
.compile with new toolchain 4.9.1 Christopher83 2014.07
.disable HTC write protection (thx Tiny)
.clean up some drivers of unneeded code
Click to expand...
Click to collapse

THERMAL
This driver is completely new to everyone. Its a much simpler driver and it works much more effectively for me. I have coded 3 steps into the base thermal driver.
-> low, 45C, 1350MHz
-> mid, 55C, 918MHz
-> max, 70C, 486MHz
So when it gets to these temps it will throttle to the freq set. I have tried to keep these steps similar to previous releases for simplicity. This can be changed in sysfs or through the scripts attached. Kangaroo script is the one by headless updated for new thermal and cpu cap drivers. Other script is for thermal only. Please use these as the new thermal driver is NOT supported by any apps.
new driver also features smart polling. This means, that when the CPU temp is low it will poll less frequently, to save power. When the cpu nears the low temp point, it will poll slightly faster. As the CPU gets hotter and hotter it will poll faster and faster. This is to make sure that the CPU is throttled when it needs to be, otherwise it may get too hot before polling again
This is the new thermal driver we are using from now on. I will be working on refining it, and if anyone wants to review the code its on my git. Its been reliable in my testing
The thermal driver is derived from myfluxi thermal for N4, ported to HTC One and with some new features and enhancements added by me.
Click to expand...
Click to collapse

POWER SAVER CPU CAP
Since we use PNPMGR in this kernel, its previously not possible to set a max freq (for anyone who wants to set for battery or whatever reason)
Now it is possible.
Via the thermal driver, the user is able to set a max freq for the kernel to scale to. Default is OFF, and 1458MHz. If you switch ON, it will cap to this. Thermal steps however will ALWAYS take priority. The point is to have POWERSAVE FREQ > ANY THERMAL LEVEL FREQ. Obviously it will work if you set anything but I designed it to be more. Hopefully the bugs have been ironed out in the past weeks that this has been in testing.
You can set this freq, and turn the power saver on or off in the scripts below.

This is NOT a replacement for HTC powersaver. This is more of a way to cap max freq/underclock.
Click to expand...
Click to collapse

2w:
based on flar's implementation, but we now have bi directional support. Thanks Savoca for the original idea and some pointers. Sweep2wake is now default 1, so now we have sweep2wake and doubletap2wake on by default. Of course you can change in script or something.
The new feature means that you can sweep left or right with screen on or off to toggle the screen. Both directions will register a power button press. We have found that it is much more convenient especially with one handed use. Hopefully you guys will find the same.
Click to expand...
Click to collapse

DOWNLOAD:
http://www.androidfilehost.com/?fid=23578570567723498
md5: 44da3ff0229250d41c53f696bb2a2a3a
Click to expand...
Click to collapse

as you guys can see I have been busy coding new features. These will eventually come to 4.4.3 kernel. EVENTUALLY. Enjoy the new kernel :)

EDIT:
For scripts see here pls
http://xdaforums.com/showpost.php?p=54423681&postcount=4878
 
  • Like
Reactions: Banished Angel, tflogic, unconfuse and 5 others
Mortal Eternity

Mortal Eternity

Senior Member
Nov 23, 2012
447
145
Karachi
poondog said:
I started on this kernel last week, idk if I will support wls and wlv
Click to expand...
Click to collapse

Sorry for a noob question:

in Pakistan there is no CDMA services... please let me know if ur kernel works with the GSM version phones as well, also will it go well with Viper one?... also if you can please enlighten me if it supports 2 way call recording as well.

Thanks :)
 
Last edited:
poondog

poondog

Inactive Recognized Contributor
Dec 11, 2011
14,998
12,953
Melbourne
Mortal Eternity said:
Sorry for a noob question:

in Pakistan there is no CDMA services... please let me know if ur kernel works with the GSM version phones as well, also will it go well with Viper one?... also if you can please enlighten me if it supports 2 way call recording as well.

Thanks :)
Click to expand...
Click to collapse

??
idk what you mean. You have m7, not m7wl?? International kernel can be found in the link in first post. For GSM phones. Thats what I have. No 2 way call recording support has been addded. I hear theres an app for that, idk what it is
 
Mortal Eternity

Mortal Eternity

Senior Member
Nov 23, 2012
447
145
Karachi
poondog said:
??
idk what you mean. You have m7, not m7wl?? International kernel can be found in the link in first post. For GSM phones. Thats what I have. No 2 way call recording support has been addded. I hear theres an app for that, idk what it is
Click to expand...
Click to collapse

i have htc one m7 Verizon... i got it unlocked so that i could use it with local network (Telenor Pakistan) its a GSM network not CDMA... so can i flash your kernal on my phone? :)
 
A

AznInertia

Senior Member
Jun 12, 2010
523
68
Poon, you have done a phenomenal job with this kernel. Can't wait for v 132! Or hopefully we can find someone to compile it if you are too busy? =)
 
DroidOnRoids

DroidOnRoids

Senior Member
Dec 30, 2011
1,168
398
Baltimore, MD
So I have a question. I usually rock stock kernels on a stock rooted rom. I want to venture over to this kernel as it's not too flashy. My question is if I enable Power Saver mode, will the kernel automatically underclock and do its job just like the stock HTC kernel?
 
poondog

poondog

Inactive Recognized Contributor
Dec 11, 2011
14,998
12,953
Melbourne
AznInertia said:
Poon, you have done a phenomenal job with this kernel. Can't wait for v 132! Or hopefully we can find someone to compile it if you are too busy? =)
Click to expand...
Click to collapse

Can if you want, I don't really mind





DroidOnRoids said:
So I have a question. I usually rock stock kernels on a stock rooted rom. I want to venture over to this kernel as it's not too flashy. My question is if I enable Power Saver mode, will the kernel automatically underclock and do its job just like the stock HTC kernel?
Click to expand...
Click to collapse

Yep
 
  • Like
Reactions: DroidOnRoids
hightech316

hightech316

Senior Member
Jan 27, 2011
1,281
220
github.com
OnePlus 8T
poondog said:
I started on this kernel last week, idk if I will support wls and wlv
Click to expand...
Click to collapse

Hmmm I thought it would support 4.4.3 ... Now I might downgrade to give this beast a test!

Are there plans in the future for OverClocking & UnderClocking?


Thanks for bringing it over to the Verizon, poon.

I had posted a while back in the original HTC One forums, requesting a verizon kernel. Did you by chance see that?
 
Last edited:
Uzephi

Uzephi

Inactive Recognized Contributor
Apr 20, 2012
3,439
1,892
Phoenix
Google Pixel 3a
hightech316 said:
Hmmm I thought it would support 4.4.3 ... Now I might downgrade to give this beast a test!


