**NEW KERNEL IN AUR** kernel26zenmm-git

rooloo · 2008-08-11 17:16:22

ianegg wrote:

The package installs correctly with no errors. In this instance I adjusted the PKGBUILD for convenience but when I tried yesterday I noticed the PKGBUILD now has _kernver=`uname -r` at the top, which I assumed was a workaround for the date string being appended, so I rebooted with the new kernel before running makepkg. If _kernver is wrong, makepkg finishes before any actual compilation begins.
pacman -Ql nvidia-beta-zenmm                                                                                              ~/packages/nvidia-beta-zenmm
nvidia-beta-zenmm /lib/
nvidia-beta-zenmm /lib/modules/
nvidia-beta-zenmm /lib/modules/2.6.27-rc1-zenmm0-20080810/
nvidia-beta-zenmm /lib/modules/2.6.27-rc1-zenmm0-20080810/kernel/
nvidia-beta-zenmm /lib/modules/2.6.27-rc1-zenmm0-20080810/kernel/drivers/
nvidia-beta-zenmm /lib/modules/2.6.27-rc1-zenmm0-20080810/kernel/drivers/video/
nvidia-beta-zenmm /lib/modules/2.6.27-rc1-zenmm0-20080810/kernel/drivers/video/nvidia.ko
cat PKGBUILD | grep kernver                                                                                                    ~/packages/nvidia-beta-zenmm
#_kernver=`uname -r`
_kernver="2.6.27-rc1-zenmm0-20080810"
  echo $_kernver
  echo ${_kernver}
  make SYSSRC=/lib/modules/${_kernver}/build module
  mkdir -p $startdir/pkg/lib/modules/${_kernver}/kernel/drivers/video/
  install -m644 nvidia.ko $startdir/pkg/lib/modules/${_kernver}/kernel/drivers/video/
  sed -i -e "s/KERNEL_VERSION='.*'/KERNEL_VERSION='${_kernver}'/" $startdir/*.install
edit: I really would like to test a known working version of the kernel before I recompile the whole thing again. I want to stick to vanilla so I know I'm not breaking anything, but my system takes a very long time to compile all the stuff I don't need. Have I misunderstood some sublety of the lines that are to be commented out and `uname -r`? I'll do it again if you think it will make a difference, but I don't want to spend hours compiling the exact same kernel with a different name...

OK, Let me upload the packages I have to server and you can try them. For now you can just download using wget and install with pacman -U.

ianegg · 2008-08-11 18:55:02

Thanks. I've just double checked with the standard arch kernel and I get the same problem, so I guess this is a problem unique to me... Is there anything I should be looking out for in my environment variables?

ianegg · 2008-08-14 10:02:21

Sorry to double post... Have you uploaded anything yet rooloo, or has anyone else on x86_64 got a kernel + driver I can try?

rooloo · 2008-08-14 11:38:11

ianegg wrote:

Sorry to double post... Have you uploaded anything yet rooloo, or has anyone else on x86_64 got a kernel + driver I can try?

I have e-mailed you a link to try.

ianegg · 2008-08-14 13:26:05

Thankyou. Frustratingly I get the same error trying to start X, so it seems there's some problem with the 177.13 drivers on my system.

thisllub · 2008-08-15 22:43:10

Installed it and the nvidia driver.

I had to manually remove the conflicting files before the nvidia-utils package would build.

On first starting Firefox the npviewer program went into meltdown and started over 30 instances of sh.
I have had to remove the flash plugin to stop this.

I am hoping that this kernel solves the intermittent hanging problem. http://bbs.archlinux.org/viewtopic.php?id=43932

Cotton · 2008-08-16 10:45:43

Any reason why no disk activity is reported by gkrellm when using this kernel?

rooloo · 2008-08-17 04:04:25

Cotton wrote:

Any reason why no disk activity is reported by gkrellm when using this kernel?

probably don't have the right module compiled for the sensor. This is a vanilla kernel config, You are probably missing a module.

Last edited by rooloo (2008-08-17 04:04:37)

ianegg · 2008-08-17 19:38:39

rooloo, you said you were thinking of creating a zenmm+nvidia repo and keeping it strictly vanilla but there's one thing I think would be a good idea to change; 32 bit binary support on 64 bit kernels seems to be disabled by default for some reason.

bbs - is that deliberate or an oversight?

