How-To: 2.6.14 Vanilla Kernel (latest) + ck Patchset (Enhanced Performance kernel)

[RubenGonc]

This How-To will guide you through the compilation/installation of the 2.6.14 Vanilla Kernel (more recent than the one distribuited with Ubuntu 5.10) with the performance patchset by Kon Colivas.

1-Download what is needed:

Type in the command line:

sudo apt-get install build-essential bin86 kernel-package

sudo apt-get install libqt3-headers libqt3-mt-dev (needed for make xconfig)

Then download the following files to your Home directory:

http://www.kernel.org/pub/linux/kernel/v2.6/linux-2.6.14.tar.bz2

http://ck.kolivas.org/patches/2.6/2.6.14/2.6.14-ck1/patch-2.6.14-ck1.bz2

2-Now, lets unpack the kernel source to /usr/src:

*Copy the source to /usr/src:

sudo cp linux-2.6.14.tar.bz2 /usr/src

*Change to /usr/src:

cd /usr/src

*Unpack it:

sudo tar -xvjf linux-2.6.14.tar.bz2

*Now lets change the created directory name:

sudo mv linux-2.6.14/ linux-2.6.14cK1

*Remove the symlink if there is one directory called linux:

sudo rm -rf linux

*Make the new symlink pointing for our 2.6.14 kernel source:

sudo ln -s /usr/src/linux-2.6.14cK1 linux

3-Now it's time to patch the kernel:

*Change to /usr/src/linux:

cd /usr/src/linux

*Switch to root user:

sudo -s -H
Password: <specify user password>

*Apply the patch:

sudo bzcat /home/username/patch-2.6.14-ck1.bz2| patch -p1

4-Kernel configuration:

*Import the configuration of the running kernel:

uname -r To see what kernel are you running (in my case it is 2.6.12-9-686).

sudo cp /boot/config-2.6.12-9-686 .config To copy the config file and use it as base for the new kernel configuration (Don't forget to choose the correct config file).

*Configure the kernel:

sudo make xconfig

*While you may tweak your kernel configuration to your needs I will sugest you some tweaks:

In "General Setup" activate:

-Support for paging of anonymous memory (swap)
--Support for prefetching swapped memory

In "Processor type and features":

-Processor family Choose the model of your processor.

Activate:

-Preemption Model
--Voluntary Kernel Preemption (Desktop)

-High Memory Support
--off -if you have less than 1 GB of RAM
--1GB Low Memory Support -if you have 1GB of RAM
--4GB -if you have more than 1GB of RAM

-Timer frequency
--1000 Hz

In "Device drivers" go to "Block devices" and in "IO Schedulers" leave only the "CFQ I/O scheduler" activated, which provides the best performance.

In "Kernel hacking" uncheck "Kernel debugging".

Ctrl+S to save the kernel configuration and then close the window.

5-Let's build the kernel:

*In a terminal make sure you are in /usr/src/linux with full root access.

We will build a ".deb" file that can be installed in our Ubuntu system, using make-kpkg.

*In a terminal type:

make-kpkg clean

make-kpkg -initrd --revision=ck1 kernel_image

If there wasn't errors this will build the kernel and a ".deb" file will be created at /usr/src.
*To install it:

sudo dpkg -i kernel-image-2.6.14*.deb


6-Reboot and everything should be running ok!
*Try:

uname -r

to see that you are running the new kernel.


7-Nvidia graphic cards users:

To install Nvidia driver follow the guide (no need to install linux-headers as we have the source in /usr/src/linux:

http://www.ubuntuforums.org/showthread.php?t=57368&highlight=nvidia


If you have any problem ask here.

Hope you like it!

Rúben Gonçalves:p

[23meg]

Shall we retain the Ubuntu devs' patches to the 2.6.12 kernel (found in the linux-patch-ubuntu-2.6.12 package) by using this kernel? Are Ubuntu's kernel patches sent upstream and merged, or do they have to be applied manually?

[XDevHald]

If you're not sure of it's changes please see here: http://www.kernel.org/pub/linux/kernel/v2.6/ChangeLog-2.6.14

And also other news here: http://linuxdevices.com/news/NS7213494018.html4

[RubenGonc]

Shall we retain the Ubuntu devs' patches to the 2.6.12 kernel (found in the linux-patch-ubuntu-2.6.12 package) by using this kernel? Are Ubuntu's kernel patches sent upstream and merged, or do they have to be applied manually?


I think that it isn't needed. I haven't applied them and everything is alright:)

[crypto178]

Nice guide, worked flawlessy for me, thanks :D

[Dracontopes]

root@ubuntu:/usr/src/linux# sudo make xconfig
*
* Unable to find the QT installation. Please make sure that the
* QT development package is correctly installed and the QTDIR
* environment variable is set to the correct location.
*
make[1]: *** [scripts/kconfig/.tmp_qtcheck] Fout 1
make: *** [xconfig] Fout 2


and


root@ubuntu:/usr/src/linux# sudo apt-get install libqt3-headers
Pakketlijsten worden ingelezen... Klaar
Boom van vereisten wordt opgebouwd... Klaar
libqt3-headers is reeds de nieuwste versie.
0 pakketten opgewaardeerd, 0 nieuwe pakketten geïnstalleerd, 0 verwijderen en 0 niet opgewaardeerd.

(it says libqt3-headers is already the newest version.)

[Rob2687]

Is pentium-m strictly for the newer mobile chips or can I use that option with my PIII-M?

[berserker]

root@ubuntu:/usr/src/linux# sudo make xconfig
*
* Unable to find the QT installation. Please make sure that the
* QT development package is correctly installed and the QTDIR
* environment variable is set to the correct location.
*
make[1]: *** [scripts/kconfig/.tmp_qtcheck] Fout 1
make: *** [xconfig] Fout 2


and


root@ubuntu:/usr/src/linux# sudo apt-get install libqt3-headers
Pakketlijsten worden ingelezen... Klaar
Boom van vereisten wordt opgebouwd... Klaar
libqt3-headers is reeds de nieuwste versie.
0 pakketten opgewaardeerd, 0 nieuwe pakketten geïnstalleerd, 0 verwijderen en 0 niet opgewaardeerd.

(it says libqt3-headers is already the newest version.)

Try installing the libqt3-mt-dev package.

[crypto178]

root@ubuntu:/usr/src/linux# sudo make xconfig
*
* Unable to find the QT installation. Please make sure that the
* QT development package is correctly installed and the QTDIR
* environment variable is set to the correct location.
*
make[1]: *** [scripts/kconfig/.tmp_qtcheck] Fout 1
make: *** [xconfig] Fout 2


and


root@ubuntu:/usr/src/linux# sudo apt-get install libqt3-headers
Pakketlijsten worden ingelezen... Klaar
Boom van vereisten wordt opgebouwd... Klaar
libqt3-headers is reeds de nieuwste versie.
0 pakketten opgewaardeerd, 0 nieuwe pakketten geïnstalleerd, 0 verwijderen en 0 niet opgewaardeerd.

(it says libqt3-headers is already the newest version.)
a simple workaround would be to use 'make menuconfig' instead (text-based), but maybe there's a more proper way.

[RubenGonc]

root@ubuntu:/usr/src/linux# sudo make xconfig
*
* Unable to find the QT installation. Please make sure that the
* QT development package is correctly installed and the QTDIR
* environment variable is set to the correct location.
*
make[1]: *** [scripts/kconfig/.tmp_qtcheck] Fout 1
make: *** [xconfig] Fout 2


and


root@ubuntu:/usr/src/linux# sudo apt-get install libqt3-headers
Pakketlijsten worden ingelezen... Klaar
Boom van vereisten wordt opgebouwd... Klaar
libqt3-headers is reeds de nieuwste versie.
0 pakketten opgewaardeerd, 0 nieuwe pakketten geïnstalleerd, 0 verwijderen en 0 niet opgewaardeerd.

(it says libqt3-headers is already the newest version.)


Yes, I forgot to mention to install libqt3-mt-dev. How-To updated. Or you can use "sudo make menuconfig" or "sudo make gconfig" (you need gtk+-2.0, glib-2.0 and libglade-2.0) instead.

[berserker]

My Logitech MX1000 laser mouse works in 2.6.13.4 but not in 2.6.14 (with or without the patch). Here's my xorg.conf:

Section "InputDevice"
Identifier "Configured Mouse"
Driver "mouse"
Option "Protocol" "evdev"
Option "Dev Name" "Logitech USB RECEIVER" #cat /proc/bus/input/devices
Option "Dev Phys" "usb-0000:02:0c.2-2.2/input0" #cat /proc/bus/input/devices
Option "Buttons" "12"
Option "ZAxisMapping" "11 12 10 9"
Option "Resolution" "800"
Option "Emulate3Buttons" "false"
EndSection

Here's the output of /var/log/Xorg.0.log:

(**) Configured Mouse: Protocol: evdev
(**) Option "CorePointer"
(**) Configured Mouse: Core Pointer
(**) Option "Buttons" "12"
(**) Option "Emulate3Buttons" "false"
(**) Option "ZAxisMapping" "11 12 10 9"
(**) Configured Mouse: ZAxisMapping: buttons 11, 12, 10 and 9
(**) Configured Mouse: Buttons: 12
(II) XINPUT: Adding extended input device "Configured Mouse" (type: MOUSE)
(II) XINPUT: Adding extended input device "evdev brain" (type: evdev brain)
(II) XINPUT: Adding extended input device "Configured Mouse" (type: MOUSE)
(II) XINPUT: Adding extended input device "Generic Keyboard" (type: KEYBOARD)
(II) XINPUT: Adding extended input device "NVIDIA Event Handler" (type: Other)
(**) Option "Dev Name" "Logitech USB RECEIVER"
(**) Option "Dev Phys" "usb-0000:02:0c.2-2.2/input0"
(EE) Configured Mouse: cannot open input device
(**) Option "Dev Name" "Logitech USB RECEIVER"
(**) Option "Dev Phys" "usb-0000:02:0c.2-2.2/input0"
(EE) Configured Mouse: cannot open input device
No core pointer

Fatal server error:
failed to initialize core devices

I changed nothing in xorg.conf and used the config file from 2.6.13.4.

The difference between "cat /proc/bus/input/devices" is that there's no "event1" associated with the 2.6.14 kernel:

I: Bus=0003 Vendor=046d Product=c50e Version=2510
N: Name="Logitech USB RECEIVER"
P: Phys=usb-0000:02:0c.2-2.2/input0
H: Handlers=event1 mouse0
B: EV=7
B: KEY=ffff0000 0 0 0 0 0 0 0 0
B: REL=143


Any ideas?

[RubenGonc]

My Logitech MX1000 laser mouse works in 2.6.13.4 but not in 2.6.14 (with or without the patch). Here's my xorg.conf:

Section "InputDevice"
Identifier "Configured Mouse"
Driver "mouse"
Option "Protocol" "evdev"
Option "Dev Name" "Logitech USB RECEIVER" #cat /proc/bus/input/devices
Option "Dev Phys" "usb-0000:02:0c.2-2.2/input0" #cat /proc/bus/input/devices
Option "Buttons" "12"
Option "ZAxisMapping" "11 12 10 9"
Option "Resolution" "800"
Option "Emulate3Buttons" "false"
EndSection

Here's the output of /var/log/Xorg.0.log:

(**) Configured Mouse: Protocol: evdev
(**) Option "CorePointer"
(**) Configured Mouse: Core Pointer
(**) Option "Buttons" "12"
(**) Option "Emulate3Buttons" "false"
(**) Option "ZAxisMapping" "11 12 10 9"
(**) Configured Mouse: ZAxisMapping: buttons 11, 12, 10 and 9
(**) Configured Mouse: Buttons: 12
(II) XINPUT: Adding extended input device "Configured Mouse" (type: MOUSE)
(II) XINPUT: Adding extended input device "evdev brain" (type: evdev brain)
(II) XINPUT: Adding extended input device "Configured Mouse" (type: MOUSE)
(II) XINPUT: Adding extended input device "Generic Keyboard" (type: KEYBOARD)
(II) XINPUT: Adding extended input device "NVIDIA Event Handler" (type: Other)
(**) Option "Dev Name" "Logitech USB RECEIVER"
(**) Option "Dev Phys" "usb-0000:02:0c.2-2.2/input0"
(EE) Configured Mouse: cannot open input device
(**) Option "Dev Name" "Logitech USB RECEIVER"
(**) Option "Dev Phys" "usb-0000:02:0c.2-2.2/input0"
(EE) Configured Mouse: cannot open input device
No core pointer

Fatal server error:
failed to initialize core devices

I changed nothing in xorg.conf and used the config file from 2.6.13.4.