Thanks for bringing it over to the Verizon, poon.

I had posted a while back in the original HTC One forums, requesting a verizon kernel. Did you by chance see that?
Click to expand...
Click to collapse

Not even on my full drain and gsam stats doubled for me. We will see after a few drains though.

Running the 4.4.2 version of Nusense 6, which is based off official Verizon release, unlike the newer versions.

Sent from my HTC6500LVW using XDA Free mobile app
 
poondog

poondog

Inactive Recognized Contributor
Dec 11, 2011
14,998
12,953
Melbourne
hightech316 said:
Hmmm I thought it would support 4.4.3 ... Now I might downgrade to give this beast a test!

Are there plans in the future for OverClocking & UnderClocking?


Thanks for bringing it over to the Verizon, poon.

I had posted a while back in the original HTC One forums, requesting a verizon kernel. Did you by chance see that?
Click to expand...
Click to collapse

Not for CPU. underclock is available anyway :) use the cpu limiter feature, to cap cpu even with pnpmgr enabled
What do you mean? Is this when you asked for support or recently?
 
S

shine_118

Senior Member
Sep 29, 2012
134
22
Hi, guys. I saw two updated v132 and v138 on international m7, but Verizon m7 no update, hope you go on support update, Thank you very much.
 
You must log in or register to reply here.

Similar threads

santod040
★ ☆ [ROM][5.0.2] NuSenseSIX | Sense 6.0 | 7.17.1540.7 | v10.6 [02-13-15]
Replies
3K
Views
352K
S
solace mukhtar
santod040
★ ☆ [ROM][5.0.2] NuSenSeveN | Sense 7.0 | v4.00 [05-06-15] ☆
Replies
1K
Views
176K
Z
zeeBomb
santod040
★ ☆ [ROM][4.4.2] UnSenseONE-KitKat | Beta2.1 | Beats Audio | Zoe | Google/HTC Hybrid | 01-05-14 ☆ ★
Replies
769
Views
100K
fulishone
fulishone
santod040
★ ☆ [ROM][5.0.2][OFFICIAL] LiquidSmooth | v4.0 | Lollipop [02-28-15]
Replies
886
Views
124K
D
Dormat
Turge
[ROM][Vzw] ViperOne 7.0.1 • Tweaks, HUB, Themes, Icons, OTAs and more
Replies
4K
Views
401K
R
rgps

Top Liked Posts

24 Hours All time
  • There are no posts matching your filters.
  • 18
    poondog
    poondog
    Kangaroo Kernel for the HTC One GSM: Sense ROMs
    Linux 3.4.104+, with many features and improvements from Linux 3.12 and 3.14
    ePLKfMb52P-8.png

    For android 4.4.3 Sense ROMs

    Based off HTC 6.09.401.5 sources for 4.4.3 Sense

    Code: 
    * Standard disclaimer: 
* Your warranty is now void.  Use at your own risk.
* In other words: It is your choice to use this software,
* I am not responsible for any damage or problems!

    Naming convention
    KANGAROO KERNEL MKIV IS FOR 4.4.3 ONLY (it will have _443 at the end of the filename)
    JOEY AOSP CM (ckxxx): AOSP/CM based ROMs depreciated
    JOEY GPE (post 4): sense based 4.4.2 and GPE 4.4.2 depreciated
    KANGAROO KERNEL SENSE (kkxxx): sense based 4.4.2 ROMs and GPE only depreciated
    Click to expand...
    Click to collapse

    Features:
    Compiled with latest Linaro Christopher83 Linaro 4.9.1 toolchain
    Better performance and battery savings due to the use of more efficient algorithms and updated, optimized code
    Kernel is aimed at reducing heat while saving battery and increasing performance over stock

    IN DEPTH FEATURE LIST IN CHANGELOG AND ON GIT

    Better to look at changelog cause I miss things when I try to condense it :p

    CPU governors (italics for default): ondemand, performance, userspace, msm-dcvs, powersave
    io sched: no-op, deadline, CFQ

    KANGAROO CONTROL SCRIPT IN POST #3
    Can be used to control all options, plus more :)
    Click to expand...
    Click to collapse

    Installation:
    1. Download kernel
    2. Reboot to recovery
    3. Backup
    *if you come from faux or tbalden please reflash your rom, you don''t need a wipe. They have some trickery with their own hotplugs so they have reomved mpdecision, so without reflashing your phone will be stuck on quad core*
    4. Flash kernel
    5. Reboot
    *Please let it sit for ~10 mins before judging as it will need to settle down. Do not judge battery life unless you've run it through at least 3+ cycles after flashing*
    Click to expand...
    Click to collapse

    DISCLAIMERS:
    1) Excessive under/overvolting can cause damage to the internal parts of your phone. It can cause the phone's lifespan to deteriorate quickly so make sure you know boundaries
    2) If you OC'ed/OV'ed/UV'ed and now are in a boot loop or your phone is slow, or acting weird or just died - it means your phone does not like the changes. Voltage control is for those who know what the risks are, know what they are doing and know how to fix it. Perhaps when you change these, test them for an hour or so before setting on boot
    3) Any changes to the phone, including flashing the kernel itself is done at your own risk.
    4) Every phone is differentt
    5) Please DO NOT use any task killers (or additional init.d scripts), they DO NOT improve performance/battery. The kernel is designed to be used with stock settings: my settings and anything else can cause problems
    Click to expand...
    Click to collapse

    Download latest version:
    http://xdaforums.com/showthread.php?t=2768254&page=32#post55839871

    If you want logo options:
    http://xdaforums.com/showpost.php?p=54759427&postcount=5140

    All previous versions:
    Can now be found in changelog posts
    Older versions:
    http://www.androidfilehost.com/?w=files&flid=15247

    Sauce: https://github.com/poondog/kangaroo-m7-mkIV

    *Please do not re upload without permission, I like to count downloads*
    Changelog will be in post 2

    What is not coming via requests:
    - replacement for mpdecision
    - additional useless governors
    - aroma
    - anything useless: this kernel was made for me and I shared it with a few people until now
    Click to expand...
    Click to collapse

    With thanks to (in no particular order)
    franciscofranco
    faux123
    anarkia1976/ak
    imoseyon
    flar2
    teamseven
    n3ocort3x
    morfic
    maxwen
    labodj
    ktoonsez
    tpbklake
    janarp
    szezso
    Synthetic.Nightmare
    ZeroInfinity
    Sabermod team
    Stratosk
    Lens-F
    Savoca
    tiny4579
    neniick
    geko95gek
    brymaster
    buckmarble
    Click to expand...
    Click to collapse

    XDA:DevDB Information
    Kangaroo Kernel, Kernel for the Verizon HTC One

    Contributors
    poondog
    Kernel Special Features:

