[Kernel 3.4.103][CAF][CC4.9] XGENESIS v.350 [19/08][AOSP/CM][4.4.4]

[jljeff]

Callmealdy est quand meme le mec qui fais le hellscreed et quand je vois ce que donne le hellscreed je me dis que ça peut être du lourd

[lolobarjo]

C'est reparti

As promised here's the update.

AK.340.MAKO

AK.340.MAKO.CM

Changelog :

Paul Reioux - Sound Control: separate locking mechanism for recording and playback

Paul Reioux - Sound Control: let register lock be dependent on different hw revisions

Paul Reioux - Sound Control: Misc clean up for newer WCD9xxx SOCs

Paul Reioux - Sound Control: expose direct register manipulations to userspace

Paul Reioux - intelli_plug: older Qualcomm kernel compatibility fixup

Paul Reioux - intelli_plug: use primary CPU's info data for non-boot cpu's settings

Paul Reioux - intelli_plug: initialize ip_info struct element during driver init

Paul Reioux - intelli_plug: fix incorrect cpufreq API usage

Paul Reioux - intelli_plug: only apply suspend/resume logic if active

CallMeAldy - Linux 3.4.101

CallMeAldy - Linux 3.4.100

CallMeAldy - Linux 3.4.99

CallMeAldy - Switched to 4.9.1-2014.07.20140718.CR83 toolchain

Cheers:good:

Envoyé de mon Nexus 5

[zarancar]

C'est bon ça, je veux bien le test pour le coup, après je sais pas si ont peur toujours appeler ça du AK mais bon, a voir.

[lolobarjo]

Bah pour le moment y'a pas de changement significatif qui pourrait justifier que ça n'est pas du ak. Cette MàJ est plus une MàJ de maintenance à mon goût

Envoyé de mon Nexus 5

[zarancar]

Oki ^^, merci de la précision lolo.

[zarancar]

Omgggg un kernel ak stable, enfin c'est pas trop tôt, étrange coïncidence lol^^, je troll un peu désolé.

Blague a part il m'a sortie un score de porc que j'avais jamais vu, un truc de malade jusqu'à ce que je m'aperçois qu'il avait overclock le processeur a 1.94ghz durant le test alors que c'était pas paramétré comme ça dans synapse mais bon, c'est tout toujours marrant.

Screen pour le fun sur la dernière mahdi.

Modifié 8 août 2014 par zarancar

[barome]

Il faut vraiment arreter de critiquer Ak. CallMeAldy, je ne vais pas le critiquer mais est nettement moins bon dev que Ak. Ak est l'inventeur de nombreuses choses. CallMeAldy ne fait que reprendre des commits a droite a gauche, comme moi.

[zarancar]

Je critique pas ak, j'ai toujours suivit son travail mais a part au début de Xgenesis si je me souviens bien, j'en ai rarement vu de stable où quand on met 81mhz et 1 core mini c'est 81 et 1 core et pas un yoyo permanent où tout s'allume comme un arbre de Noël,si j'ai des options de personnalisations, j'aime pouvoir les exploiter, la c'est possible et en suis content.

[barome]

CallMeAldy n arien touché a la partie CPU, pur hasard si par magie ca marche mieux chez toi. Apres, je le concede, la gestion CPU du Ak n'est pas celle que j apprecie.

[zarancar]

AK 345 : http://forum.xda-developers.com/showpost.php?p=54754367&postcount=9089

[lolobarjo]

New AKAK.350.MAKO

Aldrin Holmes - Linux 3.4.103

Aldrin Holmes - Added Sweep2Wake

savoca - lge_touch_core: Removed unused int variable retry

savoca - lge_touch_core: Implement doubletap-wake profiles TricksterMOD supports multiple values for dt2w, let's put it to use. 1 - (Semaphore Stock) Center wake zone 2 - Full wake zone 3 - Bottom half of screen wake zone 4 - Top half of screen wake zone

savoca - lge_touch_core: Fix dt2w set with screen off & android_touch

savoca - s2w: heavy cleanup, remove debug/unnecessary lines

savoca - dt2w: remove old code

savoca - lge_touch_core: sysfs to android_touch/doubletap2wake

stratosk - lge_touch_core: implement touchwake and doubletap wake - Also remove old s2w/dt2w/ps code from lge_touch_core.c

savoca - sweep2wake: Make S2W bidirectional Also enable S2W+S2D if S2W = 1 and S2D = 1

savoca - s2d: update inline with m8/n5/falcon - assign s2d_enabled to android_touch sysfs - don't update if any value is 255 - screen off if kcal drops < 1

savoca - LCD_KCAL: Change rgb values using sweep2sleep gestures

savoca - s2w: Add bi-directional sweep2sleep

savoca - s2w: Remove s2w_s2sonly, replace w/ s2w == '2'

Dennis Rassmann - drivers/input/touchscreen/sweep2wake: if s2s_only is true, opt out early