Last edited by ianegg (2008-08-17 19:40:03)

bbs · 2008-08-17 20:41:02

all,

again thank you very much for your interests -- we are trying to keep zenmm in great shape and as bleeding edge as possible -- i've been very busy in the last month with some oddjobs, work and preperation for school -- if someone would like to list a recommended set of additions (to the PKGBUILD -- we are taking care of in kernel features @ zen) -- then i would be happy to incorperate them or even better a diff for the pkg so i can easily change it -- i'd be happy to incorperate them in the next week -- time should permit me to play a bit more this up coming week -- again -- thanks for the interest and i hope that the users really do appreciate the power of this kernel -- i should be adding a zen-rt (realtime) -- pkgbuild for those who should choose to take advantage (need to ) or realtime from molnar

hope everyone is doing well

jT

bbs · 2008-08-17 20:43:15

oh and ianegg -- i'm a pro pure 64 bit kinda guy -- you should be able to hack this relatively easily -- if it becomes a large problem feel free to drop me a line via the bulletin board email box and i'll send you a fix

thisllub · 2008-08-17 21:27:00

I recompiled it 4 times but I still can't get the pptp client to work.
I might have another go next weekend.

arch0r · 2008-08-18 06:09:48

pls explain a noob why to use this custom kernel!? do i have any vantages using it?

Last edited by arch0r (2008-08-18 06:09:58)

methuselah · 2008-08-18 09:19:57

arch0r wrote:

pls explain a noob why to use this custom kernel!? do i have any vantages using it?

For me, the new alsa and hp specific modules (I think it was wmi hp) fixed my multimedia volume keys. On my other 2.6.26-2 Arch kernel, the multimedia volume key doesn't go up or down smoothly. It always breaks the left/right channel so the volume becomes uneven and unbalanced, and it goes from 75% volume to 10% volume when trying to gently increase the volume.

Now with this zenmm and the newer alsa and hp modules, the multimedia volume is as smooth as it is in Vista, and sounds really good.

Another good thing is the new NVIDIA driver. It fixed my black tty problem, and works better with compiz. The NVIDIA beta driver can also be used in the regular kernel so it's not really a zenmm thing.

I also noticed that my dmesg seems to have changed this part from a generic Archx86_64:

Clockevents: could not switch to one-shot mode: lapic is not functional.
Could not switch to high resolution mode on CPU 0
Clockevents: could not switch to one-shot mode: lapic is not functional.
Could not switch to high resolution mode on CPU 1

to this part with zenmmx86_64 compiled for AMD64:

Switched to high resolution mode on CPU 0
Switched to high resolution mode on CPU 1

However, pm-suspend can't resume like it did in 2.6.25... it broke for me when the kernel moved to 2.6.26?????? And I've tried a few different fixes for the zenmm, but no luck so far. Maybe I'm doing something wrong. It goes into suspend normally, but it won't resume at all... it just stays suspended.

I also seem to have problems with my usb 2.0 digital camera not being auto-mounted when I plug it in.

I'm sticking with the zenmm kernel for the improved multimedia keys... otherwise I think the Arch kernel is just as fast.

rooloo · 2008-08-18 13:09:34

if your building zenmm-git, then it is now and has been for 10 days or so, 2.6.27-rc1 source code.

suspend could have been working in 2.6.26-zenmm, maybe it's a upstream bug in 2.6.27-rc1. You could always try and d/l the patch for 2.6.26 kernel from zen-sources and re-compile 2.6.26 stable with zenmm patch.

bbs · 2008-08-18 15:12:18

methuselah wrote:

arch0r wrote:
pls explain a noob why to use this custom kernel!? do i have any vantages using it?
For me, the new alsa and hp specific modules (I think it was wmi hp) fixed my multimedia volume keys. On my other 2.6.26-2 Arch kernel, the multimedia volume key doesn't go up or down smoothly. It always breaks the left/right channel so the volume becomes uneven and unbalanced, and it goes from 75% volume to 10% volume when trying to gently increase the volume.

Now with this zenmm and the newer alsa and hp modules, the multimedia volume is as smooth as it is in Vista, and sounds really good.

Another good thing is the new NVIDIA driver. It fixed my black tty problem, and works better with compiz. The NVIDIA beta driver can also be used in the regular kernel so it's not really a zenmm thing.