    Version Information
    Status: Stable

    Created 2014-05-31
    Last Updated 2014-10-05
    View
    12
    poondog
    poondog
    UPDATE: 14 September

    Condensed/translated Changelog: Kangaroo Kernel 4.4.3 v058
    .shorten some wakelocks (for better sleeping)
    .some more tweaks for better overall battery and performance
    .upstream fixes and improvements to TOPOLOGY drivers (to do with queuing of tasks to individual cores, from Linaro)
    .optimize TOPOLOGY further for Krait devices (neobuddy89)
    .introduce BFQ IOSCHED v7r5
    .introduce FIOPS IOSCHED (default)
    .enable ILLINOIS to be set as default, and set it default for TCP
    .introduce LINARO POWER EFFICIENT WORKQUEUES (heaps of battery savings)
    .some fixes to bluetooth
    .upstream backports to EVENTPOLL subsystem from 3.14+ (Efficient event retrieval implementation, thx neobuddy)
    .fix some more compile warnings for better optimized code (cl3kener)
    .drop i2c clock a bit more (to save battery, ref. @zeroblade1984/Samsung/OPPO
    .remove some more debug (less energy waste)
    .strip most debug from modules, reduce size by 90%
    .enable all sleep states, based on 4.4.2 kernel tests (for better battery especially in sleep)
    .remove depreciated sched sysfs since they are removed up upstream kernels and don't really do anything
    .some LINARO ARM code improvements
    .update to HTC 6.09.401.5 kernel source (not much really lol, a few internet related fixes)
    .introduce USB FAST CHARGE (flar2, off by default)
    .cleanup some HTC junk
    .some improvements to LOWMEMORYKILLER from moto/upstream (to do with multitasking)
    .some improvements to ONDEMAND governor for better power savings with no smoothness loss
    .fixed USB DAC audio (Thanks @hq5 for testing)
    .introduce KEXEC_HARDBOOT aka MULTIROM (still needs patched RAMDISK, I do not include this)
    .add some more doge
    .fixed compass etc for Sprint and Verizon (thx brymaster5000)

    Changelog from v045:
    .2w: change feedback slightly, now it doesn't vibrate heaps on sweep2sleep
    .further updated workqueue (for better battery and efficiency)
    .some network tweaks
    .heaps of "under the hood" improvements (lead to better efficiency for speed/battery)
    .reduce internal USB wakelocks to improve battery
    .patched kernel futex vulnerability (from CM, for security)
    .compress with LZ4 for faster boot speeds and built with latest binary
    .updated all compressors to latest versions
    .add lower GPU frequencies and scale to them (for better battery)
    .upstream backports to OOM handler from 3.16+ (related to memory use, see here https://www.kernel.org/doc/gorman/ht...rstand016.html)
    .optimized deadline scheduler for better use on flash
    .quad core cpu stats (faux123)
    .lower max i2c clock to 384MHz to save some juice, no negative effects in testing
    .some general code fixes and improvements
    .enable ILLINOIS tcp algorithm and set default
    .enable NFTS r/w support (showp1984)
    Click to expand...
    Click to collapse