The difference between "cat /proc/bus/input/devices" is that there's no "event1" associated with the 2.6.14 kernel:

I: Bus=0003 Vendor=046d Product=c50e Version=2510
N: Name="Logitech USB RECEIVER"
P: Phys=usb-0000:02:0c.2-2.2/input0
H: Handlers=event1 mouse0
B: EV=7
B: KEY=ffff0000 0 0 0 0 0 0 0 0
B: REL=143


Any ideas?


Go to http://blog.blackdown.de/page/2/ and then search for the topic "Logitech MX1000 Configuration". Hope it helps you:)

[berserker]

Go to http://blog.blackdown.de/page/2/ and then search for the topic "Logitech MX1000 Configuration". Hope it helps you:)

Thank you for the suggestion. I tried that but no luck. I think it's either that the new kernel handles USB mice differently or the NVIDIA drivers (I've tried both 7667 and 7676) aren't compatible with the new kernel.

[Poul]

can someone tell us or link to information about it?
Why/when should i bother?
What is so good about and what is changed comparing to the current one?
THX in advance.

[Rob2687]

Got it working but as usuall with compiling a new kernel like this a lot of things stop working or need fixing after this.

I really hope they put this in the repositories soon.

[Dracontopes]

Yes, I forgot to mention to install libqt3-mt-dev. How-To updated. Or you can use "sudo make menuconfig" or "sudo make gconfig" (you need gtk+-2.0, glib-2.0 and libglade-2.0) instead.

Thanks, I'm now happily compiling the kernel. :D

[Hyun]

*In a terminal make sure you are in /usr/src/linux with full root access.

tss tss...

No need to be root to compile the kernel. You need to be root
to _install_ the kernel, but that's different.
;)

just add your user to src group
sudo addsuer username src
logout/login, and start How To. :razz:

[Tomasz]

All went fine here. Only thing that stopped working is usplash (completely blank screen), not that I care, it's probably due to my problematic LCD screen... Ubuntu seems to boot faster now, or maybe it's just my imagination. Anyway, thanks for the guide, that was my first kernel compilation!

PS How do I switch the root account off now?

[HotShot]

I have some errors during the patch and I´m unsure what to do

Sample:

can't find file to patch at input line 853
Perhaps you used the wrong -p or --strip option?
The text leading up to this was:
--------------------------
|Index: linux-2.6.14-ck1/kernel/exit.c
|================================================= ==================
|--- linux-2.6.14-ck1.orig/kernel/exit.c 2005-10-28 13:03:05.000000000 +1 000
|+++ linux-2.6.14-ck1/kernel/exit.c 2005-10-28 13:06:03.000000000 +1000
--------------------------
File to patch:


just copy and paste the commands from the how-to and get some of these messages.

Edit: problem solved. the kernel-achive was corrupt :)

[Tomasz]

Does anyone else have problems with pmount? Removable drives and cdroms are only accessible through /media/cdrom0 (and so on), they don't appear on the desktop.

I'll try to translate the warning messeges when I click the icons in "Computer":

Warning: device /dev/hdc is already taken care of by /etc/fstab, given label has been ignored
mount: block device /dev/hdc is write-protected, mounting read-only
mount: /dev/hdc already mounted or /media/cdrom0 busy
mount: according to mtab, /dev/hdc is already mounted on /media/cdrom0
Error: cannot run pmount


Please, can anyone help me with this?

[RubenGonc]

tss tss...
;)

just add your user to src group
sudo addsuer username src
logout/login, and start How To. :razz:


Yes...and you think that is more easy to add the user to the src group,logout,login or to gain full root access?...
And I am not so sure if that will work because the deb file will be created at /usr/src... but I don't know... What I posted work..I didn't say that there aren't more ways of doing this:P But I will try it later!Thanks;)

[RubenGonc]

can someone tell us or link to information about it?
Why/when should i bother?
What is so good about and what is changed comparing to the current one?
THX in advance.


http://www.internetnews.com/dev-news/article.php/3560061

http://wiki.kernelnewbies.org/LinuxChanges

http://www.kernel.org/pub/linux/kernel/v2.6/ChangeLog-2.6.14

[theh0g]

Cool, it worked...had problems with Nvidia (of course) but after second reinstall of drivers it worked fine. All I have to do now is get my other disk to work, it says it's either already mounted or busy. Nice HOWTO! :)

[johannes]

Thanks for this guide. :) All worked flawlessly exept for installing the ATI drivers afterwards... Has anybody had any success with this? If so, did you need to do anything exept the things included in this how-to?

[crypto178]

Does anyone else have problems with pmount? Removable drives and cdroms are only accessible through /media/cdrom0 (and so on), they don't appear on the desktop.

I'll try to translate the warning messeges when I click the icons in "Computer":

Warning: device /dev/hdc is already taken care of by /etc/fstab, given label has been ignored
mount: block device /dev/hdc is write-protected, mounting read-only
mount: /dev/hdc already mounted or /media/cdrom0 busy
mount: according to mtab, /dev/hdc is already mounted on /media/cdrom0
Error: cannot run pmount


Please, can anyone help me with this?
Same problem here, slightly annoying, but a temporary fix is to run sudo umount '/media/drivename' and then double clicking the icon again. I have to do it every time I plug in the removable media again.

[kaamos]

Same here, pmount isn't working.

Also, I can't use the virtual terminals(?). Meaning that pressing ctrl+alt+f1 - f6 all show just a gray screen. When booting the screen is blank until gdm is started..

Any ideas how to fix these problems?

[Capilano]

:confused: Thank you for a well made Howto. It installed perfectly on my system with one annoyance: On the splash menu it says that the system failed to load local filesystems. I have no acces to partition that were previously accessible with breezy. Yhe fstab is the same as before; same for mtab. help please !

[suRoot]

Same problem here, slightly annoying, but a temporary fix is to run sudo umount '/media/drivename' and then double clicking the icon again. I have to do it every time I plug in the removable media again.


Same problem here. Same fix / hack.

[Dracontopes]

Same problem here. Shame, back to 2.6.12 kernel for now untill there is a simple fix :)

[Technoviking]

Same problem here. Shame, back to 2.6.12 kernel for now untill there is a simple fix :)
Same here.

I have been trying to add the ck performance patch to the Linux source (2.6.12) w/ Ubuntu patch added and tweak more for my Pentium M, but no luck yet.

[TakisX]

It compiled easily but when i install the nvidia drivers with tseliots how to as 100 times before it says everything ok but gnome does not start .What to do ?

I got back 2.6.12 and reinstalled nvidia drivers and it works
Maybe the new kernel does not like nvidia drivers

[berserker]

It compiled easily but when i install the nvidia drivers with tseliots how to as 100 times before it says everything ok but gnome does not start .What to do ?

I got back 2.6.12 and reinstalled nvidia drivers and it works
Maybe the new kernel does not like nvidia drivers

What does your /var/log/Xorg.0.log say?

KDE wouldn't start for me because the USB mouse couldn't be found.

[suRoot]

I somehow managed to get pmount to work. I went back and recompiled the kernel again, this time making some changes to the config based on a config I'd used before, optimized for my thinkpad t40p. After rebooting, I see that pmount works again. Not sure why - I made quite a few changes to the config file - but my intent wasn't to fix pmount :-).

Anyways, I'm attaching my config in case someone smarter than I wants to compare the stock config to see what's missing.

My only problem now is that usplash, or whatever it's called, (I like Ubuntu, but I've spent a lot of time in the RedHat world) doesn't seem to work. I just get text when booting, but that's okay.

[GoA]

Logitech MX700 USB-mouse and couldn't get x running.

here's log

(**) Option "CoreKeyboard"
(**) Generic Keyboard: Core Keyboard
(**) Option "Protocol" "standard"
(**) Generic Keyboard: Protocol: standard
(**) Option "AutoRepeat" "500 30"
(**) Option "XkbRules" "xorg"
(**) Generic Keyboard: XkbRules: "xorg"
(**) Option "XkbModel" "pc105"
(**) Generic Keyboard: XkbModel: "pc105"
(**) Option "XkbLayout" "fi"
(**) Generic Keyboard: XkbLayout: "fi"
(**) Option "CustomKeycodes" "off"
(**) Generic Keyboard: CustomKeycodes disabled
(**) Option "Protocol" "evdev"
(**) Configured Mouse: Protocol: evdev
(**) Option "Buttons" "10"
(==) Configured Mouse: Emulate3Buttons, Emulate3Timeout: 50
(**) Option "ZAxisMapping" "4 5"
(**) Configured Mouse: ZAxisMapping: buttons 4 and 5
(**) Configured Mouse: Buttons: 10
(**) Configured Mouse: Protocol: evdev
(**) Option "CorePointer"
(**) Configured Mouse: Core Pointer
(**) Option "Buttons" "10"
(==) Configured Mouse: Emulate3Buttons, Emulate3Timeout: 50
(**) Option "ZAxisMapping" "4 5"
(**) Configured Mouse: ZAxisMapping: buttons 4 and 5
(**) Configured Mouse: Buttons: 10
(II) XINPUT: Adding extended input device "Configured Mouse" (type: MOUSE)
(II) XINPUT: Adding extended input device "evdev brain" (type: evdev brain)
(II) XINPUT: Adding extended input device "Configured Mouse" (type: MOUSE)
(II) XINPUT: Adding extended input device "Generic Keyboard" (type: KEYBOARD)
(**) Option "Dev Name" "Logitech USB Receiver"
(**) Option "Dev Phys" "usb-*/input0"
(EE) Configured Mouse: cannot open input device
(**) Option "Dev Name" "Logitech USB Receiver"
(**) Option "Dev Phys" "usb-*/input0"
(EE) Configured Mouse: cannot open input device
No core pointer

Fatal server error:
failed to initialize core devices

Please consult the The X.Org Foundation support
at http://wiki.X.Org
for help.
Please also check the log file at "/var/log/Xorg.0.log" for additional information.

[RubenGonc]

:confused: Thank you for a well made Howto. It installed perfectly on my system with one annoyance: On the splash menu it says that the system failed to load local filesystems. I have no acces to partition that were previously accessible with breezy. Yhe fstab is the same as before; same for mtab. help please !

You forgot to activate the kernel support for your filesystem (ext3, reiserfs). Try to recompile with support for your filesystem enabled.

[RubenGonc]

For people experimenting problems with the Nvidia drivers:

The drivers worked for me! What I did was remove the "nvidia-glx" package. Then I installed the ".run" nvidia drivers installer (you may have to change xorg.conf as described in http://www.ubuntuforums.org/showthre...ghlight=nvidia).Then you restart X and everything shoul work.

Sorry but I can't help with ATI.

[crypto178]

I somehow managed to get pmount to work. I went back and recompiled the kernel again, this time making some changes to the config based on a config I'd used before, optimized for my thinkpad t40p. After rebooting, I see that pmount works again. Not sure why - I made quite a few changes to the config file - but my intent wasn't to fix pmount :-).

Anyways, I'm attaching my config in case someone smarter than I wants to compare the stock config to see what's missing.

My only problem now is that usplash, or whatever it's called, (I like Ubuntu, but I've spent a lot of time in the RedHat world) doesn't seem to work. I just get text when booting, but that's okay.
Thanks for the attachment, I've generated a difference chart between my config and yours (using this site : http://www.ranks.nl/cgi-bin/ranksnl/tools/difference.cgi ). I'll investigate tomorrow which setting might be behind this mystery.
You can see the differences list here : http://hlqg.free.fr/autres/diffs.html (mine is on the left, changes are marked with a | )
Upsplash works here :confused:

[berserker]

Logitech MX700 USB-mouse and couldn't get x running.

I had the same problem with a Logitech MX1000 USB mouse.

[nastya]

You forgot to activate the kernel support for your filesystem (ext3, reiserfs). Try to recompile with support for your filesystem enabled.

I had recieved the same error on boot. However I know that I have fat32 support enabled. It's only my other hard drive that is affected as I have a fat32 partition on my drive with ubuntu and it mounts just fine. I tried to mount the partitions on the offending drive in terminal, but I recieve an error stating that it is already mounted or busy. I then tried to unmount it and it tells me that neither partition is mounted. I hope that I just overlooked something, as I don't want to go back to not having dri support for my video card (via unichrome).