Dennis Rassmann - drivers/input/touchscreen/sweep2wake: add flo/deb definitions

Dennis Rassmann - drivers/input/touchscreen/lge_touch_core: fix ifdef, fixes dt2w without s2w

Dennis Rassmann - drivers/input/touchscreen/lge_touch_core: trigger on low and do not suspend the IRQ

Dennis Rassmann - drivers/input/touchscreen/sweep2wake: add mako definitions

Dennis Rassmann - drivers/input/touchscreen/lge_touch_core: release all touchscreen events when going to sleep

Dennis Rassmann - drivers/input/touchscreen/lge_touch_core: Add prevent_sleep functionality

Dennis Rassmann - drivers/input/touchscreen/sweep2wake: Fix compilation with earlysuspend

Dennis Rassmann - drivers/input/touchscreen/doubletap2wake: Fix compilation with earlysuspend

Dennis Rassmann - driver/input/touchscreen: add new s2w & dt2w implemenation from hammerhead

Dennis Rassmann - lge_touch_core: i2c_read: auto retry 10 times before erroring out The i2c devices we are connecting to may not be ready yet, or just busy. Retry with a 10ms pause in order to still obtain touch data. This improves overall touch data robustness, but the main thing it does: WE CAN FINALLY GET TOUCH DATA IN SUSPEND ALL THE TIME.

Aldrin Holmes - Revert "input: lge_touch_core: touchwake: initial support and code"

Aldrin Holmes - Revert "input: lge_touch_core: touchwake: possible fix touchwake reboot issue"

Aldrin Holmes - Revert "input: lge_touch_core: touchwake: Use delayed work to handle powerkey"

Aldrin Holmes - Revert "input: lge_touch_core: touchwake: Fix checkpatch warnings"

Aldrin Holmes - Revert "input: lge_touch_core: touchwake: Remove unused code"

Aldrin Holmes - Revert "input: lge_touch_core: touchwake: Introduce LED trigger"

Aldrin Holmes - Revert "input: lge_touch_core: touchwake: Handle touch suspend/resume only if touchwake is enabled"

Aldrin Holmes - Revert "Revert "input: lge_touch_core: touchwake: Handle touch suspend/resume only if touchwake is enabled""

Aldrin Holmes - Revert "Revert "Revert "input: lge_touch_core: touchwake: Handle touch suspend/resume only if touchwake is enabled"""

Aldrin Holmes - Revert "input: lge_touch_core: touchwake: Send input to touchwake only if it's enabled"

Aldrin Holmes - Revert "input: lge_touch_core: touchwake: Fix issue that touch remains unpowered after late_resume"

Aldrin Holmes - Revert "input: lge_touch_core: touchwake: Unlock the wakelock when disable the driver

DOWNLOAD

[lolobarjo]

AK.355.MAKO

CallMeAldy - Switch to 4.9.2-2014.08.20140831.CR83 Toolchain

mrg666 - Random: backport updates from Linux master (3.13-rc4) lib, net: make isodigit() public and use itAndy Shevchenko random: add debugging code to detect early use of get_random_bytes()Theodore Ts'o random: initialize the last_time field in struct timer_rand_stateTheodore Ts'o random: don't zap entropy count in rand_initialize()Theodore Ts'o random: printk notifications for urandom pool initializationTheodore Ts'o random: make add_timer_randomness() fill the nonblocking pool firstTheodore Ts'o random: convert DEBUG_ENT to tracepointsTheodore Ts'o random: push extra entropy to the output poolsTheodore Ts'o random: drop trickle modeTheodore Ts'o random: adjust the generator polynomials in the mixing function slightlyTheodore Ts'o random: speed up the fast_mix function by a factor of fourTheodore Ts'o random: cap the rate which the /dev/urandom pool gets reseededTheodore Ts'o random: optimize the entropy_store structureTheodore Ts'o random: optimize spinlock use in add_device_randomness()Theodore Ts'o random: fix the tracepoint for get_random_bytes(_arch)Theodore Ts'o random: account for entropy loss due to overwritesH. Peter Anvin random: allow fractional bits to be trackedH. Peter Anvin random: statically compute poolbitshift, poolbytes, poolbitsH. Peter Anvin random: mix in architectural randomness earlier in extract_buf()Theodore Ts'o random: allow architectures to optionally define random_get_entropy()Theodore Ts'o random: run random_int_secret_init() run after all late_initcallsTheodore Ts'o char: Convert use of typedef ctl_table to struct ctl_tableJoe Perches1 random: fix accounting race condition with lockless irq entropy_count updateJiri Kosina drivers/char/random.c: fix priming of last_dataJarod Wilson lib/string_helpers: introduce generic string_unescapeAndy Shevchenko random: fix locking dependency with the tasklist_lockTheodore Ts'o locking: Various static lock initializer fixesThomas Gleixner random: prime last_data value per fips requirementsJarod Wilson random: fix debug format stringsJiri Kosina random: make it possible to enable debugging without rebuildJiri Kosina random: mix in architectural randomness in extract_buf()H. Peter Anvin random: Add comment to random_initialize()Tony Luck random: add tracepoints for easier debugging and verificationTheodore Ts'o

motley - mako LCD: keep RGB custom colors alive after screen off/on

savoca - LCD_KCAL: Introduce configurable kcal_min

savoca - LCD_KCAL: Prevent out-of-range input values

savoca - LCD_KCAL: Fix mistake on out-of-range values

savoca - s2d: turn off display if sweep < kcal_min, not 1

CallMeAldy - ak-mako-build.sh: Minor changes

DOWNLOAD

Envoyé de mon Nexus 5

[lolobarjo]