    Full Changeloge:
    Project name: Kangaroo Kernel MKIV
    poondog - BUILD: v058
    poondog - KEXEC: make some comments prettier
    poondog - DEFCONFIG: missed this change in the merge of 6.09 sources
    poondog - DEFCONFIG: support for KEXEC_HARDBOOT aka MULTIROM Needs RAMDISK support still
    Vojtech Bocek - Implement kexec-hardboot
    poondog - BUILD: TEST version for sound test
    poondog - SOUND: USB: fix USB DAC functionality once and for all Massive thanks to hq5 for help in testing kernels
    poondog - BUILD: v057
    Tingwei Zhang - cpufreq: ondemand: Boost CPU frequency only for touch input
    poondog - CPUFREQ: cleanup ondemand a bit
    Swetha Chikkaboraiah - cpufreq: ondemand:kernel NULL pointer dereference at dbs_check_cpu
    Srivatsa Vaddagiri - cpufreq: ondemand: Remove unnecessary synchronization attempt
    Jingoo Han - cpufreq: Remove unnecessary initialization of a local variable
    Viresh Kumar - cpufreq: Fix sparse warning by making local function static
    Viresh Kumar - cpufreq: return early from __cpufreq_driver_getavg()
    Weijie Yang - staging: android: lowmemorykiller: set TIF_MEMDIE before send kill sig
    Yi-wei Zhao - staging: android: lowmemorykiller: select a new task to kill
    Colin Cross - lowmemorykiller: make default lowmemorykiller debug message useful
    Jeffrey Carlyle - trace: allow memkill tracing to be selectively enabled
    Jeffrey Carlyle - trace: memkill: trace LMK and OOM kills
    Liam Mark - staging: android: lowmemorykiller: Reduce debug_level to 1
    poondog - BOARD: missed this in cleanup
    poondog - BOARD: cleanup some leftover unneeded stuff from HTC
    poondog - LIB: remove DEBUG_BUGVERBOSE
    poondog - BUILD: v056
    flar2 - USB Fast Charge
    poondog - MISC: push HTC source m7ul-3.4.10-g40b189c 6.09.401.5 WWE base Seems all HTC did was delete more comments lel
    poondog - TIME: fix instance of creating something out of nothing
    poondog - BUILD: v055
    poondog - DEFCONFIG: remove SCHED_MC and SCHED_SMT, they are depreciated
    Olof Johansson - ARM: dt: Only print warning, not WARN() on bad cpu map in device tree
    Lorenzo Pieralisi - ARM: 7763/1: kernel: fix __cpu_logical_map default initialization
    Lorenzo Pieralisi - ARM: kernel: add logical mappings look-up
    Lorenzo Pieralisi - ARM: 7762/1: kernel: fix arm_dt_init_cpu_maps() to skip non-cpu nodes
    Lorenzo Pieralisi - ARM: 7585/1: kernel: fix nr_cpu_ids check in DT logical map init
    Lorenzo Pieralisi - ARM: kernel: add device tree init map function
    Vincent Guittot - sched: cpu_power: enable ARCH_POWER
    poondog - NET: allow to set ILLINOIS as default TCP algo
    Thomas Renninger - cpupower: Remove mc and smt power aware scheduler info/settings
    Hans de Goede - cgroup: Fix use after free of cgrp (cgrp->css_sets)
    Hans de Goede - cgroup: Take css_set_lock from cgroup_css_sets_empty()
    poondog - DEFCONFIG: default tcp illinois
    poondog - BUILD: v054
    Rafael J. Wysocki - Revert "epoll: use freezable blocking call"
    poondog - PM: enable all sleep states. Tests have shown that this will increase battery life particularly with screen off. Tests conducted with users. Thanks to those who contributed
    poondog - DEBUG: disable some more debug referenced here https://github.com/franciscofranco/hammerhead/commit/f0a852f51c0530bb02a9eb082cc1ea3a95517c39
    poondog - MSM: remove some debug
    poondog - BOARD: drop i2c gsbi1 clock rate even lower Based on @zeroblade1984 research
    Andrew Bresticker - CHROMIUM: clocksource: allow clocksources to be selected earlier during boot
    Yinghai Lu - initrd: fix lz4 decompress with initrd
    David Decotigny - lib: cpu_rmap: avoid flushing all workqueues
    Chet Kener - net: Fix Warnings
    Chet Kener - arch:arm: Fix Frame pointer and Arm Unwind Warnings
    poondog - BUILD: v053
    Pranav Vashi - Add new helpers
    Jason Baron - epoll: do not take the nested ep->mtx on EPOLL_CTL_DEL
    Amit Pundir - epoll: drop EPOLLWAKEUP if PM_SLEEP is disabled
    Jason Baron - epoll: do not take global 'epmutex' for simple topologies
    Jason Baron - epoll: optimize EPOLL_CTL_DEL using rcu
    Eric Dumazet - epoll: add a reschedule point in ep_free()
    Al Viro - switch epoll_ctl() to fdget
    Pranav Vashi - signals: eventpoll: do not use sigprocmask()
    Eric Wong - epoll: cleanup: use RCU_INIT_POINTER when nulling
    Eric Wong - epoll: cleanup: hoist out f_op->poll calls
    Eric Wong - epoll: lock ep->mtx in ep_free to silence lockdep
    poondog - BUILD: v052
    Eric Wong - epoll: use RCU to protect wakeup_source in epitem
    Eric Wong - epoll: trim epitem by one cache line
    Pranav Vashi - epoll: switch simple cases of fget_light to fdget
    Al Viro - switch epoll_wait(2) to fget_light()
    Al Viro - eventpoll: use-after-possible-free in epoll_create1()
    Michael Kerrisk - PM: Rename CAP_EPOLLWAKEUP to CAP_BLOCK_SUSPEND
    Rafael J. Wysocki - epoll: Fix user space breakage related to EPOLLWAKEUP
    Arve Hjønnevåg - epoll: Add a flag, EPOLLWAKEUP, to prevent suspend while epoll events are ready
    poondog - BUILD: new toolchain is 4.9.2
    poondog - BUILD: new toolchain 2014.08 from Christopher83 for a15
    poondog - BUILD: v051
    Pranav Vashi - Bluetooth: Prevent Bluetooth L2CAP Remote Heap Memory Corruption
    poondog - DEFCONFIG: enable Linaro POWER EFFICIENT WORKQUEUES
    Mark Brown - ASoC: jack: Use power efficient workqueue
    Mark Brown - regulator: core: Use the power efficient workqueue for delayed powerdown
    Mark Brown - ASoC: pcm: Use the power efficient workqueue for delayed powerdown
    Viresh Kumar - PHYLIB: queue work on system_power_efficient_wq
    Viresh Kumar - block: queue work on power efficient wq
    Viresh Kumar - workqueue: Add system wide power_efficient workqueues
    Viresh Kumar - workqueues: Introduce new flag WQ_POWER_EFFICIENT for power oriented workqueues
    poondog - BUILD: v050
    poondog - DEFCONFIG: disable GPIO debug
    poondog - DEFCONFIG: disable some more debugging
    poondog - DEFCONFIG: disable USERSPACE governor
    poondog - DEFCONFIG: enable automatic process group scheduling
    poondog - DEFCONFIG: enable ILLINOIS TCP congestion algorithm
    poondog - BUILD: v049
    poondog - DEFCONFIG: set FIOPS as default scheduler
    poondog - DEFCONFIG: support for FIOPS IOSCHED
    Shaohua Li - block: fiops add some trace information
    Shaohua Li - block: fiops bias sync workload
    Shaohua Li - block: fiops preserve vios key for deep queue depth workload
    Shaohua Li - block: fiops add ioprio support
    Shaohua Li - block: fiops sync/async scale
    Shaohua Li - block: fiops read/write request scale
    Shaohua Li - block: fiops ioscheduler core
    poondog - BUILD: v048
    poondog - DEFCONFIG: support for BFQ IOSCHED
    Mauro Andreolini - block, bfq: add Early Queue Merge (EQM) to BFQ-v7r5 for 3.4.0
    Paolo Valente - block: introduce the BFQ-v7r5 I/O sched for 3.4
    Arianna Avanzini - block: cgroups, kconfig, build bits for BFQ-v7r2-3.4
    poondog - BUILD: v047
    Pranav Vashi - ARM: topology: Suppress clock-frequency error
    Pranav Vashi - ARM: topology: Add efficiency coefficient for Krait
    Mark Brown - ARM: 8008/1: topology: Coding style fixes
    Mark Brown - ARM: 7920/1: topology: Staticise non-exported symbols