[Drifter]

I see Linux.org a new kernel released: 2.6.14-git4, is this the new Vanilla kernel if so where is the patchset. I need the new one in hopes that freeze ups while online will be taken care of. Please inform me if this is so and how to get the info.

[Drifter]

This How-To will guide you through the compilation/installation of the 2.6.14 Vanilla Kernel (more recent than the one distribuited with Ubuntu 5.10) with the performance patchset by Kon Colivas.

1-Download what is needed:

Type in the command line:



Then download the following files to your Home directory:

http://www.kernel.org/pub/linux/kernel/v2.6/linux-2.6.14.tar.bz2

http://ck.kolivas.org/patches/2.6/2.6.14/2.6.14-ck1/patch-2.6.14-ck1.bz2

2-Now, lets unpack the kernel source to /usr/src:

*Copy the source to /usr/src:

sudo cp linux-2.6.14.tar.bz2 /usr/src

*Change to /usr/src:

cd /usr/src

*Unpack it:

sudo tar -xvjf linux-2.6.14.tar.bz2

*Now lets change the created directory name:

sudo mv linux-2.6.14/ linux-2.6.14cK1

*Remove the symlink if there is one directory called linux:

sudo rm -rf linux

*Make the new symlink pointing for our 2.6.14 kernel source:

sudo ln -s /usr/src/linux-2.6.14cK1 linux

3-Now it's time to patch the kernel:

*Change to /usr/src/linux:

cd /usr/src/linux

*Switch to root user:

sudo -s -H
Password: <specify user password>

*Apply the patch:

sudo bzcat /home/username/patch-2.6.14-ck1.bz2| patch -p1

4-Kernel configuration:

*Import the configuration of the running kernel:

uname -r To see what kernel are you running (in my case it is 2.6.12-9-686).

sudo cp /boot/config-2.6.12-9-686 .config To copy the config file and use it as base for the new kernel configuration (Don't forget to choose the correct config file).

*Configure the kernel:

sudo make xconfig

*While you may tweak your kernel configuration to your needs I will sugest you some tweaks:

In "General Setup" activate:

-Support for paging of anonymous memory (swap)
--Support for prefetching swapped memory

In "Processor type and features":

-Processor family Choose the model of your processor.

Activate:

-Preemption Model
--Voluntary Kernel Preemption (Desktop)

-High Memory Support
--off -if you have less than 1 GB of RAM
--1GB Low Memory Support -if you have 1GB of RAM
--4GB -if you have more than 1GB of RAM

-Timer frequency
--1000 Hz

In "Device drivers" go to "Block devices" and in "IO Schedulers" leave only the "CFQ I/O scheduler" activated, which provides the best performance.

In "Kernel hacking" uncheck "Kernel debugging".

Ctrl+S to save the kernel configuration and then close the window.

5-Let's build the kernel:

*In a terminal make sure you are in /usr/src/linux with full root access.

We will build a ".deb" file that can be installed in our Ubuntu system, using make-kpkg.

*In a terminal type:

make-kpkg clean

make-kpkg -initrd --revision=ck1 kernel_image

If there wasn't errors this will build the kernel and a ".deb" file will be created at /usr/src.
*To install it:

sudo dpkg -i kernel-image-2.6.14*.deb


6-Reboot and everything should be running ok!
*Try:

uname -r

to see that you are running the new kernel.


7-Nvidia graphic cards users:

To install Nvidia driver follow the guide (no need to install linux-headers as we have the source in /usr/src/linux:

http://www.ubuntuforums.org/showthread.php?t=57368&highlight=nvidia


If you have any problem ask here.

Hope you like it!

Rúben Gonçalves:p

I did everything just like the guide said it all worked, but when I checked to see if I was using the new kernel, it said I was using 2.6.12-9-386. I rebooted aftwards and then it said this.

[RubenGonc]

There is a new version of ck patchset.I will try it today...maybe it solves the pmount problem:???:
Anyone tried it?

[RubenGonc]

No:( It doesn't solves the pmount problem. Did anybody found the solution for this???
crypto178 did you realize what was the option to activate in the kernel config?

[RubenGonc]

Very easy solution:D

Download http://darkstar.ist.utl.pt/ubuntu/archive/pool/main/p/pmount/pmount_0.9.6-1_i386.deb

Install doing
sudo dpkg -i pmount_0.9.6-1_i386.deb


I think it will work!:D

[btarlinian]

I'm a newb, so forgive me for my ignorance.
What exactly is the ck Patchset. Is it necessary, or does it just improve performance?

[crypto178]

Very easy solution:D

Download http://darkstar.ist.utl.pt/ubuntu/archive/pool/main/p/pmount/pmount_0.9.6-1_i386.deb

Install doing
sudo dpkg -i pmount_0.9.6-1_i386.deb


I think it will work!:D
It's working !
Thanks :)

[_Pete_]

I did according to this how-to and everyrthing went well. There's a little problem:

In X/gnome if I don't move the mouse for a while the pointer freezes. Only way to get it move again is to go to the text console by pressing alt-Fx and then come back to X ...

Any ideas what might be causing this ? It's a bit annoying ...

[nszabolcs]

What exactly is the ck Patchset. Is it necessary, or does it just improve performance?
Not necessary.
http://members.optusnet.com.au/ckolivas/kernel/

[_Pete_]

Not necessary.
http://members.optusnet.com.au/ckolivas/kernel/


Btw, how are you supposed to notice this the boost to the performance that's caused by this patch? At least I didn't (yet) notice any difference. Maybe the difference can be noticed better with low end machines? I'm running P4@2.6GHz / 1G mem / nvidia6600 gt machine...

And any hints to that mouse problem ? :)

[techstop]

You forgot to activate the kernel support for your filesystem (ext3, reiserfs). Try to recompile with support for your filesystem enabled.

At what step do I need to start again to recompile? I tried starting from

"sudo make xconfig"

but I got this error;

make: *** No rule to make target `xconfig'. Stop.

Do I need to start the whole process again? I have rebooted into my old kernel because the new one wouldn't let X start, and also have the problem of the filesystems failing to load. Cheers.

[RubenGonc]

At what step do I need to start again to recompile? I tried starting from

"sudo make xconfig"

but I got this error;

make: *** No rule to make target `xconfig'. Stop.

Do I need to start the whole process again? I have rebooted into my old kernel because the new one wouldn't let X start, and also have the problem of the filesystems failing to load. Cheers.

Go to /usr/src/linux (cd /usr/src/linux) and then:

sudo -s -H
<password>
make clean
make xconfig

[Hitman1]

The compile finishes with no errors but there is no .deb package generated. here is what I get last:

make[1]: Leaving directory `/usr/src/linux-2.6.14'
/usr/bin/make \
ARCH=i386 prepare
make[1]: Entering directory `/usr/src/linux-2.6.14'
CHK include/linux/version.h
SPLIT include/linux/autoconf.h -> include/config/*
make[1]: Leaving directory `/usr/src/linux-2.6.14'
echo done > stamp-kernel-configure
echo done > stamp-configure

Anyone know what could the problem for not building the .deb file? I searched the whole filesystem for it incase it was placed somewhere else, but it is not there.

[Wally68]

Thanks, Ruben for this how-to. The instructions led me down a path of errorless compilation and a headache free kernel upgrade. It's not the first kernel I've compiled, but it is the first time it worked for me on the first try. Now that I've got that far, I'll do it again, but try removing a few things (such as old cdrom support, extra network drivers, etc). Very clear, concise, and to the point. Thanks again.
Wally

[nastya]

If anyone who's compiled this kernel with the via unichrome support is reading this and you have recieved no errors, would you be willing to email me the deb? I cannot escape my mounting filesystems error with any of the debs I have compiled myself and would really like to have support for my video card.

[Dany]

:confused: Thank you for a well made Howto. It installed perfectly on my system with one annoyance: On the splash menu it says that the system failed to load local filesystems. I have no acces to partition that were previously accessible with breezy. Yhe fstab is the same as before; same for mtab. help please !

I have the same problem, my SATA drive dissapeared after the install, but I dont get the "system failed to load local filesystem" message. ex If I create a ext3 filsystem on the drive and mount it only root has access to the mounted dir...and when I reboot its not mounted anymore... and when I try to umount it, it says " /dev/sda1 not mounted". Then I try and mount it again and it says "busy or allready mounted".
So if somebody could please paste a correct config-file (x config)of the kernel compilation , so that all disks SATA,IDE or whatever works properly I and many other n00bs would be greatful :) other then this exelent guide ;)

[sk545]

Keep in mind people, that just going to kernel 2.6.14 all of a sudden, will make things not work even if you copy over the config from the old kernel.

[techstop]

Go to /usr/src/linux (cd /usr/src/linux) and then:

sudo -s -H
<password>
make clean
make xconfig

Many thanks!

OK. one more question. I'm not confident in getting this thing working at my current level of knowledge, how do I "uninstall" the broken kernel? ie. so that my old kernel is the default in grub etc. Cheers

[desht]

I have the same problem, my SATA drive dissapeared after the install, but I dont get the "system failed to load local filesystem" message. ex If I create a ext3 filsystem on the drive and mount it only root has access to the mounted dir...and when I reboot its not mounted anymore... and when I try to umount it, it says " /dev/sda1 not mounted". Then I try and mount it again and it says "busy or allready mounted".
So if somebody could please paste a correct config-file (x config)of the kernel compilation , so that all disks SATA,IDE or whatever works properly I and many other n00bs would be greatful :) other then this exelent guide ;)

I also have this problem, and the system I'm using doesn't have SATA - just a pair of IDE drives. There's obviously some more general problem here.

I have my /boot, /, /home & swap partititions on /dev/hdb, and I have a Windows partition & a reiserfs data partition on /dev/hda. 2.6.12 (i.e. stock with 5.10) works fine. I've tried to build both 2.6.13.4 and 2.6.14 kernels, and both have the same problem: the /dev/hdb* partitions mount OK, but /dev/hda* partitions won't mount, failing with the "busy or already mounted" error.

I definitely have the right filesystem support; /home is also reiserfs and it mounts fine. Loopback device mounting (e.g. mkisofs and "mount -o loop ...") also works fine.

New kernel config is based on /boot/config-2.6.12-9-386 + "make oldconfig".

[Saiboogu]

Many thanks!

OK. one more question. I'm not confident in getting this thing working at my current level of knowledge, how do I "uninstall" the broken kernel? ie. so that my old kernel is the default in grub etc. Cheers

The kernel should be installed as "kernel-image-2.6.14-ck1" but just to make sure, search for it first...

saiboogu@GREEN:~$ dpkg --list | grep 2.6.14
ii kernel-image-2.6.14-ck1 ck1 Linux kernel binary image for version 2.6.14
saiboogu@GREEN:~$ sudo apt-get remove kernel-image-2.6.14-ck1

That should take care of it.

Unfortunately I've had to do the same a few times now. :( Only my ext3 / partition will mount, my Reiser "data" disk won't mount, or my Samba shares. Haven't done a lot in terms of troubleshooting yet - Anyone had any luck with this? The pmount .deb that was shared here was just supposed to fix problems with automounting removable devices, right? Seems my problem is more with the core "mount" command.

And yes, I made sure I enabled reiserfs support in my config.. Going to go back through the config again this evening to make sure I did not miss anything obvious.

[darknuala]

When I did this, I followed the instructions to a "T". I had no errors whatsoever. It worked like a charm. Until I rebooted and tried to log in. It said that my user id was navyn, but my home directory no longer existed. I tried logging in anyways, because it gave me that option, and threw me to an .xsession error, saying that it could not find .gnome2lib and kicked me back to the login screen. I logged in to failsafe, and it threw me to the root directory as my home directory. When I browsed to see what was in /home, it was completely empty. I logged out, and logged into kde and I get an error about dcop something or other, then that $HOME/.drmc had the wrong file permissions, and should belong to the users group.

[jg123]

I have a similar problem in that my external usb drive won't mount after installing 2.6.14-cks2. It works fine if I reload my 2.6.12 kernel. I am using autofs4 to mount the drive and automatically unmount it after a minute of activity. I get these errors in my log:

26 localhost automount: >> mount: /dev/sda1 already mounted or /misc/usbmaxtor busy
26 localhost automount: mount(generic): failed to mount /dev/sda1 (type vfat) on /misc/usbmaxtor
26 localhost automount: failed to mount /misc/usbmaxtor

I'm also getting these:
1 localhost kernel: [154356.036678] device-mapper: dm-linear: Device lookup failed

[techstop]

The kernel should be installed as "kernel-image-2.6.14-ck1" but just to make sure, search for it first...

saiboogu@GREEN:~$ dpkg --list | grep 2.6.14
ii kernel-image-2.6.14-ck1 ck1 Linux kernel binary image for version 2.6.14
saiboogu@GREEN:~$ sudo apt-get remove kernel-image-2.6.14-ck1

That should take care of it.