AK.360.MAKO

[HIDE]

CallMeAldy - Kernel Version: AK.360.MAKO

ktoonsez - sched: remove redundant update_runtime notifier

ktoonsez - mm: optimize Kernel Samepage Merging(KSM)

Puneet Kumar - CHROMIUM: mm: Fix calculation of dirtyable memory

franciscofranco - arm: compile with the generic rwsem

Rob Herring - ARM: 7494/1: use generic termios.h

Andi Kleen - Disable -fconserve-stack on ARM v2

Sumit Singh - hrtimer: enhance power efficiency

Sumit Singh - sched: Trying to reduce power usage

Sumit Singh - seqlock: enhance power efficiency

Sumit Singh - cpu: enhance power efficiency

Sumit Singh - ARM: including asm-generic/processor.h

Sumit Singh - asm-generic: processor.h: adding new header file

Sumit Singh - llist: defining relaxed version of llist_empty

Sumit Singh - asm-generic: relaxed.h: defined relaxed.h

Sumit Singh - sched: defining relaxed version of idle_cpu

Sumit Singh - arm: asm: relaxed.h: defined relaxed.h

Swetha Chikkaboraiah - kernel: cgroup: protecting css_set object

Srinivasarao P - Revert "kernel: cgroup: protect cgroup linked lists"

Hannes Frederic Sowa - random: check for increase of entropy_count because of signed conversion

Theodore Ts'o - random: fix nasty entropy accounting bug

Joe Perches - random: convert use of typedef ctl_table to struct ctl_table

Theodore Ts'o - random: fix BUG_ON caused by accounting simplification

Theodore Ts'o - random: use the architectural HWRNG for the SHA's IV in extract_buf()

Greg Price - random: entropy_bytes is actually bits

Greg Price - random: simplify accounting code

Greg Price - random: forget lock in lockless accounting

Greg Price - random: simplify accounting logic

Greg Price - random: simplify loop in random_read

Greg Price - random: fix description of get_random_bytes

Greg Price - random: fix comment on proc_do_uuid

Greg Price - random: fix typos / spelling errors in comments

Nick Kossifidis - random: Mix cputime from each thread that exits to the pool

Matt Wagantall - ARM: Improve vmalloc address checking in show_data()

Srinivasarao P - kernel: cgroup: protect cgroup linked lists

Pavankumar Kondeti - USB: msm_otg: Fix a bug in charger detection procedure

Sujeet Kumar - USB: msm_otg: Fix host mode suspend bug

Venkat Gopalakrishnan - mmc: msm_sdcc: Remove unnecessary delay in interrupt context

Li Zefan - cgroup: fix exit() vs rmdir() race

Stanislaw Gruszka - sched: fix divide by zero at {thread_group,task}_times

Eric Paris - SELinux: possible NULL deref in context_struct_to_string

Himangi Saraogi - SELinux: use ARRAY_SIZE

Lai Jiangshan - kthread_work: wake up worker only when the worker is idle

Tetsuo Handa - kthread: fix return value of kthread_create() upon SIGKILL.

Nishanth Aravamudan - kthread: ensure locality of task_struct allocations

Tetsuo Handa - kthread: make kthread_create() killable

Tom Herbert - net: Performance fix for process_backlog

Alok Chauhan - msm: msm_bus: Suspected fix for deadlock

Paul Reioux - Sound Control: add line gain to cached and register lock

Paul Reioux - Sound Control: fix copy and paste error for separate record lock

franciscofranco - sched: set sched_mc_power_savings to 2, which is the highest power saving possible value for this tunable. Basically according to http://elinux.org/images/1/1d/Comparing_Power_Saving_Techniques_For_Multicore_ARM_Platforms.pdf this tunable provides serious power saving improvements close to the power savings of hotpluging cores. Since in this Kernel we're using always a minimum of 2 cpus online at any time during screen off we might actually benefit from this because this feature packs tasks together and try to bind them to cpu0, which in theory will let cpu1 idle longer, thus improving battery life.

barome - net: wireless: bcmdhd: Disable OOB interrupt when WLAN is off

CallMeAldy - Linux 3.4.104

[/HIDE]

DOWNLOAD

Envoyé de mon Nexus 5