I also noticed that my dmesg seems to have changed this part from a generic Archx86_64:
Clockevents: could not switch to one-shot mode: lapic is not functional.
Could not switch to high resolution mode on CPU 0
Clockevents: could not switch to one-shot mode: lapic is not functional.
Could not switch to high resolution mode on CPU 1
to this part with zenmmx86_64 compiled for AMD64:
Switched to high resolution mode on CPU 0
Switched to high resolution mode on CPU 1
However, pm-suspend can't resume like it did in 2.6.25... it broke for me when the kernel moved to 2.6.26?????? And I've tried a few different fixes for the zenmm, but no luck so far. Maybe I'm doing something wrong. It goes into suspend normally, but it won't resume at all... it just stays suspended.

I also seem to have problems with my usb 2.0 digital camera not being auto-mounted when I plug it in.

I'm sticking with the zenmm kernel for the improved multimedia keys... otherwise I think the Arch kernel is just as fast.

hey -- you can alternatively run git without the pkgbuild and choose the specific tags that you would like to run -- namely run 2.6.26-rc8-zenmm3 if you like -- (i believe that was a release) -- i'm just doing stuff right now with tp_smapi and trying to keep it as close to the edge as possible so i created this pkgbuild to always track master head -- it would be a sinch to make the pkgbuild utilize another tag and git checkout v2.6.26-rc8-zenmmX -- and also make a branch for it with -b option -- just a few thoughts on that --

my real recommendation for this kernel is it isn't _really_ meant for newer people coming over to linux -- its really meant for people who have been around linux for a year or so and want to stray away from the generic linux kernels provided in binary packages for specific distrobutions -- another reason to use this kernel is for situations (like above) -- where newer more bleeding edge kernels provide cooler/better support -- this kernel essentially _always_ tracks linux-next -- its not meant to be stable -- its meant to provide a neat featureset that is stable enough to use and actually get work done -- if i based it on mortons daily snapshots -- namely mmotm -- it would be as he puts it -- a barely even know if this will compile and if it does won't run well POS --

another caveat is the fact that arch provides a way for pkgbuilds to be made readily available -- and since we take such good care of the repository -- it normally builds well all the time -- making it easy for newcomers to try something crazy out and live on the edge with their linux kernel -- anyways just a few more thoughts as i surface back to the more fun world of opensource

hope everyone is smiling -- hopefully we can get a 2.6.27-rc2/3 where i can actaully get TOI to work under zenmm -- vmscan.c is being a little B

jT

bbs · 2008-08-18 15:17:36

rooloo -- thanks for staying on top of things while i have been absent -- and nice nvidia pkgbuild -- i was thinking of encorperating it -- because most people know whether or not they have nvidia ;p so there would be a prompt instead of an automated check and a parse of the lspci output or somesuch -- but i think i'll let yours shine for a week or two more because its a great addition/solution

peace

jT

methuselah · 2008-08-18 18:08:02

rooloo wrote:

if your building zenmm-git, then it is now and has been for 10 days or so, 2.6.27-rc1 source code.

suspend could have been working in 2.6.26-zenmm, maybe it's a upstream bug in 2.6.27-rc1. You could always try and d/l the patch for 2.6.26 kernel from zen-sources and re-compile 2.6.26 stable with zenmm patch.

I've been using the testing packages for Arch since I switched from xubuntu last year... so, before I tried my first zenmm kernel 2.6.27-rc1-20080805, I had been using the testing version of the Arch 2.6.26-2... I just recently noticed that the 2.6.26-2 kernel became a core package.

It was when I switched to the new testing 2.6.26-1 kernel when my "sudo pm-suspend/hibernate" stopped resuming (the synaptic touchpad doesn't even get noticed when trying to resume, only my multimedia power buttons do)... I had originally thought that the zenmm was a kernel pre-patched with tuxonice and that was one of the main reasons I decided to try zenmm out in the first place... aircrack-ng would of been the second reason except I already had a fully functioning aircrack-ng-svn with a injection using the rtl8187 driver and my AWUS036H.

Well, since August 5th, I must of rebuilt the kernel at least 20 different ways, using different tunables and researching my computer and each possible setting, and what I could leave out, sometimes totally failing... I'm still learning a bunch of things about compiling a kernel... this being my true first successful go at it, continuously rebuilding it and checking out my dmesg and performance.