Unfortunately I've had to do the same a few times now. :( Only my ext3 / partition will mount, my Reiser "data" disk won't mount, or my Samba shares. Haven't done a lot in terms of troubleshooting yet - Anyone had any luck with this? The pmount .deb that was shared here was just supposed to fix problems with automounting removable devices, right? Seems my problem is more with the core "mount" command.

And yes, I made sure I enabled reiserfs support in my config.. Going to go back through the config again this evening to make sure I did not miss anything obvious.

That worked perfectly, thanks for your help. I'm just going to stick with my 2.6.12-9-k7 kernel for now.

[kashms]

I followed the instructions and succeeded in installing the new kernel with the patch. I even was able to build a nvidia module deb to install. However a couple of things, showstoppers in fact, did not work. So far I discovered that the side scolling on my touchpad is not working, and more importantly suspend stopped working as well. I'm just going to stick with 2.6.12-9-686 for now.

[z!cHz@cH]

This howto worked great for me:p

People with notebooks:

This kernel wil stop cooking ur CPU. It handels CPU stepping a lot better now.

Thanks.

[sfabel]

Using the above method, my centrino notebook stopped its wireless support. There's always an error "unknown device" when I try to activate my wlan interface.

Any thoughts? Maybe these offical patches should be applied after all?

Thanks,
Stephan

[tseliot]

Using the above method, my centrino notebook stopped its wireless support. There's always an error "unknown device" when I try to activate my wlan interface.

Any thoughts? Maybe these offical patches should be applied after all?

Thanks,
Stephan
Perhaps you need ndiswrapper (I'm not sure though). If you need it here is a way to compile its modules http://www.ubuntuforums.org/showthread.php?t=85064.

[timczer]

I followed this guide and everything for the most part seems to work. I have an external hard drive that isn't mounted but so far that appears to be it.

my question is, once completed, what can be removed from the /usr/src folder. I have 432mb in that folder after the compiling. Can any of this be removed? That is a lot of memory to give up.

[prasys]

It really works , thanks for the superb guide

the .deb file is located in the /usr/src folder

[shartrec]

All went fine here. Only thing that stopped working is usplash (completely blank screen), not that I care, it's probably due to my problematic LCD screen... Ubuntu seems to boot faster now, or maybe it's just my imagination. Anyway, thanks for the guide, that was my first kernel compilation!

PS How do I switch the root account off now?

This and the problem mentioned by another post about the tty consoles being blank are both related to the same problem. Easy fix is remove any vga=nnn bootparam from your grub menu.lst file.

The source of the problem is that the CONFIG_FB_VESA option becomes deselected when you start xconfig. You need to manually select the option in xconfig. I think this occurs because in the standard ubuntu config this is set ot 'm' and I think this is not supported, at least not in xconfig. It needs to be 'y'.

[FiggyG]

It worked for me :D Just gotta iron out a few things as usual, like choosing the correct drivers to get my SATA drive workin :lol: This is the first kernel I've compiled on this machine and on Ubuntu. I've done kernels on basically every distro besides Ubuntu. I was pretty damn slick at it with Gentoo... ahhh good old days :)

Oh, forgot, it busted direct rendering with the ATI drivers. Nothing I can't fix ;)

[fukusayo]

It worked for me :D Just gotta iron out a few things as usual, like choosing the correct drivers to get my SATA drive workin :lol: This is the first kernel I've compiled on this machine and on Ubuntu. I've done kernels on basically every distro besides Ubuntu. I was pretty damn slick at it with Gentoo... ahhh good old days :)

Oh, forgot, it busted direct rendering with the ATI drivers. Nothing I can't fix ;)
I also lost 3D with my ATI card. However, I was unable to fix it. When you are successful, perhaps you could help a newbie out? Thanks in advance.

[blakken]

Using the above method, my centrino notebook stopped its wireless support. There's always an error "unknown device" when I try to activate my wlan interface.

Any thoughts? Maybe these offical patches should be applied after all?

Thanks,
Stephan
I downloaded the debian package for the kernel and let's say 2 things (and this without any kernel compilation matters as it is already compiled !) :
-first on my laptop my eth1(wireless device) stopped working
-second the scrolling isn't working with my mouse(and I didn't change my xorg though!!)

but for a change when mounting devices such as external hard drive or bluetooth key both of them are working!Didn't check the sound or my cd yet!
By the way anybody have any idea what's new about pocketpc stuff(synce) kernel adds and how to use it,and what is suppose the patch suppose to do theoritically for the pentium m?

[BIGtrouble77]

When I try to install the ATI drivers this is what I get:

bob@bobbyhotep:/lib/modules/fglrx$ sudo sh make_install.sh
- creating symlink
- recreating module dependency list
- trying a sample load of the kernel module
FATAL: Error inserting fglrx (/lib/modules/2.6.14-ck1/kernel/drivers/char/drm/fglrx.ko): Unknown symbol in module, or unknown parameter (see dmesg)
failed.

Hopefully someone knows what to do about this.

[zarathustra]

I get this error when trying to compile the new kernel:

drivers/built-in.o: In function `init_nandsim':
nandsim.c:(.text+0x7bca8): undefined reference to `nand_flash_ids'
drivers/built-in.o: In function `ns_init_module':
: undefined reference to `nand_scan'
drivers/built-in.o: In function `ns_init_module':
: undefined reference to `nand_default_bbt'
drivers/built-in.o: In function `ns_cleanup_module':
nandsim.c:(.exit.text+0x35e): undefined reference to `nand_release'
make[1]: *** [.tmp_vmlinux1] Error 1
make[1]: Leaving directory `/usr/src/linux-2.6.14cK1'
make: *** [stamp-build] Error 2

[StefanoCole]

I just wanted to know the meaning of the term "Vanilla" used for a Kernel.

Thanks, Stefano

[Lifesteeler]

My wlan stopped working with this kernel. What can I do about it? Also, should I use the "Pentium M" choice with my Celeron M?

[tare]

Logitech MX700 USB-mouse and couldn't get x running.


First of all, thanks for a great HowTo!
I got finally my MX700 to work. I had configured evdev as loadable module but could not see it with
lsmod | grep evdev
and /proc/bus/input/devices had no "eventx" as a mouse handler. So I changed kernel .config to look like this:

#
# Userland interfaces
#
CONFIG_INPUT_MOUSEDEV=y
CONFIG_INPUT_MOUSEDEV_PSAUX=y
CONFIG_INPUT_MOUSEDEV_SCREEN_X=1024
CONFIG_INPUT_MOUSEDEV_SCREEN_Y=768
CONFIG_INPUT_JOYDEV=m
CONFIG_INPUT_TSDEV=m
CONFIG_INPUT_TSDEV_SCREEN_X=240
CONFIG_INPUT_TSDEV_SCREEN_Y=320
CONFIG_INPUT_EVDEV=y
# CONFIG_INPUT_EVBUG is not set


Recompile and voila! I can define evdev as Protocol and all my buttons and wheel work! :lol:

[tseliot]

I just wanted to know the meaning of the term "Vanilla" used for a Kernel.

Thanks, Stefano

Vanilla means "plain and simple" and it means that it has no specific features (i.e. it is not adapted to any distro in particular) and it is not patched (at least in theory) and of course it is the source code made by Linux Torvalds & colleagues.

Somebody correct me if I'm wrong, please.

[benplaut]

anyone know how to get restricted modules working? atheros based wireless, not being found by ifconfig :(

<<edit>>

this thread:
http://ubuntuforums.org/showthread.php?t=75451
totally rocks

[GoA]

First of all, thanks for a great HowTo!
I got finally my MX700 to work. I had configured evdev as loadable module but could not see it with
lsmod | grep evdev
and /proc/bus/input/devices had no "eventx" as a mouse handler. So I changed kernel .config to look like this:

#
# Userland interfaces
#
CONFIG_INPUT_MOUSEDEV=y
CONFIG_INPUT_MOUSEDEV_PSAUX=y
CONFIG_INPUT_MOUSEDEV_SCREEN_X=1024
CONFIG_INPUT_MOUSEDEV_SCREEN_Y=768
CONFIG_INPUT_JOYDEV=m
CONFIG_INPUT_TSDEV=m
CONFIG_INPUT_TSDEV_SCREEN_X=240
CONFIG_INPUT_TSDEV_SCREEN_Y=320
CONFIG_INPUT_EVDEV=y
# CONFIG_INPUT_EVBUG is not set


Recompile and voila! I can define evdev as Protocol and all my buttons and wheel work! :lol:

Could I get more specific details about the how did you do that. :D I'm a noob you know. :)

[Lifesteeler]

My network card is Intel Pro Wireless 2200BG, i found out that i should use ndiswrapper, but how do i do this? Should i include it in the kernel when building or what? Need help, cause i want my speedy kernel to work...

[tare]

Could I get more specific details about the how did you do that. :D I'm a noob you know. :)

Just do your normal make xconfig and go to "Input device support - Generic input layer" and select "Event interface" to be built in, not as module.
And recompile!

[jewelz]

What about conexant usb adsl modems? There seems to be built-in drivers, but i don't know how to make them work!

[quintok]

hi, is there anything in this how-to that could/is disabiling ntfs support? I can mount the device fine on the ubuntu 5.1 livecd but following this I cannot load ntfs even if it's built into the kernel (not as a module)

[Maverick911]

Vanilla means "plain and simple" and it means that it has no specific features (i.e. it is not adapted to any distro in particular) and it is not patched (at least in theory) and of course it is the source code made by Linux Torvalds & colleagues.

Somebody correct me if I'm wrong, please.

So does that mean that I'm really not running ubuntu anymore with this kernel?

Wish You Were Here - Pink Floyd

[quintok]

no, it's still ubuntu.

[allans]

Just wanted to say thanks for the great how to, everything fine here and definitely a little faster!

Allan

[berserker]

I got finally my MX700 to work.

Thank you! I couldn't figure out what was going on with my MX1000. For some reason the new 2.6.14 and 2.6.14.1 kernels need EVDEV compiled into the kernel.

[spoilerhead]

if someone would like a try:

http://stud3.tuwien.ac.at/~e0326004/kernel/kernel-headers-2.6.14-sph1-686_2_i386.deb
http://stud3.tuwien.ac.at/~e0326004/kernel/kernel-image-2.6.14-sph1-686_2_i386.deb

read more at:
http://www.ubuntuusers.de/viewtopic.php?t=14688&start=30 (in german)

ck3 patchset included, vesa-tng, speedstep fix for pentium M (Dothan)

for feedback
spoilerhead_remove_m3_(at)_brennercom.net

[Eazy©]

Has anyone found a solution to the "mount problem"?
I cant mount my ntfs partition though I think ntfs support its activated as a module (If I run menuconfig I see that its activated).
If I try to mount it it says that its already mounted or busy.

[tseliot]

So does that mean that I'm really not running ubuntu anymore with this kernel?

Wish You Were Here - Pink Floyd

It's just that you're not using a kernel with Ubuntu patches. Ubuntu is more than kernel patches, it's an entire OS

[ibanez88]

ipw2200 should be built as a module by default. You need to download the firmware from sourceforge (use version 2.2) and extract the tarball into /usr/lib/hotplug/firmware. Worked for me. Don't keep the tarball itself in that folder.

I can boot from this new kernel but I'm getting errors on bootup related to my Realtek 8139 ethernet. Also my SD slot not working (nothing new there, but I thought the new Winbond SD support might help).. I probably don't have the right devices / fstab; nothing shows up in dmesg. Nonetheless any pointers related to this or my Realtek 8139 are appreciated.

THANKS for the how-to!

[ColdWind]

Thanks!
I compiled it with cK5 Patchset and works fine in my laptop (at the moment).

[adamb10]

How big of a performance boost does this kernel provide?

[curtis]

Hi,
I am using 2.6.14.1-cK1 fine here.
That is on a Dell Inspiron 5150.
Just having an issue with the nvidia module.
I end up needing to re-install it every boot up.
It is in /etc/modules though, and modprobing nvidia does not work.
I have a feeling it could be InitNG possibily, will check now.

Any one else running the 2.6.14.x kernel with a Nvidia graphics card?
Even better if they are using InitNG.