    Sudeep KarkadaNagesha - ARM: topology: remove hwid/MPIDR dependency from cpu_capacity
    Zhang Yanfei - arm: remove cast for kzalloc return value
    Lorenzo Pieralisi - ARM: kernel: update topology to use new MPIDR macros
    Lorenzo Pieralisi - ARM: kernel: enhance MPIDR macro definitions
    Venkatraman Sathiyamoorthy - ARM: 7482/1: topology: fix section mismatch warning for init_cpu_topology
    Vincent Guittot - ARM: 7463/1: topology: Update cpu_power according to DT information
    Vincent Guittot - ARM: 7462/1: topology: factorize the update of sibling masks
    Vincent Guittot - ARM: 7461/1: topology: Add arch_scale_freq_power function
    poondog - BUILD: v046
    Sankara Muthukrishnan - irq: Set CPU affinity right on thread creation
    Joonsoo Kim - mm, vmalloc: remove useless variable in vmap_block
    Joonsoo Kim - mm, vmalloc: use well-defined find_last_bit() func
    Jan Kara - block: Reserve only one queue tag for sync IO if only 3 tags are available
    poondog - MKCOMPILE: such name so kernel very doge wow
    Katish Paran - diag: dci: Safeguard to prevent Integer Underflow and Memory Leak
    Andrew Ruder - fs/super.c: sync ro remount after blocking writers
    poondog - BCMDHD: further reduce wakelock time
    poondog - BCMDHD: reduce rx wakelock time by more experimental
    imoseyon - bcmdhd: tweaked wakelocks a bit
    Peter Hurley - n_tty: Fix n_tty_write crash when echoing in raw mode
    TripNRaVeR - suspend: shorten time to enter sleep state
    TripNRaVeR - mmc: decrease wakelock timeout
    poondog - BUILD: v045
    poondog - Revert "ARM: be strict about FP exceptions in kernel mode"
    poondog - Revert "ARM: add support for bit sliced AES using NEON instructions"
    poondog - Revert "crypto: create generic version of ablk_helper"
    poondog - Revert "crypto: ablk_helper - Replace memcpy with struct assignment"
    poondog - Revert "ARM: move AES typedefs and function prototypes to separate header"
    poondog - Revert "ARM: pull in <asm/simd.h> from asm-generic"
    poondog - Revert "ARM: 7626/1: arm/crypto: Make asm SHA-1 and AES code Thumb-2 compatible"
    poondog - Revert "ARM: 7837/3: fix Thumb-2 bug in AES assembler code"
    poondog - Revert "ARM: 7723/1: crypto: sha1-armv4-large.S: fix SP handling"
    poondog - Revert "defconfig: Enable NEON optimized crypto routines"
    poondog - Revert "arm: vfpmodule: Fix warning procfs vfp_bounce reporting failed"
    poondog - Revert "ARCH: ARM: VFP: fix build"
    poondog - Revert "DEFCONFIG: support for new crypto"
    poondog - BUILD: v044
    poondog - M7: i2c: change max i2c clock to 384KHz Devices support upto 400KHz but QUP I2C is recommended to use 384KHz with the following message.
    Sergey Senozhatsky - zlib: clean up some dead code
    poondog - FS: namespace: initialize variables
    Stratos Karafotis - cpufreq: Break out early if freq equals to target_freq
    Viresh Kumar - cpufreq: Optimize cpufreq_frequency_table_verify()
    poondog - DEFCONFIG: enable quad core cpu stats
    faux123 - overall_stats: forward port to kernel 3.4+
    faux123 - overall_stats: make dual/quad core stats configurable via meunconfig
    Paul Reioux - overall_stats: add overall stats for all available cores
    Archana Sathyakumar - cpufreq: Resolve CPUFREQ_NOTIFY issue
    poondog - BUILD: v043
    Srivatsa Vaddagiri - cpufreq: Retain only online cpus in managed_policy->cpus
    mrg666 - af_unix: speedup /proc/net/unix
    myfluxi - block: deadline: Optimize for non-rotational
    xiaobing tu - deadline: Allow 0ms deadline latency, increase the read speed
    Joe Perches - jiffies conversions: Use compile time constants when possible
    Schichan Nicolas - ARM: 7597/1: net: bpf_jit_32: fix kzalloc gfp/size mismatch.
    Schichan Nicolas - ARM: 7598/1: net: bpf_jit_32: fix sp-relative load/stores offsets.
    Nicolas Schichan - ARM: net: bpf_jit: fix emit_swap16() for non ARMv6+.
    poondog - BUILD: v042
    Sasha Levin - mm/ksm.c: use new hashtable implementation
    Sasha Levin - block,elevator: use new hashtable implementation
    Tejun Heo - workqueue: fix find_worker_executing_work() brekage from hashtable conversion
    Sasha Levin - workqueue: use new hashtable implementation
    Sasha Levin - hashtable: introduce a small and naive hashtable
    poondog - FS: derp from previous commit
    poondog - BUILD: v041
    Will Deacon - asm-generic: rwsem: de-PPCify rwsem.h
    Will Deacon - asm-generic: rwsem: ensure sem->cnt is only accessed via atomic_long_*
    Linus Torvalds - Merge branch 'rwsem-optimizations'
    Alex Shi - rwsem: Implement writer lock-stealing for better scalability
    poondog - SCHED: LOAD_FREQ (4*HZ+61) avoids loadavg Moire LOAD_FREQ is (5*HZ+1) to avoid high load average when idle: http://kerneltrap.org/mailarchive/linux-kernel/2007/10/3/328568
    Jeff Liu - binfmt_elf.c: use get_random_int() to fix entropy depleting
    Sana Venkat Raju - msm: buspm: Correct size type in buspm_xfer_req
    Alok Chauhan - msm: msm_bus: Fix the type error causing bandwidth overflow
    ChandanaKishori Chiluveru - usb: gadget: Fix synchronization issue between f_audio_source
    poondog - BUILD: v040
    Pranav Vashi - staging:android:lmk: Use for_each_thread helper
    Oleg Nesterov - introduce for_each_thread() to replace the buggy while_each_thread()
    David Rientjes - mm, oom: base root bonus on current usage
    David Rientjes - mm, oom: prefer thread group leaders for display purposes
    Oleg Nesterov - oom_kill: add rcu_read_lock() into find_lock_task_mm()
    Oleg Nesterov - oom_kill: has_intersects_mems_allowed() needs rcu_read_lock()
    Oleg Nesterov - oom_kill: change oom_kill.c to use for_each_thread()
    Rusty Russell - mm/oom_kill: remove weird use of ERR_PTR()/PTR_ERR().
    David Rientjes - mm, oom: remove redundant sleep in pagefault oom handler
    David Rientjes - mm, oom: cleanup pagefault oom handler
    Lai Jiangshan - oom: use N_MEMORY instead N_HIGH_MEMORY
    David Rientjes - mm, oom: allow exiting threads to have access to memory reserves
    poondog - BUILD: v039
    David Rientjes - mm, memcg: move all oom handling to memcontrol.c
    David Rientjes - mm, oom: reduce dependency on tasklist_lock
    David Rientjes - mm, memcg: introduce own oom handler to iterate only over its own threads
    David Rientjes - mm, oom: introduce helper function to process threads during scan
    David Rientjes - mm, oom: move declaration for mem_cgroup_out_of_memory to oom.h
    David Rientjes - mm, oom: fix potential killing of thread that is disabled from oom killing
    David Rientjes - mm, oom: do not schedule if current has been killed
    David Rientjes - mm, oom: fix and cleanup oom score calculations
    David Rientjes - mm, oom: fix badness score underflow
    David Rientjes - mm, oom: normalize oom scores to oom_score_adj scale only for userspace
    poondog - BUILD: v038
    poondog - mm: patch the following to stock google msm kernel: memory controller out of memory
    Igor Kovalenko - reduce memory usage by input driver
    Chris Fries - msm: acpuclock-krait: Init to max speed
    Yuanyuan Zhong - msm: rng: pass correct size to kzalloc
    Alexander Filonenko - msm: bam_dmux: optimize UL inactivity timeout