I'll put "resume=/dev/sda5" back into my regular Arch's /boot/grub/menu.lst and see if the problem has been fixed since core.

I have also researched the problem on google and the hal quirks page, but I have a pretty new computer (hp pavilion dv9920us) and there isn't a lot of info out there for my dv9920us other than people having the same problems, without any answers solved yet.

...as for the zenmm tunable for resume, I "*" yes'ed them both and added resume=/dev/sda5 (which is my swap partition) to the resume string. I also don't know if it has to do with a compression issue... I did not see the module LZF, so I added the module "M" for LZO in the encryption and compression section... I have also followed all directions on this page: http://wiki.archlinux.org/index.php/Suspend_to_Disk for the Tuxonice method... with and without compression... but no luck so far... regular sudo pm-suspend also doesn't work... I might try the uswsusp method... or I might just wait for someone else to fix it and post the exact directions of how to make it work with my hp.

I have also tried to change something that I've never seen on any wiki or forum before... what I tried was I used the uuid instead of resume=/dev/sda5... so my grub menu.lst looks like this for zenmm:

# (0) zenmm
title  zenmm
root   (hd0,2)
kernel /vmlinuz26zenmm root=/dev/disk/by-uuid/fe976406-f6d8-4cd0-99ea-e6e8aab2c397 resume=/dev/disk/by-uuid/bb1fd790-0f86-48b3-a744-f6ae2511b83b iommu=noaperture ro
initrd /kernel26zenmm.img

... maybe I should go back to the other two ways I've tried which are:

resume=/dev/sda5

and

resume=swap:/dev/sda5

Anyway, I thank you both for your zenmm-git and the nvidia-beta packages, I'm going on about 2 solid years of Linux and during these 2 years I spent as much time as I could learning to use the console/terminal and the proper way to use Linux... I switched to Arch last year because I wanted a faster more hands on system than xubuntu (which I already used as many different "config speed hacks" as possible), and the learning curve for Arch has been exactly what I wanted from a Linux OS, and now that I'm learning the kernel and "patching/tuning"... this is my next step to learning more about Linux, so thank you again.

... and I agree with bbs, you should have a bit of Linux experience and not need a "stable" kernel... For me, I switched to Linux 2 years ago because I like to learn the inner workings of the computer, and if I break something I enjoy fixing it on my own... I have the 2.6.26-2 kernel to fallback on, as well as my Vista partition...

I still say that the alsa and nvidia were major improvements and I prefer that to the suspend/resume and usb 2.0 automount... which I'm sure will get fixed by me or someone else.

thanks again,
-c.

methuselah · 2008-08-18 18:26:42

hey -- you can alternatively run git without the pkgbuild and choose the specific tags that you would like to run -- namely run 2.6.26-rc8-zenmm3 if you like -- (i believe that was a release) -- i'm just doing stuff right now with tp_smapi and trying to keep it as close to the edge as possible so i created this pkgbuild to always track master head -- it would be a sinch to make the pkgbuild utilize another tag and git checkout v2.6.26-rc8-zenmmX -- and also make a branch for it with -b option -- just a few thoughts on that --

Thanks, that's good to know.

my real recommendation for this kernel is it isn't _really_ meant for newer people coming over to linux -- its really meant for people who have been around linux for a year or so and want to stray away from the generic linux kernels provided in binary packages for specific distrobutions -- another reason to use this kernel is for situations (like above) -- where newer more bleeding edge kernels provide cooler/better support -- this kernel essentially _always_ tracks linux-next -- its not meant to be stable -- its meant to provide a neat featureset that is stable enough to use and actually get work done -- if i based it on mortons daily snapshots -- namely mmotm -- it would be as he puts it -- a barely even know if this will compile and if it does won't run well POS --

I totally agree... Linux for me is my new hobbie (2 solid years)... it's like being a beginner mechanic, or building models... even if I am totally happy with my OS, I am always trying to improve or learn/try something new.

another caveat is the fact that arch provides a way for pkgbuilds to be made readily available -- and since we take such good care of the repository -- it normally builds well all the time -- making it easy for newcomers to try something crazy out and live on the edge with their linux kernel -- anyways just a few more thoughts as i surface back to the more fun world of opensource