[tseliot]

Hi,
I am using 2.6.14.1-cK1 fine here.
That is on a Dell Inspiron 5150.
Just having an issue with the nvidia module.
I end up needing to re-install it every boot up.
It is in /etc/modules though, and modprobing nvidia does not work.
I have a feeling it could be InitNG possibily, will check now.

Any one else running the 2.6.14.x kernel with a Nvidia graphics card?
Even better if they are using InitNG.
Have a look at Point 7 of my guide:
http://www.ubuntuforums.org/showthread.php?t=85064
it explains how to compile the nvidia modules together with your kernel (as packages, just like Ubuntu's restricted modules)

[quintok]

Has anyone found a solution to the "mount problem"?
I cant mount my ntfs partition though I think ntfs support its activated as a module (If I run menuconfig I see that its activated).
If I try to mount it it says that its already mounted or busy.

I'm glad I'm not the only one in this boat :)

[nastya]

I'm glad I'm not the only one in this boat :)

There's definately a few of us with that problem, but no one seems to know what is going on. I've even had 3 or 4 others take a look at my .config, but none of them could find anything wrong.

[synd7]

Cheers for the guide
2.6.14-ck4 is running fine. Except that i no longer have fglrx working. does anyone know how to get it up and running? ive tried the method in the 'kernel compilation for newbies' thread but I end up tith a .deb that requires the fglrx-kernel-common package which is not in the repositories.

Is there some way to build the source for the xorg-driver-fglrx package? I really hate having to reboot to a different kernel when i want 3d accelearation.

[daenforcer]

Cheers for the guide
2.6.14-ck4 is running fine. Except that i no longer have fglrx working. does anyone know how to get it up and running? ive tried the method in the 'kernel compilation for newbies' thread but I end up tith a .deb that requires the fglrx-kernel-common package which is not in the repositories.

Is there some way to build the source for the xorg-driver-fglrx package? I really hate having to reboot to a different kernel when i want 3d accelearation.

:D I got the latest ATI driver (8.18.10) to work with 2.6.14+ck5 just fine. Follow the instructions located in the below 2 threads:

http://ubuntuforums.org/showthread.php?t=78466

http://ubuntuforums.org/showthread.php?t=76116&page=26

Let me know if you need any assistance.

[richardh]

Brilliant. Its fixed the slow boot problem on my Toshiba.

BUT I cant get Suspend (hibernate) from the System->Logout->hibernate menu

The computer switches off but then does not completely power down.

I would like to instal suspend2, but the patches given elsewhere in the forum are for the 6.12 ubuntu kernel

[Zed]

After doing all this, hddtemp didn't work anymore. Following these instructions:

http://www.thinkwiki.org/wiki/Problems_with_SATA_and_Linux

I found the right patch for 2.6.14 to fix it.

Thanks for the howto -- it was easier than it looked.

[gar]

hi ,
with , my msi S270 notebook , i have some sound with the new kernel , great , thanks
however , the USB did'nt work ... i havn't change anything in the menuconfig , should i ?
is there a patch ? i just begin in kernel compilation , so i don't understand everything ,sorry
if someone could help ...

[curtis]

Have a look at Point 7 of my guide:
http://www.ubuntuforums.org/showthread.php?t=85064
it explains how to compile the nvidia modules together with your kernel (as packages, just like Ubuntu's restricted modules)

Hi,
I managed to fix it by unistalling the current driver and re-installing it.
Wierd that just doing the installer did not work (Does show uninstalling)
I did already have the modules done.
Thank you for suggesting I read through that guide though.

[LPK]

With this kernel 2.6.14 dont work my fat32 disc in Breezy. With defoult kernel 2.6.12-9 is OK.
?????????

[Eazy©]

Has anyone found a solution to the "mount problem"?
I cant mount my ntfs partition though I think ntfs support its activated as a module (If I run menuconfig I see that its activated).
If I try to mount it it says that its already mounted or busy.

I'm glad I'm not the only one in this boat :)

I think I maybe have found a solution to the mount problem I have. If I try to mount my ntfs partition on hda with: "sudo mount -o loop /media/windows" I can access my ntfs partiton. I'm going to once again (5:th time :P ) compile the kernel and remove "Loopback device support" (located under "Device Drivers") and see if that does the trick. I'm also using the latest ck5-patch instead

I see myselfe still after 2 years with linux (off and on) as a newbee as only thing I do is to test things in guides, but I always seems to forget every thing afterwards and not lerning anything :P

Edit: Nope that didnt work to remove "Loopback device support" . Now I cant mount it even with: "sudo mount -o loop /media/windows" (offcourse...stupid me). Anyone knows how the line should look in fstab with the "-o loop" command in it?

"/dev/hda1 /media/windows -o loop ntfs nls=utf8,umask=0222 0 0" dont work.

I dont get the error mesage when "mouting local filesystem" when it boot with that line, but its not mounted when gnome starts. Perhaps the line in fstab should look different?

Going to compile 1 last time now and not leaving out "Loopback device support"

Edit2: Ok, I have now compiled for the last time (wont do it again) exactly as the guide, with the only diffrence that I selected write support for ntfs (dont think it makes anny difference). I edited my fstab to look like this:

/dev/hda1 /media/windows ntfs nls=utf8,umask=0222,loop 0 0

Changes in red

This makes my ntfs partition to load at startup. Hope someone else than me have some use of it...

[lmandrake]

@richardh: I'm just trying to follow this guide on the suspend2 Website (http://wiki.suspend2.net/DistroAndHardwareSetup/Ubuntu_Breezy_Badger?highlight=(ubuntu)) with 2.6.14-2. Sadly it's not working yet. But i'll post an update when it's hopefully working.

One question: How is ck working with 2.6.14-2? When I try to apply the ck patch it complains about nearly ervery file...

[curtis]

@richardh: I'm just trying to follow this guide on the suspend2 Website (http://wiki.suspend2.net/DistroAndHardwareSetup/Ubuntu_Breezy_Badger?highlight=(ubuntu)) with 2.6.14-2. Sadly it's not working yet. But i'll post an update when it's hopefully working.

One question: How is ck working with 2.6.14-2? When I try to apply the ck patch it complains about nearly ervery file...

What cK patchset are you using?
I am using the cK1 one perfect with the 2.6.14.2 GNU/Linux kernel.

[kaamos]

One question: How is ck working with 2.6.14-2? When I try to apply the ck patch it complains about nearly ervery file...

ck5 includes 2.6.14-2 patches, so you have to patch the 2.6.14-1 source. Linux-militia also offers pre-ck-patched kernel sources so you could try one of those.

http://www.linux-militia.net/kernel/ck/

[nastya]

/dev/hda1 /media/windows ntfs nls=utf8,umask=0222,loop 0 0

Changes in red

This makes my ntfs partition to load at startup. Hope someone else than me have some use of it...

Yay, now my partitons on my other drive will mount! Now I can play planet penguin racer at fullspeed AND access my other drive with my data. Thank you so much! :)

[RedDragon]

Kernel update: ok

Only problem is that it doesn 't detect my 'WMP54G' wifi adapter anymore. Is there an explanation. Did i forget an option during kernel compilation??

[lmandrake]

@curtis: Thanks. I now applied ck1 too 2.4.16-2 without errors.

Kernel compiles and installs fine now. But I can't boot it because root seems not to be mountable.

mount: mounting /dev/root on /root failed: Device or resource busy
...
mount: Mounting /root/dev on /dev/.static/dev failed: No such device or directory
mount: Mounting /dev on /root/dev failed: No such device or directory

I included my ide chipset driver, sata, all filetypes into the from the original ubuntu kernel inheritated kernel config. Don't know if it's related to the initrd scripts or if I'm missing something important in my config (see attached file).

[timczer]

I asked this question earlier, but was unable to get a response. I am new to compiling a kernel, and this went really well. My question is, once done, what can I remove fromthe source folder I built the kernel in? It is using a lot of memory that I would like to free up. Thanks.

[lmandrake]

You could remove the whole folder /usr/src/linux, don't forget to install kernel-headers as well, perhaps you need it later for some other kernel modules.
But creating an archive and burning it to cdr/dvdr is recommended.

[outr]

Tried this how-to with the current ck5 patch.
Compiled ok and everything seems to work.
A few quirks though:
iwconfig gives the following warning:
Driver for device eth1 has been compiled with version 19 of Wireless Extension, while this program supports up to version 18.
Some things may be broken...

it works ok still. waproamd seems to have some problems, will have to look into that.

Even the boot-up splash screen works, but now it's displaying xubuntu instead of ubuntu ? not a big problem but wtf?

[LeTon]

Hi,
After installation of new kernel ( which went flawlesly, and I finally got SOUND with Intel ICH 6) , Firofox would not launch:(
I was messing around with the browser befor new kernel...so that must be it...just don't know how to fix it. Please...
In terminal it returns:

INTERNAL ERROR on Browser End: Expected version > 5! Version = 4

System error?: : No such file or directory

Tried reinstallation already.
Thank you in advance

[domino]

Reading through the 12 pages I was able to gather almost all the information I need to compile this myself. I just need some more info though.

ATI drivers: Solved
Mount ntfs: Solved
Mount FAT32: ?
Mount HFS+: ?

Please comment on the last 2.

[tasca]

I compiled kernel 2.6.15 and now it usplash does not work and vmware has that to configure all time to load ("runme.pl"), and no reboot (halt ok).
I am use patch-2.6.15-rc1-ck1.bz2.

somebody can help?

[beeldings]

I followed this guide on two different machines, and it worked flawlessly on both occassions. On the one machine, I followed your guide, step-by-step, and it worked perfectly. On the second machine, I took my time and tweaked the kernel to suit this machine, removing all unnecessary modules and drivers, and wow, there was a very noticeable increase in the overall performance of this machine (PII 400 Mhz, 128 MB SDRAM).

[waydaws]

One question I have is that when I go to configure my kernel options, everything is already selected. For instance, where it says,

...*While you may tweak your kernel configuration to your needs I will sugest you some tweaks:

In "General Setup" activate:

-Support for paging of anonymous memory (swap)
--Support for prefetching swapped memory

In "Processor type and features":

-Processor family Choose the model of your processor.

In my config everything in general setup is already selected. Am I supposet to deselect everything but the above.

And...

For processor type, I'm not sure what to use. In my case it is running in VMWare. The processor type is listed as unknown in device manager. I'm not sure if I should just use 386 or the chose Pentium 4 (what the host is).

Anyone done this in vmware?

[redlaw28]

Hi
nice guide
i followed it when i compiled my nitro2 kernel it worked flawless

thx!

[Locutus]

I compiled the kernel but there are some errors. Is there a way for me to uninstall it so that I can give it a second go?

I get the following when I try and uninstall:

root@ubuntu:/usr/src# dpkg -r kernel-image-2.6.14-ck5_ck1_i386.deb
dpkg: you must specify packages by their own names, not by quoting the names of the files they come in

Type dpkg --help for help about installing and deinstalling packages ;
Use `dselect' or `aptitude' for user-friendly package management;
Type dpkg -Dhelp for a list of dpkg debug flag values;
Type dpkg --force-help for a list of forcing options;
Type dpkg-deb --help for help about manipulating *.deb files;
Type dpkg --licence for copyright licence and lack of warranty (GNU GPL) .

Options marked produce a lot of output - pipe it through `less' or `more' !
root@ubuntu:/usr/src#

Any ideas?

Thanks for any help you can give me.

[Rob2687]

You can search for it in synaptic and remove it from there.

[Locutus]

That worked like a charm. Thank you.

[mistergq]

I built the kernel, and then get this error when I try to install it.


dpkg: status database area is locked by another process

I am the root user in the terminal window. Any other suggestions?
------------------------------
never mind, I had the Synaptic opened.

[Raoul Duke]

Is there a list of patches used to produce the default Ubuntu kernels? I would like to apply them selectively...
Thanks

[henriquemaia]

Nice howto.

Worked flawlessly. Nothing like our own kernels!

[devanb]

I just want to thank you for the great guide. I've tried to install my wireless drivers using Fedora and Gentoo, but unsuccessfully when trying to recompile my kernel.

Your guide was excellent and helped me get Ubuntu working.