    Jayachandran C - Delay UL powerdown if BAM is in polling mode
    Alexander Filonenko - msm: bam_dmux: UL inactivity timeout is too high
    poondog - BUILD: v037
    poondog - SLUB: fix uninitialized variable
    anarkia1976 - slub.c: Remove PAGE_ALLOC_COSTLY_ORDER - +3% boost
    myfluxi - msm: kgsl: Fix policy sysfs implementation
    myfluxi - msm: kgsl: report gpu frequency correctly in sysfs
    Steve Kondik - m7: Resync GPU configuration with upstream
    poondog - BUILD: v036
    Greg Kroah-Hartman - lz4: fix another possible overrun
    Lasse Collin - lib/xz: enable all filters by default in Kconfig
    Lasse Collin - lib/xz: add comments for the intentionally missing break statements
    Rashika Kheria - lib/decompress_inflate.c: include appropriate header file
    Alexandre Courbot - lib/decompressors: fix "no limit" output buffer length
    poondog - BUILD: v035
    Paul Bolle - decompressors: fix typo "POWERPC"
    Florian Fainelli - decompressors: make the default XZ_DEC_* config match the selected architecture
    Florian Fainelli - decompressors: drop dependency on CONFIG_EXPERT
    Florian Fainelli - decompressors: group XZ_DEC_* symbols under an if XZ_BCJ / endif
    Markus F.X.J. Oberhumer - lib/lzo: Optimize code for CPUs with inefficient unaligned access
    Greg Kroah-Hartman - lzo: properly check for overruns
    Markus F.X.J. Oberhumer - lib/lzo: Update LZO compression to current upstream version
    Markus F.X.J. Oberhumer - lib/lzo: Rename lzo1x_decompress.c to lzo1x_decompress_safe.c
    Greg Kroah-Hartman - lz4: add overrun checks to lz4_uncompress_unknownoutputsize()
    Greg Kroah-Hartman - lz4: ensure length does not wrap
    poondog - BUILD: v034
    Chanho Min - lib: add lz4 compressor module
    Jan Beulich - lib/decompress_unlz4.c: always set an error return code on failures
    Sunil Khatri - genalloc: Correct nbytes calculation on long boundary
    Jan Kara - lib/flex_proportions.c: fix corruption of denominator in flexible proportions
    Jan Kara - lib: Fix possible deadlock in flexible proportion code
    Jan Kara - lib: Proportions with flexible period
    poondog - DEFCONFIG: update for LZ4 compression
    Dennis Rassmann - defconfig: enable NTFS r/w support
    P J P - initramfs: read CONFIG_RD_ variables for initramfs compression
    Sergey Senozhatsky - lz4: fix compression/decompression signedness mismatch
    Richard Laager - lib/lz4: correct the LZ4 license
    Markus Trippelsdorf - .gitignore: ignore *.lz4 files
    Chanho Min - crypto: add lz4 Cryptographic API
    Chanho Min - lib: add lz4 compressor module
    Kyungsik Lee - arm: add support for LZ4-compressed kernel
    Kyungsik Lee - lib: add support for LZ4-compressed kernel
    Kyungsik Lee - decompressor: add LZ4 decompressor module
    poondog - FUTEX: remove whitespace errors
    Thomas Gleixner - futex: Make lookup_pi_state more robust
    Thomas Gleixner - futex: Always cleanup owner tid in unlock_pi
    Thomas Gleixner - futex: Validate atomic acquisition in futex_lock_pi_atomic()
    Thomas Gleixner - futex-prevent-requeue-pi-on-same-futex.patch futex: Forbid uaddr == uaddr2 in futex_requeue(..., requeue_pi=1)
    Thomas Gleixner - futex: Add another early deadlock detection check
    Thomas Gleixner - futex: Prevent attaching to kernel threads
    poondog - BUILD: v032
    Peter Zijlstra - sched/fair: Optimize cgroup pick_next_task_fair()
    Peter Zijlstra - sched/fair: Optimize find_busiest_queue()
    Pavankumar Kondeti - USB: Prevent system suspend when HSIC device is active
    Pavankumar Kondeti - USB: Allow skipping device resume during system resume
    motley - drivers:usb decrease PHY_SUSPEND_TIMEOUT_USEC to 200 (from 500) to hopefully reduce the msm_hsic_host wakelock duration and yield more deep sleep.
    Ingo Molnar - nohz: Reduce overhead under high-freq idling patterns
    Davidlohr Bueso - lib/int_sqrt.c: optimize square root algorithm
    poondog - BUILD: v030
    poondog - TCP: remove whitespace errors
    anarkia1976 - tcp: network speed tweak
    anarkia1976 - tcp: add auto corking feature
    anarkia1976 - tcp: fix conflicting types for tcp_write_xmit
    Yuchung Cheng - tcp: early retransmit
    Eric Dumazet - tcp: tcp small queues
    poondog - gitignore: ignore hz.bc
    H. Peter Anvin - kernel: Replace timeconst.pl with a bc script
    poondog - BUILD: v030
    Paul Reioux - arch/arm/kernel/armksyms: fix merge derp
    faux123 - ARM: 7593/1: nommu: do not enable DCACHE_WORD_ACCESS when !CONFIG_MMU
    Will Deacon - ARM: 7592/1: nommu: prevent generation of kernel unaligned memory accesses
    Rob Herring - ARM: 7492/1: add strstr declaration for decompressors
    faux123 - ARM: 7493/1: use generic unaligned.h
    Will Deacon - ARM: dcache: select DCACHE_WORD_ACCESS for little-endian ARMv6+ CPUs
    faux123 - ARM: use generic strnlen_user and strncpy_from_user functions
    Paul Mackerras - lib: Fix generic strnlen_user for 32-bit big-endian machines
    Linus Torvalds - lib: add generic strnlen_user() function
    Linus Torvalds - word-at-a-time: make the interfaces truly generic
    faux123 - lib: Sparc's strncpy_from_user is generic enough, move under lib/
    faux123 - kernel: Move REPEAT_BYTE definition into linux/kernel.h
    poondog - BUILD: v029
    Tejun Heo - workqueue: cond_resched() after processing each work item
    Lai Jiangshan - workqueue: ensure @task is valid across kthread_stop()
    Tejun Heo - workqueue: consider work function when searching for busy work items
    Tejun Heo - workqueue: convert BUG_ON()s in __queue_delayed_work() to WARN_ON_ONCE()s
    Mike Galbraith - workqueue: exit rescuer_thread() as TASK_RUNNING
    Tejun Heo - workqueue: add missing smp_wmb() in process_one_work()
    poondog - MM: remove redundant statement
    poondog - 2w: don't vibrate in response to sweep when screen is on, there is enough vibration from button feedback
    poondog - INCLUDE: add back some comments
    poondog - DEFCONFIG: support for new crypto
    poondog - BUILD: v028
    poondog - ARCH: ARM: VFP: fix build
    myfluxi - arm: vfpmodule: Fix warning procfs vfp_bounce reporting failed
    myfluxi - defconfig: Enable NEON optimized crypto routines
    Russell King - ARM: add .gitignore entry for aesbs-core.S
    Ard Biesheuvel - ARM: 7723/1: crypto: sha1-armv4-large.S: fix SP handling
    Ard Biesheuvel - ARM: 7837/3: fix Thumb-2 bug in AES assembler code
    Dave Martin - ARM: 7626/1: arm/crypto: Make asm SHA-1 and AES code Thumb-2 compatible
    Ard Biesheuvel - ARM: pull in <asm/simd.h> from asm-generic
    Ard Biesheuvel - ARM: move AES typedefs and function prototypes to separate header
    kbuild test robot - crypto: ablk_helper - Replace memcpy with struct assignment
    Ard Biesheuvel - crypto: create generic version of ablk_helper
    Ard Biesheuvel - ARM: add support for bit sliced AES using NEON instructions
    poondog - BUILD: v027
    Click to expand...
    Click to collapse
    Sorry, another big update :D
    Big thanks this time to @hq5 for his (I am guessing?) time testing out USB DAC functionality. Finally this is fixed (something that I have not been able to test myself) but with his help testing (I don't have a USB DAC) it could be narrowed down and fixed. Make sure to hit his thanks button a few times, let's double hq5's meter!