Yes, the AUR and PKGBUILDS rule!!!!! I've been using the AUR since I switched to Arch last year... occasionally I still have to build a package the old way, or write my own PKGBUILD and .install file... I should start learning to use the ABS one of these days, but the AUR and the regular repositories have been all I needed for the last year.

hope everyone is smiling -- hopefully we can get a 2.6.27-rc2/3 where i can actaully get TOI to work under zenmm -- vmscan.c is being a little B
jT

Totally, I saw this on digg the other day: http://digg.com/linux_unix/2_6_27_rc3_T … almed_Down

Thanks again,
-c.

ianegg · 2008-08-21 14:47:35

I'd just like to mention I can actually start using this kernel now with the new nvidia beta 177.67 which compiles, installs and runs perfectly. (so far)

bbs · 2008-08-26 08:49:39

umm -- i'm sorry to tempt you all -- but i've added a new one -- kernel26zenmmotm-git -- its based on morton's DAILY broken-out patches -- its running like absolute butter at the moment -- there is some intel microcode issues -- but other than that its ok -- again this is _not_ a stable kernel -- its for those who want to live a bit more on the edge and for people who play with linux-next type stuff

/jT

pogeymanz · 2008-09-28 23:06:34

There seems to be a problem, at least for me.

I get this error when compiling:

fs/ext4/defrag.c: In function 'ext4_defrag_ioctl':
fs/ext4/defrag.c:677: error: 'struct task_struct' has no member named 'fsuid'
fs/ext4/defrag.c: In function 'ext4_defrag_alloc_blocks':
fs/ext4/defrag.c:1388: error: implicit declaration of function 'ext4_ext_calc_credits_for_insert'
make[2]: *** [fs/ext4/defrag.o] Error 1
make[1]: *** [fs/ext4] Error 2
make[1]: *** Waiting for unfinished jobs....

So, it seems that the source code is incorrect, so I went into the build folder and just commented out a bunch of stuff in ext4/defrag.c (and I disabled ext4 support anyway)

Then when I ran makepkg -c on the directory it somehow undid my changes...

bbs · 2008-09-28 23:58:51

pogeymanz,
yea this latest mm was POS -- well not that i'm expecting miracles from a highly unstable tree. alternatively if you want to get this to work now -- i don't have a new zenmm but i _do_ have a very working latest mm -- are you familiar with git? if so you can be up and running the latest working mm w/o the patches so i can rebase properly when he releases the next mm -- i expect relatively shortly -- if you are feeling a bit on the crazy side do this:

git clone http://git.zen-sources.org/kernel/zenmm.git
git checkout -b latest 2008-09-23-11-27

now you are in v2.6.27-rc7-mm1
^^ thats the latest mm you can get right now -- i'm working out the allyesconfig kinks -- but we will see -- looks ok for now -- if i add the zen branches in i'm _sure_ bad things will happen -- so i'm holding off until the next official zen release emerges

hope that helps

/jT

pogeymanz · 2008-09-29 00:17:37

I might have some time to give that a shot tomorrow.

We'll see how it works out.

Thanks.

methuselah · 2008-09-29 00:49:58

@ bbs:

Do you know a way to configure the resume function to work properly with the zen kernels? I've built multiple versions of the zenmm, the zen2 2.6.27-rc5, and the zen2 2.6.26.

With your zenmm kernels, I couldn't resume from suspend like the regualr -ARCH kernel.

In both the zen2 patched vanilla kernels that I compiled (2.6.26 and 2.6.27-rc5), they both came with tuxonice patches. Neither of those would suspend properly because of my nvidia module and for some reason my webcam would always start when I tried to suspend.....

When I made my current zen2 2.6.26 kernel that I use now I didn't add the tuxonice modules and I configured the swap/resume the same way as the stock -ARCH kernel does in the wiki, but it still won't resume.... it just stays a blank screen.

I also tried the uswsusp method which would suspend correctly and print out a verbose message saying so, but it would break my swap when resuming and just boot into a regular boot up. Then I would have to start my swap with "swapon -a".

Any knowledge of what I might need to do to get the zen2 patch to resume correctly?