Thanks again,

Devan

[obscenic]

When I tried to do my apt-getting, I got this:

mark@susty:~$ sudo apt-get install libqt3-headers libqt3-mt-dev
Reading package lists... Done
Building dependency tree... Done
Some packages could not be installed. This may mean that you have
requested an impossible situation or if you are using the unstable
distribution that some required packages have not yet been created
or been moved out of Incoming.
The following information may help to resolve the situation:

The following packages have unmet dependencies:
libqt3-mt-dev: Depends: xlibs-static-dev (>= 4.3.0.dfsg.1-4) but it is not going to be installed
Depends: libxext-dev (>= 4.3.0.dfsg.1-4) but it is not going to be installed
Depends: libxrandr-dev (>= 4.3.0.dfsg.1-4) but it is not going to be installed
Depends: libsm-dev (>= 4.3.0.dfsg.1-4) but it is not going to be installed
Depends: libxmu-dev (>= 4.3.0.dfsg.1-4) but it is not going to be installed
Depends: libice-dev (>= 4.3.0.dfsg.1-4) but it is not going to be installed
Depends: libx11-dev (>= 4.3.0.dfsg.1-4) but it is not going to be installed
Depends: libxt-dev (>= 4.3.0.dfsg.1-4) but it is not going to be installed
Depends: libmng-dev (>= 1.0.3) but it is not going to be installed
Depends: libpng12-0-dev
Depends: libjpeg62-dev but it is not going to be installed
Depends: libfreetype6-dev but it is not going to be installed
Depends: xlibmesa-gl-dev but it is not going to be installed or
libgl-dev
Depends: xlibmesa-glu-dev but it is not going to be installed or
libglu1-mesa-dev but it is not going to be installed or
libglu-dev
Depends: libxft-dev but it is not going to be installed
Depends: libxrender-dev but it is not going to be installed
Depends: libxcursor-dev but it is not going to be installed
Depends: libxinerama-dev but it is not going to be installed
E: Broken packages

[mistergq]

Is there way to get have this kernel built with Alsa 1.0.10? If so how?

Thanks

[shaggydoo]

Hi, I was wondering, how long should this take? I've been waiting a couple of hours now, but it's still building the kernel. (Duron 800mhz, 256mb RAM)

[Rob2687]

Probably atleast an hour...depending on how much stuff you have left in to compile and stuff. It takes over an hour on my 1Ghz PIII and I probably have a lot of useless crap that could probably be disabled since my PC doesn't need it.

[shaggydoo]

Oh, I see. Great, thanks Rob!

[sswords]

Hi,
I am using 2.6.14.1-cK1 fine here.
That is on a Dell Inspiron 5150.
Just having an issue with the nvidia module.
I end up needing to re-install it every boot up.
It is in /etc/modules though, and modprobing nvidia does not work.
I have a feeling it could be InitNG possibily, will check now.

Any one else running the 2.6.14.x kernel with a Nvidia graphics card?
Even better if they are using InitNG.

I'm not using InitNG, but I am having a similar problem where I can't start gdm or x on reboot without first running the nvidia installer. I've tried a number of things I thought sure would work - uninstalled all other nvidia packages including linux-restricted-modules for the other kernels I have, turned off the nvidia-glx service. I even made a debian package using
sudo checkinstall --nodoc --strip=no NVIDIA-Linux-x86-1.0-7174-pkg1.run
-- then next time I rebooted, x wouldn't start. I extracted the package to its own directory and checked the differences between what was in that folder and what was installed ("sudo diff -qr . / | grep -v 'Only in /'") and everything was the same. Once I ran the installer again, everything worked, but there were several changes evident when I ran that diff again, including the nvidia.ko module and several numbered files in /var/lib/nvidia.

The specific error that occurs is (in /var/log/Xorg.0.log)
(EE) NVIDIA(0): Failed to load the NVIDIA kernel module!
(EE) NVIDIA(0): *** Aborting ***
(II) UnloadModule: "nvidia"
(EE) Screen(s) found, but none have a usable configuration.
However, the nvidia module is in fact loaded according to lsmod, (it's listed in /etc/modules), and doing "rmmod nvidia; modprobe nvidia" doesn't give any sort of error. (I've tried "insmod [actual path to nvidia.ko]" just in case it was somehow loading a different nvidia module, too.)

Any ideas? At this point I'm tempted to make an init script that runs the nvidia installer every time I reboot.

I have a Geforce2 GTS so I use an older version of the installer (and the nvidia-legacy packages when using an official kernel), but it doesn't seem like that's any problem because it does work; just requires a reinstall every boot.

[curtis]

I got it working by adding modules_install to the build line.
Then installing the deb created by that...

[taciturn]

I just gave this a shot and everything went great, with the exception of one problem. My samba shares are no longer automounted at boot, I can still mount them after I log into gnome, but they're no longer mounted automatically.

I cant figure out why this is, anyone have any suggestions?

[sswords]

I figured out why I need to reinstall the nvidia drivers every time I reboot: the nodes /dev/nvidia[0-9] and /dev/nvidiactl don't appear when the module is loaded, but get created by the installer. I suppose there's probably some problem with my udev configuration (though I haven't changed anything.) Does anyone know how to make them appear?

[xUmaRix]

hi guys.
I already compile kernel 2.6.14.3 with ck-6 patch on hoary
but got error on bootup can't mount to devfs maybe i didn't config it on kernel and device-mapper:dm-linear:Device lookup fail.My orinoco gold driver also not load automatically.The worst part is i can't load my fluxbox. I read the xorg.log and it shows like font error coz i use artwiz font.I also can't boot to blackbox.
Izzit this patch not suitable for hoary or is there any extra setting or step i should add.

[coaxx]

2.6.14 worked well, thank You very much, this is a usefull HowTo!

some problems here with 2.6.15 / rc2 /rc5, I tried w/ and w/o cK Patch!

("Fehler" is "error" in english :)

CC kernel/irq/handle.o
In file included from include/asm/fixmap.h:27,
from include/asm/smp.h:16,
from include/linux/irq.h:13,
from kernel/irq/handle.c:9:
include/asm/acpi.h: In Funktion »__acpi_acquire_global_lock«:
include/asm/acpi.h:67: Fehler: »boot_cpu_data« nicht deklariert (erste Benutzung in dieser Funktion)
include/asm/acpi.h:67: Fehler: (Jeder nicht deklarierte Bezeichner wird nur einmal aufgeführt
include/asm/acpi.h:67: Fehler: für jede Funktion in der er auftritt.)
include/asm/acpi.h: In Funktion »__acpi_release_global_lock«:
include/asm/acpi.h:79: Fehler: »boot_cpu_data« nicht deklariert (erste Benutzung in dieser Funktion)
make[3]: *** [kernel/irq/handle.o] Fehler 1
make[2]: *** [kernel/irq] Fehler 2
make[1]: *** [kernel] Fehler 2
make[1]: Verlasse Verzeichnis »/usr/src/linux-2.6.15-rc5«
make: *** [stamp-build] Fehler 2


Anybody?

[limit223]

2.6.14 worked well, thank You very much, this is a usefull HowTo!

some problems here with 2.6.15 / rc2 /rc5, I tried w/ and w/o cK Patch!

("Fehler" is "error" in english :)

CC kernel/irq/handle.o
In file included from include/asm/fixmap.h:27,
from include/asm/smp.h:16,
from include/linux/irq.h:13,
from kernel/irq/handle.c:9:
include/asm/acpi.h: In Funktion »__acpi_acquire_global_lock«:
include/asm/acpi.h:67: Fehler: »boot_cpu_data« nicht deklariert (erste Benutzung in dieser Funktion)
include/asm/acpi.h:67: Fehler: (Jeder nicht deklarierte Bezeichner wird nur einmal aufgeführt
include/asm/acpi.h:67: Fehler: für jede Funktion in der er auftritt.)
include/asm/acpi.h: In Funktion »__acpi_release_global_lock«:
include/asm/acpi.h:79: Fehler: »boot_cpu_data« nicht deklariert (erste Benutzung in dieser Funktion)
make[3]: *** [kernel/irq/handle.o] Fehler 1
make[2]: *** [kernel/irq] Fehler 2
make[1]: *** [kernel] Fehler 2
make[1]: Verlasse Verzeichnis »/usr/src/linux-2.6.15-rc5«
make: *** [stamp-build] Fehler 2


Anybody?


I've got my 2.6.15-rc1-ck1 patched from here, a week or two ago, I copiled without any error you have above...you may find the latest kernel as well ...already patched it.
http://www.linux-militia.net/modules/news/

[coaxx]

thanks for the hint. Meanwhile I have dist-upgraded to dapper :cool: 2.6.15-x compiles now easily.

[ZedLord]

I Followed the Guide upto 5-Lets Build the kernel:

I'm Trying to compile kernel 2.6.14.3 along with patch-2.6.14-ck6

when i type-in : sudo make-kpkg -initrd --revision=ck6 kernel_image

i get the Following error after sometime and return to the shell

CC kernel/time.o
CC kernel/softirq.o
CC kernel/resource.o
kernel/resource.c:480: warning: ‘__check_region’ is deprecated (declared at kernel/resource.c:468)
CC kernel/sysctl.o
kernel/sysctl.c:671: error: redefinition of ‘one_hundred’
kernel/sysctl.c:665: error: previous definition of ‘one_hundred’ was here
kernel/sysctl.c:748: error: ‘vm_swappiness’ undeclared here (not in a function)
make[2]: *** [kernel/sysctl.o] Error 1
make[1]: *** [kernel] Error 2
make[1]: Leaving directory `/usr/src/linux-2.6.14.3cK6'
make: *** [stamp-build] Error 2


What am I Doing Wrong, please help Me!

[curtis]

I Followed the Guide upto 5-Lets Build the kernel:

I'm Trying to compile kernel 2.6.14.3 along with patch-2.6.14-ck6

when i type-in : sudo make-kpkg -initrd --revision=ck6 kernel_image

i get the Following error after sometime and return to the shell



What am I Doing Wrong, please help Me!
Hi,
Patch the 2.6.14 kernel with the 2.6.14-ck6 patchset.
As it includes the patches in 2.6.14.1, 2.6.14.2, 2.6.14.3

[windowsrefund]

-High Memory Support
--off -if you have less than 1 GB of RAM
--1GB Low Memory Support -if you have 1GB of RAM
--4GB -if you have more than 1GB of RAM


Rúben Gonçalves:p


I don't see this option in 2.6.14. Anyone else?

[interzoneuk]

Hi people.

What is the correct procedure for installing the latest kernel and latest ck patchset?

I tried using the guide in the original post but used kernel 2.6.14.3 and patch-2.6.14-ck6.bz2.

I used the command

sudo bzcat /home/username/patch-2.6.14-ck6.bz2| patch -p1

and answered 'y' to all.

When i tried the recompile the kernel it errored (i did not record the error).

However I then wiped the /usr/src/linux directory and started again.

This time I used the 2.6.14.3 kernel and the patch-2.6.14-ck1.bz2, it patched ok.

I then patched the source with these updates

2.6.14-ck1_to_2.6.14-ck2.patch.bz2
2.6.14-ck2_to_2.6.14-ck3.patch.bz2
2.6.14-ck3_to_2.6.14-ck4.patch.bz2
2.6.14-ck4_to_2.6.14-ck5.patch.bz2
2.6.14-ck5_to_2.6.14-ck6.patch.bz2

I used this command (is this correct)

sudo bzcat /home/username/2.6.14-ck1_to_2.6.14-ck2.patch.bz2| patch -p1

This time I could recompile (nvidia and nforce driver are also running fine)

Is there a way to patch the ck6 without having to upgrade from 1 - 6?

Cheers

[Bloot]

I don't see this option in 2.6.14. Anyone else?

Yes, me too. Maybe on x86-64 they are not avaliable.

[hoodwink]

Hi people.

What is the correct procedure for installing the latest kernel and latest ck patchset?

. . .

Is there a way to patch the ck6 without having to upgrade from 1 - 6?

I, too, would like to know the answer.

As an aside, this HOWTO leaves out a lot of preparatory work for building a kernel on Ubuntu. I could not get make xconfig to work; not surprising, since that has been problematic on other distros. So I switched to the old standby make menuconfig only to discover that ncurses-devel is required. Considerable searching to find that this is really the libncurses5-dev package.

make-kpkg clean and make-kpkg -initrd --revision=ck1 kernel_image were uneventful, but dpkg -i kernel-image-2.6.14*.deb died because a boatload of dependencies around mkinitrd-tools were not present. After tracking this down, the kernel .deb installed ok.

Reboot brought a few more surprises. The standard bootscripts (I haven't changed any) try to setup raid devices, and this failed. Also errors stating that udev was already started, and something called dirmgr failed.

Net result: I'm running from the kernel, but either the .config loaded from /boot doesn't describe everything that's required, or some of the support packages needed for the bootscripts have hidden kernel dependancies, or ???