    Download is attached.

    COMING UP:
    .tests for audio
    .some more tests mystery!    gsm version
    .additional governors
    .thermal driver kernel side
    View
    10
    poondog
    poondog
    UPDATE TIME! Finally


    ^ For all the haters out there

    4 October
    Changelog: Kangaroo Kernel MKIV v068
    .Update toolchain to Linaro 2014.09 4.9.2 for a15 chipsets by Christopher83
    .some updates to SELINUX (to do with security)
    .introduce OPTIMAX governor (see below)
    .introduce PRESERVATIVE governor (see below)
    .introduce SMARTMAX governor (see below)
    .GPU overclocking (see below)
    .imported ULTRASOUND_B sound drivers (see below)
    .upstream Linux update to 3.4.104
    .allow stock EXFAT and CRYPTO modules to work (thx to bgcngm from dual sim m7, this is experiemental since idk if it works)
    .fixed random reboots and stability from prev version
    Click to expand...
    Click to collapse

    New things
    Optimax governor
    This is based on ONDEMAND, like almost all governors that have arisen from XDA. It contains some enhancements from LG, particularly to freq boost handling so it will boost to a set level, almost like HTC's governor. It has different tunables to the HTC governor but it behaves pretty similar, the tunables it comes with default are a bit more conservative.
    I picked it from Cl3kener's Uber kernel for Nexus 5, where it has quite a reputation for battery life
    I believe it originates from the LG G2 (correct me if I'm wrong)

    Preservative governor
    This is based on the idea that the CPU will consume a lot of power when it changes frequency. It is based on the conservative governor. The idea is that it will stay at the step specified (702MHz selected by the creator Bedalus) unless needed. You will notice it will hover around 702 a lot, and not go above too much, and only to min freq when NOTHING is happening at all. This is most beneficial when you are doing something like reading; the screen is static or playing light games that won't need boosting any more
    The governor comes from Moob kernel for nexus 4

    Smartmax governor
    Does it need an explanation? Its basically an updated quad core version of Smartass V2, the benefit of this over any other gov is that its highly tunable. Because we have Ondemand for performance and smoothness, its tuned for battery with @Rad Ryan's findings
    The author is maxwen, the gov was originally written for One V, One X and original OPPO phones

    GPU Overclock
    This is different to any other kernel I believe. Instead of replacing the max freq qith a higher one, I added another freq step and allowed the kernel to scale to it. Therefore when the GPU speed is not really needed it can still sit at 320 or 400MHz, only boosting to 487.5 max when needed. The GPU is very conservative with its boosting. It will like to boost to medium freqs a lot. The default freqs now are 27, 128, 200, 320, 400, 487.5MHz. Max freq is able to be set with an app like TricksterMOD. But, since it will not boost to 487.5MHz very often the battery impact is negligible under "normal" use.
    Thanks goes to @Brymaster5000 and @Buckmarble from VZW forums for some help on this :) Plus to faux123 where the original extra freq comes from

    ULTRASOUND_B
    These drivers actually are made for the Snapdragon 80x devices. You guys picked it from the blind testing. The tests were:
    a: ULTRASOUND_A (made for our devices but newer than whats stock)
    b: ULTRASOUND_B (made for m8, z2, SGS5 generation devices)
    c: CONTROL (completely stock)
    See back a few pages for the discussion on this

    In short new sound drivers for extra eargasm experience :D :D :D
    Click to expand...
    Click to collapse