Last edited by methuselah (2008-09-29 00:53:06)

Arch Linux

#51 2008-08-11 17:16:22

Re: **NEW KERNEL IN AUR** kernel26zenmm-git

#52 2008-08-11 18:55:02

Re: **NEW KERNEL IN AUR** kernel26zenmm-git

#53 2008-08-14 10:02:21

Re: **NEW KERNEL IN AUR** kernel26zenmm-git

#54 2008-08-14 11:38:11

Re: **NEW KERNEL IN AUR** kernel26zenmm-git

#55 2008-08-14 13:26:05

Re: **NEW KERNEL IN AUR** kernel26zenmm-git

#56 2008-08-15 22:43:10

Re: **NEW KERNEL IN AUR** kernel26zenmm-git

#57 2008-08-16 10:45:43

Re: **NEW KERNEL IN AUR** kernel26zenmm-git

#58 2008-08-17 04:04:25

Re: **NEW KERNEL IN AUR** kernel26zenmm-git

#59 2008-08-17 19:38:39

Re: **NEW KERNEL IN AUR** kernel26zenmm-git

#60 2008-08-17 20:41:02

Re: **NEW KERNEL IN AUR** kernel26zenmm-git

#61 2008-08-17 20:43:15

Re: **NEW KERNEL IN AUR** kernel26zenmm-git

#62 2008-08-17 21:27:00

Re: **NEW KERNEL IN AUR** kernel26zenmm-git

#63 2008-08-18 06:09:48

Re: **NEW KERNEL IN AUR** kernel26zenmm-git

#64 2008-08-18 09:19:57

Re: **NEW KERNEL IN AUR** kernel26zenmm-git

#65 2008-08-18 13:09:34

Re: **NEW KERNEL IN AUR** kernel26zenmm-git

#66 2008-08-18 15:12:18

Re: **NEW KERNEL IN AUR** kernel26zenmm-git

#67 2008-08-18 15:17:36

Re: **NEW KERNEL IN AUR** kernel26zenmm-git

#68 2008-08-18 18:08:02

Re: **NEW KERNEL IN AUR** kernel26zenmm-git

#69 2008-08-18 18:26:42

Re: **NEW KERNEL IN AUR** kernel26zenmm-git

#70 2008-08-21 14:47:35

Re: **NEW KERNEL IN AUR** kernel26zenmm-git

#71 2008-08-26 08:49:39

Re: **NEW KERNEL IN AUR** kernel26zenmm-git

#72 2008-09-28 23:06:34

Re: **NEW KERNEL IN AUR** kernel26zenmm-git

#73 2008-09-28 23:58:51

Re: **NEW KERNEL IN AUR** kernel26zenmm-git

#74 2008-09-29 00:17:37

Re: **NEW KERNEL IN AUR** kernel26zenmm-git

#75 2008-09-29 00:49:58

Re: **NEW KERNEL IN AUR** kernel26zenmm-git

Board footer

Re: NEW KERNEL IN AUR kernel26zenmm-git

Re: NEW KERNEL IN AUR kernel26zenmm-git

Re: NEW KERNEL IN AUR kernel26zenmm-git

Re: NEW KERNEL IN AUR kernel26zenmm-git

Re: NEW KERNEL IN AUR kernel26zenmm-git

Re: NEW KERNEL IN AUR kernel26zenmm-git

Re: NEW KERNEL IN AUR kernel26zenmm-git

Re: NEW KERNEL IN AUR kernel26zenmm-git

Re: NEW KERNEL IN AUR kernel26zenmm-git

Re: NEW KERNEL IN AUR kernel26zenmm-git

Re: NEW KERNEL IN AUR kernel26zenmm-git

Re: NEW KERNEL IN AUR kernel26zenmm-git

Re: NEW KERNEL IN AUR kernel26zenmm-git

Re: NEW KERNEL IN AUR kernel26zenmm-git

Re: NEW KERNEL IN AUR kernel26zenmm-git

Re: NEW KERNEL IN AUR kernel26zenmm-git

Re: NEW KERNEL IN AUR kernel26zenmm-git

Re: NEW KERNEL IN AUR kernel26zenmm-git

Re: NEW KERNEL IN AUR kernel26zenmm-git

Re: NEW KERNEL IN AUR kernel26zenmm-git

Re: NEW KERNEL IN AUR kernel26zenmm-git

Re: NEW KERNEL IN AUR kernel26zenmm-git

Re: NEW KERNEL IN AUR kernel26zenmm-git

Re: NEW KERNEL IN AUR kernel26zenmm-git

Re: NEW KERNEL IN AUR kernel26zenmm-git