So, I doubt that this kernel is a keeper, but it's been a learning experience.

[kaamos]

Hi people.

What is the correct procedure for installing the latest kernel and latest ck patchset?

I tried using the guide in the original post but used kernel 2.6.14.3 and patch-2.6.14-ck6.bz2.
[...]
Is there a way to patch the ck6 without having to upgrade from 1 - 6?


You just apply the ck6 patch to the 2.6.14 sources. Note that not to the sources of 2.6.14.3 as the ck patch includes the patches from 2.6.14.1 to 2.6.14.3.

[coaxx]

maybe you shouldt also tell 'em to install


sudo apt-get install kernel-package g++


Otherwise the "make" command is not known.

[coaxx]

-deleted-

[coaxx]

I've got my 2.6.15-rc1-ck1 patched from here, a week or two ago, I copiled without any error you have above...you may find the latest kernel as well ...already patched it.
http://www.linux-militia.net/modules/news/

Now I know what the problem was. It only happens if 386 is chosen for CPU model. With Pentium-M everything works well here compiling the 2.6.15rc5!

[Mizzou_Engineer]

I think that you should pick the PIII normal processor type for a PIII-M as even though the P-M is based on the PIII's architecture, the P-M has things like SSE2 that your PIII-M does not.

I have a P4-M and use the Pentium 4 type as my P4-M is basically a Pentium 4 with the multiplier able to be dropped from 22x to 12x (and lower-power transistors.) Your PIII-M is the same to a PIII desktop chip as mine is to a P4 Northwood desktop chip.

[jeffreyvergara.NET]

anyone got their automount working after update to 2.6.14 kernel? I tried using pmount but it wont work either. I have this problem before and fixed by editing HAL users & groups but it seems I can't find it anymore... or just forgot what I typed.. Y_Y

[dk_pa]

Hmm...what should I do now for Linux Headers? I would assume i would need a new package. Just looking at the other ones for 2.6.12-10 they seem to just be certain folders from the kernel package? Are these available from a Debian repository?

EDIT:: Sorry, Checked over at the Documentation project. Answer seems to be here http://doc.gwos.org/index.php/Kernel_Compilation. Just have to wait 45 minz, looks like it's going to be okay. Maybe the headers should be added to the how to? I read through it again and didn't see it. Thanks for the how to btw, got my ITE8212 ATA RAID adapter working and made my day =) Another edit, I'm a little slow. I guess you dont need "headers" when you have the source. I just didn't know how that all worked. Or maybe I will need the headers? Anyone have answer on that...haha. I love learning =)

[Bloot]

anyone got their automount working after update to 2.6.14 kernel? I tried using pmount but it wont work either. I have this problem before and fixed by editing HAL users & groups but it seems I can't find it anymore... or just forgot what I typed.. Y_Y

Yes automount won't work anymore, and pmount doesn't work either. If I can't fix it I'll probably turn back soon to 2.6.12-10.

[jeffreyvergara.NET]

how do you uninstall the kernel(2.6.14) together with it's files?

[just-hatched]

Am new to the world of kernel compiles, have successfully compiled and installed 2.6.14.3 from the sticky instructions in this thread, but cannot get 2.6.15-rc5 to actually compile, and I havent a clue how to troubleshoot this problem or even where to start. Used exact same commands for this compile as the 2.6.14.3

The reason am looking to build2.6.15-rc5 is to be able to use a Leadtek DVT 1000 DVB-T card, which according to another forum(whirlpool), support for this card has been built into 2.6.15-rc5. If I can get to work, then I dont need to go and buy something with drivers...

The compile error, not the whole output, just the end before it blows up:
CC kernel/sched.o
CC kernel/fork.o
CC kernel/exec_domain.o
CC kernel/panic.o
CC kernel/printk.o
CC kernel/profile.o
CC kernel/exit.o
CC kernel/itimer.o
CC kernel/time.o
CC kernel/softirq.o
CC kernel/resource.o
kernel/resource.c:480: warning: ‘__check_region’ is deprecated (declared at kernel/resource.c:468)
CC kernel/sysctl.o
CC kernel/capability.o
CC kernel/ptrace.o
CC kernel/timer.o
CC kernel/user.o
CC kernel/signal.o
CC kernel/sys.o
CC kernel/kmod.o
CC kernel/workqueue.o
CC kernel/pid.o
CC kernel/rcupdate.o
CC kernel/intermodule.o
kernel/intermodule.c:178: warning: ‘inter_module_register’ is deprecated (declared at kernel/intermodule.c:38)
kernel/intermodule.c:179: warning: ‘inter_module_unregister’ is deprecated (declared at kernel/intermodule.c:78)
kernel/intermodule.c:181: warning: ‘inter_module_put’ is deprecated (declared at kernel/intermodule.c:159)
CC kernel/extable.o
CC kernel/params.o
CC kernel/posix-timers.o
CC kernel/kthread.o
CC kernel/wait.o
CC kernel/kfifo.o
CC kernel/sys_ni.o
CC kernel/posix-cpu-timers.o
CC kernel/futex.o
CC kernel/dma.o
CC kernel/uid16.o
CC kernel/module.o
CC kernel/kallsyms.o
CC kernel/irq/handle.o
In file included from include/asm/fixmap.h:27,
from include/asm/smp.h:16,
from include/linux/irq.h:13,
from kernel/irq/handle.c:9:
include/asm/acpi.h: In function ‘__acpi_acquire_global_lock’:
include/asm/acpi.h:67: error: ‘boot_cpu_data’ undeclared (first use in this function)
include/asm/acpi.h:67: error: (Each undeclared identifier is reported only once
include/asm/acpi.h:67: error: for each function it appears in.)
include/asm/acpi.h: In function ‘__acpi_release_global_lock’:
include/asm/acpi.h:79: error: ‘boot_cpu_data’ undeclared (first use in this function)
make[3]: *** [kernel/irq/handle.o] Error 1
make[2]: *** [kernel/irq] Error 2
make[1]: *** [kernel] Error 2
make[1]: Leaving directory `/home/flip/linuxbuild'
make: *** [stamp-build] Error 2

Where do i start looking, the .config file didnt make much sense, had a few goes trying to turn unnecessary items off without any change in the result...

helpless kernel newbie, any suggestions?

[Bloot]

how do you uninstall the kernel(2.6.14) together with it's files?

That's what I did:

sudo aptitude remove linux-image-2.6.14-ck6

where ck6 is the patchset I applied.

Then removed linux from /usr/src and all the 2.6.14 files and folders.

[just-hatched]

thanks for the hint. Meanwhile I have dist-upgraded to dapper :cool: 2.6.15-x compiles now easily.

So did you find out the cause of the problem?

[luminoso]

Everything ok.. but

i am getting this:

kernel-image-2.6.14-ck1_ck1_i386.deb

what about i686+sse2+O3 optimizations?!

[coaxx]

So did you find out the cause of the problem?


hi yea just look ahead, I did reply to my own problem (use threaded view in this forum to have a better discussion overview): Sulution for me: Chose a different CPU in kernel config (not 386)

[interzoneuk]

kaamos. Thank you very much.

I used 2.6.14 kernel (not 2.6.14.3) and the ck6 patch works fine, it installed o.k and runs really nice.

...however it is not running quite as fast as the Supersuse kernel...

Cheers

[tud]

I generated my own kernel using your guide and adding support for ndiswrapper and last ati drivers. Everything is ok but now I need to install VmWare workstation and when I run vmware-config.pl I have an error
The path "/usr/src/linux/include" is a kernel header file directory, but it does
not contain the file "linux/version.h" as expected. This can happen if the
kernel has never been built, or if you have invoked the "make mrproper" command
in your kernel directory. In any case, you may want to rebuild your kernel.

How can I generate linux-headers-2.6.14-ck1 package?
Thanks for your work :KS

EDIT: solved after recompiling the kernel with:
sudo make-kpkg --initrd --append-to-version=-custom kernel_image kernel_headers modules_image

[hubbadub]

Thanks, this worked perfectly the 1st time on both my friends Ubuntu box, and my Kubuntu one.

[interzoneuk]

Hi everybody.

For the people who have been following the thread there is now ck7 patch out.

http://ck.kolivas.org/patches/2.6/2.6.14/2.6.14-ck7/


cheers.

[getaceres]

I've compiled the 2.6.14 kernel with the ck7 patch but I've got the pmount bug (Device /dev/something is already mounted on /media/somepoint). The problem is that this file: http://darkstar.ist.utl.pt/ubuntu/archive/pool/main/p/pmount/pmount_0.9.6-1_i386.deb that solves this isn't there anymore, and this file http://darkstar.ist.utl.pt/ubuntu/archive/pool/main/p/pmount/pmount_0.9.6-1~breezy1_i386.deb doesn't solve the problem (the file http://darkstar.ist.utl.pt/ubuntu/archive/pool/main/p/pmount/pmount_0.9.7-1ubuntu1_i386.deb gives me unresolved dependencies)
Anyone knows how to solve this?

[Bloot]

I've compiled the 2.6.14 kernel with the ck7 patch but I've got the pmount bug (Device /dev/something is already mounted on /media/somepoint). The problem is that this file: http://darkstar.ist.utl.pt/ubuntu/archive/pool/main/p/pmount/pmount_0.9.6-1_i386.deb that solves this isn't there anymore, and this file http://darkstar.ist.utl.pt/ubuntu/archive/pool/main/p/pmount/pmount_0.9.6-1~breezy1_i386.deb doesn't solve the problem (the file http://darkstar.ist.utl.pt/ubuntu/archive/pool/main/p/pmount/pmount_0.9.7-1ubuntu1_i386.deb gives me unresolved dependencies)
Anyone knows how to solve this?

Yeah, I'd really like to solve the pmount problem on my 64bit ubuntu too.

[getaceres]

For the moment I'm back to 2.6.12-10 kernel.
It's a shame, the 2.6.14-ck7 kernel resolved a couple of bugs I have right now with kernels greater than 2.6.10.

[yoshic]

:p Thanks for the How-to, it worked perfect for me :razz::razz::razz::razz:

[Tomasz]

Yeah, pmount 0.9.6 does not fix the problem. 0.9.7 gives dependency issues. This is open source software, there *surely* must be a fix to this problem somewhere. Come on, it's the only problem with the 2.6.14 series. Anyone knows pmount enough to hack up a working version for breezy?

[medivh]

I followed the instructions and compiled the kernel correctly it works fine but i want to see splash screen in higher resolution and also the console.How can i make it appear like it.It turns into a blank screen when i write vga=0x318 at the and of the menu.lst Is there any other way to do it?

and it says its i386 in its name i am on a laptop and i think i need i686 am i wrong how can i optimize it for pentium M (centrino 1.5)

[limit223]

A day ago I compiled a new kernel 2.6.15-rc6-ck1 with smp/smt modules enabled for my HT Cpu. I made kernel-image, kernel-headers, fglrx-kernel .deb's and everything went fine without errors.
Kernel boot was fine, speed really impressing, modules loaded fine...

Though, I've got a problem which I really do not have any idea about: Conky shows me in the cpu1, cpu2 loading values of over 100% ( sometimes over 500% :eek: ). I tried using that sensor=cpu in sperkaramba applet and it shows me as well big values in cpu loading... Everything seems ok...the applications open without problems , the fan of cpu which for values of loading more than 20% usually starts, doesn't sound much as for over 100%....
Is anyone can enlighten me what's all about?



P.S. Merry Christmas, Everyone!!

[limit223]

No one has any clue for me why this big loading values in cpu1, cpu2 ?
The only real values that I belive are right are shown in #top command but doesn't help much ...

I've been searching all over the net for this kind abnormal cpu monitoring...no luck...:(

[tseliot]

No one has any clue for me why this big loading values in cpu1, cpu2 ?
The only real values that I belive are right are shown in #top command but doesn't help much ...

I've been searching all over the net for this kind abnormal cpu monitoring...no luck...:(
Perhaps you should wait for the stable version to be released.

EDIT: kernel 2.6.15 stable is already out!!!

[limit223]

Perhaps you should wait for the stable version to be released.

EDIT: kernel 2.6.15 stable is already out!!!

Cool...:)
Is it out already???? ....I haven't seen it on kernel.org, yet...Is there any other places where new vanilla versions are found?...

I've got also 2.6.15-rc1-ck1 which I'm very happy with it...that's why I tried another unstable ver. 2.6.15-rc6-ck1..

[tseliot]

Cool...:)
Is it out already???? ....I haven't seen it on kernel.org, yet...Is there any other places where new vanilla versions are found?...