    Download is attached
    Click to expand...
    Click to collapse

    About upstream Linux:
    Patches come from kernel.org for the mainline 3.4.y kernel. I merge them into the m7 kernel
    The maintainer has changed a month and a bit ago. There used to be about weekly updates from Greg KH, but now the new maintainer left it a while to update lol
    These updates consist of a number of patches that are picked to improve the kernel and fix bugs. Usually new features are not included, they are saved for new linux versions
    Here is the changelog for 3.4.104:
    Code: 
    bb4a05a Linux 3.4.104
618dea4 alpha: add io{read,write}{16,32}be functions
3a8f613 ext2: Fix fs corruption in ext2_get_xip_mem()
f79bb94 alpha: Fix fall-out from disintegrating asm/system.h
92a6e26 UniCore32-bugfix: fix mismatch return value of __xchg_bad_pointer
4e36063 UniCore32-bugfix: Remove definitions in asm/bug.h to solve difference between native and cross compiler
62cdcce unicore32: select generic atomic64_t support
82a938ab 8250_pci: fix warnings in backport of Broadcom TruManage support
69db2d4 slab/mempolicy: always use local policy from interrupt context
ab22539 openrisc: add missing header inclusion
c1bc007 MIPS: Fix accessing to per-cpu data when flushing the cache
9a8fa93 MIPS: perf: Fix build error caused by unused counters_per_cpu_to_total()
c804743 USB: serial: fix potential heap buffer overflow
62148f7 USB: serial: fix potential stack buffer overflow
2f3e285 ARM: 8129/1: errata: work around Cortex-A15 erratum 830321 using dummy strex
5624bb3 ARM: 8128/1: abort: don't clear the exclusive monitors
e78c127 HID: picolcd: sanity check report size in raw_event() callback
e115f02 HID: magicmouse: sanity check report size in raw_event() callback
0bfb000 NFSv4: Fix problems with close in the presence of a delegation
818ee41 USB: sisusb: add device id for Magic Control USB video
86d165e HID: logitech-dj: prevent false errors to be shown
2f2c704 USB: whiteheat: Added bounds checking for bulk command response
1d37c3e usb: xhci: amd chipset also needs short TX quirk
42494f0 xhci: Treat not finding the event_seg on COMP_STOP the same as COMP_STOP_INVAL
d232e2c USB: ftdi_sio: Added PID for new ekey device
68e9929 USB: serial: pl2303: add device id for ztek device
262548e USB: ftdi_sio: add Basic Micro ATOM Nano USB2Serial PID
78aea3a USB: option: add VIA Telecom CDS7 chipset device id
c2aa9b7 staging: et131x: Fix errors caused by phydev->addr accesses before initialisation
f40751b xtensa: fix a6 and a7 handling in fast_syscall_xtensa
bcf20fd xtensa: fix TLBTEMP_BASE_2 region handling in fast_second_level_miss
fb2ae73 xtensa: fix address checks in dma_{alloc,free}_coherent
2297927 xtensa: replace IOCTL code definitions with constants
483320c ALSA: hda/realtek - Avoid setting wrong COEF on ALC269 & co
d06e4b0 CIFS: Fix wrong directory attributes after rename
a0e5b9d md/raid6: avoid data corruption during recovery of double-degraded RAID6
77e5657 iommu/amd: Fix cleanup_domain for mass device removal
db83744 pata_scc: propagate return value of scc_wait_after_reset
f0634a3 kvm: iommu: fix the third parameter of kvm_iommu_put_pages (CVE-2014-3601)
33df36f MIPS: OCTEON: make get_system_type() thread-safe
5ccd3e2 isofs: Fix unbounded recursion when processing relocated directories
c945ed6 HID: logitech: perform bounds checking on device_id early enough
dab2f9b HID: fix a couple of off-by-ones
9220628 ibmveth: Fix endian issues with rx_no_buffer statistic
c27d3b5 ASoC: pxa-ssp: drop SNDRV_PCM_FMTBIT_S24_LE
b402b72 stable_kernel_rules: Add pointer to netdev-FAQ for network patches
    Click to expand...
    Click to collapse

    That is all. Thanks for sticking by the Kangaroo :) Make sure you flash the correct version
    View
    9
    poondog
    poondog
    UPDATE! 23 August

    Changelog: Kangaroo Kernel v138 (4.4.2)
    .patch to upstream 3.4.103
    .overclock the GPU to 487MHz. I added an extra step, not replaced what we had (see below)
    .disable some more debug (less wastage of energy for greater battery life)
    .disable F2FS, who even knew it was here anyway
    .updates to kernel processes from CAF/upstream
    .lower max i2c clock to 384KHz, to save battery
    .minor updates to ONDEMAND gov from tiny4579/stratosk
    .import ULTRASOUND audio drivers from latest CAF. We are using ULTRASOUND_A, made for msm8960 (improves audio playback)

    Changelog from v132 that you skipped sorry
    .cleaned up touchscreen drivers and removed whitespaces
    .build version is changed slightly to show kangaroo kernel mkiii, to differentiate between 4.4.3
    .cleaned up thermal drivers since last release
    .some changes to thermal algorithm to be more reliable and more accurate
    .tuned thermal for a bit cooler (see below)
    .remove some debugging in sched (less wasted energy)
    .misc optimizations/updates for compression/decompression (more reliable, ref. cl3kener git)
    .misc optimizations/updates for CPUFREQ
    .updated RANDOM32 driver from Linux 3.16.y kernel
    .updated RANDOM driver from Linux 3.16.y kernel (see below, ref. dorimanx git)
    .ported new random driver for use on m7 s600 (me)
    .cleaned up vibrator driver (more optimized and usable code now)
    .lower default 2w vibration to 18, down from 20
    .SMARTMAX: use new relation_C to better choose frequencies
    .PRESERVATIVE: use new relation_C to better choose frequencies
    .upstream improvements to kernel sync drivers (effects general operations, and io related things)
    .upstream improvements to sched (will effect general performance and battery)
    .updated LZ4 kernel compression binary to r120
    .merge upstream Linux kernel 3.4.102
    Click to expand...
    Click to collapse

    THERMAL:

    New thermal defaults are as follows:

    #define DEFAULT_TEMP_MIN 43
    #define DEFAULT_TEMP_MID 54
    #define DEFAULT_TEMP_MAX 65
    Click to expand...
    Click to collapse

    RANDOM:
    What is random number generation?
    https://wiki.archlinux.org/index.php/Random_number_generation

    basically it will effect overall smoothness and performance. In testing we found the kernel to be smoother and faster, while battery was slightly improved
    Click to expand...
    Click to collapse

    GPU Overclock:
    Should really smooth out games. GPU max freq is 487MHz BUT it will ONLY reach it in extreme circumstances, not all the time unlike 1728MHz CPU step. It now scales between the freqs 27, 128, 200, 320, 400, 487MHz. Fixed all instances where it wouldn't boost. PNPMGR and the GPU subsystem has been manipulated to cope with this extra step. By default, max is 487 and I recommend you keep it this way. Otherwise can use the script or TricksterMOD to cap it again at 400
    Click to expand...
    Click to collapse

    Download:
    Click to expand...
    Click to collapse
    https://www.androidfilehost.com/?fid=23622183712461199
    md5: b0db16602eaf3924faabeaa6eea110ee
    ***** ^Updated link at last time of edit^ *****
    Click to expand...
    Click to collapse

    Enjoy! Sorry for the wait again :D
    View
    9
    poondog
    poondog
    UPDATE

    Kangaroo Kernel v085 for 4.4.2 Sense and GPE (WILL NOT BOOT ON 4.4.3 GPE)
    Changelog: 8th June 2014
    .misc updates to congestion stuff and sched (for greater efficiency and battery)
    .fix bluetooth for Sprint and Verizon (thx Brymaster)
    .thermal: a bit cooler on the higher phases
    .cpufreq: ported CONSERVATIVE gov from Linux 3.14+ (stratosk, kernel.org)
    .improvements to memory model section of kernel (to do with processing stuff)
    .mutex enhancements from CAF
    .update SELINUX from 4.4.3 google source (kernel side security)
    .some misc updates from google 4.4.3 source
    .under the hood stuff (see git)

    Sorry it took so long to get the bluetooth fix out, I had already started working on this next version which usually goes out on weekends

    Download: (make sure on vzw you are flashing vzw)
    http://www.androidfilehost.com/?fid=23501681358543965
    md5; a6efc6314d79fd906ca13d3fe801c747

    enjoy!
    View

New posts