I've got also 2.6.15-rc1-ck1 which I'm very happy with it...that's why I tried another unstable ver. 2.6.15-rc6-ck1..
No, I was wrong and out of my mind :p .
Unfortunately only 2.6.14.5 is out :(

[atlas95]

thx for this guide,
i have a little problem...after compiling this kernel, my ntfs partitions on s-ata work correctly, but my 5 ntfs partitions on other disk which are in ata133 don't work :s, the mount failed...

can you say me what i have forgot in config?
thanks, sorry for my english

[MighMoS]

This How-To will guide you through the compilation/installation of the 2.6.14 Vanilla Kernel (more recent than the one distribuited with Ubuntu 5.10) with the performance patchset by Kon Colivas.
I think you mispelled his name. Isn't it "Con Kolivas"?

[SirKillalot]

I cannot mount partitions from my second HDD anymore!
It says:
/dev/hdb7 already mounted or /mount/point already in use

[muchmusic]

yes it is Con Kolivas

CK patchset thus named =)

http://members.optusnet.com.au/ckolivas/kernel/

[limit223]

New compilation of 2.6.15-rc7-ck1 vanished my problems... Found some bugs in kernel mail list archive for 2.6.15-rc4, rc5 and rc6 ..just for those who want to try prereleased versions.

[theh0g]

I cannot mount partitions from my second HDD anymore!
It says:
/dev/hdb7 already mounted or /mount/point already in use

You have to remove EVMS, I had the same problem and this fixed it.

[AntiMattR]

Awesome!

I have a Mini-ITX system that I have been trying to turn into a MythTV box. Is there anything special I need to do to optimize this kernel for the EPIA?

This board also has a Unichrome graphics accelerator and needs patches to the kernel (as referenced in https://wiki.ubuntu.com/ViaEpiaDriHowto) - could I apply these patches and compile without any problems?

Thanks!

[juantxorena]

I've found a lot of people, like me, with mounting local system problems after compiling and installing cK patched kernel. I've found this post with a solution for the problem: http://www.ubuntuforums.org/showthread.php?t=103900&page=1&pp=10
which few people have visited, but it worked for me.

[juantxorena]

I've just installed the 2.6.14 kernel with ck7 patch. The first attemp I got the pmount error. Then I configured the kernel again, but not compiling neither as a module) some options, like tape, PCMCIA, all laptop-related things, etc. Also I patched the kernel as I said above. And then installed the newest NVIDIA drivers, 8178 (the first time I got the refresh rate at maximun possible)

Now I've got pmount working, and everything (for now) working flawlessly.

I don't know what this patch do in the kernel, but now my computer flies.

[mach]

Very nice howto! First time poster here, but allow me to make a couple points..

I had to remove the linux-restricted-modules a second time, to no surprise, to get ati's fglrx drivers working again.

CK1 is an old version of Kon Colivas's patches. CK8 is now out (includes patches from linux 2.6.14.5). I realize the how-to was made a couple months ago, just letting other peeps know about CK8. ;) (link: http://ck.kolivas.org/patches/2.6/2.6.14/2.6.14-ck8/)

once again, thanks for the how-to.

[newuser111]

after compiling 2.6.14 with ck8 patch the blackscreen bug when logging out is fixed! SMP support is enabled also

[drfalkor]

EDIT: nevermind

[Gandalf]

kernel 2.6.15 released, download here (http://www.kernel.org/pub/linux/kernel/v2.6/linux-2.6.15.tar.bz2)
also colin kalivas released his fist patch today, download here (http://ck.kolivas.org/patches/2.6/2.6.15/2.6.15-ck1/patch-2.6.15-ck1.bz2)

I'm currently installing it, i'll report if any problems occured...

[Gandalf]

sadly usplash did not work, plus there are multiple modules that won't boot, i'm trying ck8 with 2.6.14 now

[mach]

Sadly, 2.6.15 doesn't fix pmount/mount issues.

[imranj]

Yup the Usplash doesnt work, dont know why can some one please check into ti and tell me why isnt it working please?

sound intially was dead, but i tweaking a button and bang on my face
ahahahah

[imranj]

Which IO scheduler should i use?

CFQ or Anticpatory?

i have desktop but i am thinking about Anticipatory?

should or shouldnt and why?

urgently pls

[limit223]

I do not have any problem with the new final release; compiled with smp, patched with new driver from SysKonnect for one of my network card, reinstalled Radeon ati 8.20.8 driver...everything went smooth without any error.
This is my hardware configuration:

P4HT 915ExpressChipset@3Ghz|ASUS P5GD1 LGA775 IT8212 chipset|Intel-PRO/100 + Yukon Gigabit Ethernet 10/100/1000Base-T Adapter|512MB DDR@400Mhz|160SATA+120PATA both WD|ATI Radeon X600Pro(RV380 3E50) PCI 128MB|Leadtek Winfast 2000XP TV-tuner|

Usplash absence wouldn't bother me too much...

[Gandalf]

i tried dapper upslash which suppose to work with 2.6.15 but nothing!!! once i saw there is an error , insmod /lib/modules/...../vga.. don't quite remember it and i didn't find where it suppose to be in the logs, i searched all /var dir but couldn't find it, this error will be written right after grub uncompress the kernel... i think that's why upslash ain't working

[Ainvar]

I've just installed the 2.6.14 kernel with ck7 patch. The first attemp I got the pmount error. Then I configured the kernel again, but not compiling neither as a module) some options, like tape, PCMCIA, all laptop-related things, etc. Also I patched the kernel as I said above. And then installed the newest NVIDIA drivers, 8178 (the first time I got the refresh rate at maximun possible)

Now I've got pmount working, and everything (for now) working flawlessly.

I don't know what this patch do in the kernel, but now my computer flies.

Can you list what you did not compile this time around? I am getting the pmount issue. I have also tried the evms patch.

[TecnoVM64]

Hello people, I've been trying to compile the recent linux kernel 2.6.15 but i have the following problem:

root@h4x:/usr/src/linux# make xconfig
HOSTCC scripts/basic/fixdep
In file included from /usr/include/sys/socket.h:35,
from /usr/include/netinet/in.h:24,
from /usr/include/arpa/inet.h:23,
from scripts/basic/fixdep.c:115:
/usr/include/bits/socket.h:304:24: error: asm/socket.h: No such file or directory
make[1]: *** [scripts/basic/fixdep] Error 1
make: *** [scripts_basic] Error 2

Any ideas?

(I can do it with the 2.6.14ck3 source that still have somewhere) :(

[limit223]

From my short experience in Debian system, loading X as root will never work!
I had the same error trying to do that as well.
Exit root and do:
sudo make xconfig

[juantxorena]

Can you list what you did not compile this time around? I am getting the pmount issue. I have also tried the evms patch. Now I'm using ck8 patch, not ck7. I attach my config file. Maybe you can use it without modifying it (k7 processor).

[goombah88]

I followed the instructions and succeeded in installing the new kernel with the patch. I even was able to build a nvidia module deb to install. However a couple of things, showstoppers in fact, did not work. So far I discovered that the side scolling on my touchpad is not working, and more importantly suspend stopped working as well. I'm just going to stick with 2.6.12-9-686 for now.
i have a similar problem. last night i installed the 2.6.14 kernel with the ck8 patch and now whenever i try to use the scroll buttons next to the synaptic touchpad on my gateway laptop nothing happens. on the desktop i get what looks like a switch desktop dialog popup and in firefox it won't scroll at all. very annoying to not be able to scroll.

if anyone knows a way around this or how to make it work again please let me know. otherwise i may have to go back to the 2.6.12-10-686-smp kernel i was using b4 this.

thanks in advance.

[imranj]

Well instead of root , be in normal command prompt and type this

xhost +

the su <<passwprd>>

u are redy to go

;)



From my short experience in Debian system, loading X as root will never work!
I had the same error trying to do that as well.
Exit root and do:
sudo make xconfig

[imranj]

when you complie the kernel, select automout v3 and 4,

and also check out the some VMS thing, ok, some file needs to be edited yeah.

good luck.

Sadly, 2.6.15 doesn't fix pmount/mount issues.

[imranj]

Well at least i should know,whts happening, what is wroking or has failed hmm, and with out the boot screen, uff.

not good.


I do not have any problem with the new final release; compiled with smp, patched with new driver from SysKonnect for one of my network card, reinstalled Radeon ati 8.20.8 driver...everything went smooth without any error.
This is my hardware configuration:

P4HT 915ExpressChipset@3Ghz|ASUS P5GD1 LGA775 IT8212 chipset|Intel-PRO/100 + Yukon Gigabit Ethernet 10/100/1000Base-T Adapter|512MB DDR@400Mhz|160SATA+120PATA both WD|ATI Radeon X600Pro(RV380 3E50) PCI 128MB|Leadtek Winfast 2000XP TV-tuner|

Usplash absence wouldn't bother me too much...

[Ainvar]

Now I'm using ck8 patch, not ck7. I attach my config file. Maybe you can use it without modifying it (k7 processor).


Thanks, I will look at the config and compare it to mine, unfortunately I am on a laptop with a Pent M class cpu. So a few changes will need to be made.

I remember on aone of the first few pages of this thread a link that will compare the configs to one another, I think I will give that a shot also.

[limit223]

Well at least i should know,whts happening, what is wroking or has failed hmm, and with out the boot screen, uff.

not good.


You are able to see every module at boot time without usplash (this is just a eyecandy of boot time). Go to /boot/grub/menu.lst and delete "splash" word from the end of the line:
kernel /boot/vmlinuz-2.6.15-ck1-smp root=/dev/hdb10 ro quiet splash

PS This setting could be disable in writing menu.lst at every kernel compilation by deleting the same word in the line:
# nonaltoptions=quiet splash

[limit223]

Well instead of root , be in normal command prompt and type this

xhost +

the su <<passwprd>>

u are redy to go

;)

Typing "xhost +" I've got this message:

access control disabled, clients can connect from any host

OK I've got it...it works...Thanks!

[ZedLord]

I Want the USplash Back!!!!, Is There any How-to of sorts which helps in bringing it back!,

EDIT: Even the Font (the startup and shutdown text) Seems to be messed!, I Looks kinda big.

[limit223]

There is another option to have a splash ...
A couple of days ago Upower was lake of one package depency...Now a new version is released and you may find in this repo:
deb http://repo.nanofreesoft.org breezy main

More info here (http://nanofreesoft.org/index.php?module=subjects&func=viewpage&pageid=1).

I just installed it and it works in my new kernel...now I try to figure out how to customize its theme.

[Gandalf]

hmmmm This didn't even work with breezy kernel for me :o :o

[limit223]

Gandalf,
I may say that original kernel from breezy lived just a week in my system( I have my own config file that I use it in compilation)...I really do not have any ideea why things do not work to you...but I found something probably would help you figure out something:
http://www.ubuntuforums.org/showpost.php?p=287946&postcount=208

[Rob2687]

upower doesn't uninstall :(

Removing upower ...
Searching for GRUB installation directory ... found: /boot/grub .
Testing for an existing GRUB menu.list file... found: /boot/grub/menu.lst .
Searching for splash image... none found, skipping...
Found kernel: /boot/vmlinuz-2.6.15-ck1
Found kernel: /boot/vmlinuz-2.6.12-10-686
Found kernel: /boot/memtest86+.bin
Updating /boot/grub/menu.lst ... done

update-rc.d: /etc/init.d/upower exists during rc.d purge (use -f to force)
update-rc.d: /etc/init.d/upower10 exists during rc.d purge (use -f to force)
update-rc.d: /etc/init.d/upower20 exists during rc.d purge (use -f to force)
update-rc.d: /etc/init.d/upower30 exists during rc.d purge (use -f to force)
update-rc.d: /etc/init.d/upower40 exists during rc.d purge (use -f to force)
update-rc.d: /etc/init.d/upower50 exists during rc.d purge (use -f to force)
update-rc.d: /etc/init.d/upower60 exists during rc.d purge (use -f to force)
dpkg: error processing upower (--remove):
subprocess post-removal script returned error exit status 1
Searching for GRUB installation directory ... found: /boot/grub .
Testing for an existing GRUB menu.list file... found: /boot/grub/menu.lst .
Searching for splash image... none found, skipping...
Found kernel: /boot/vmlinuz-2.6.15-ck1
Found kernel: /boot/vmlinuz-2.6.12-10-686
Found kernel: /boot/memtest86+.bin
Updating /boot/grub/menu.lst ... done

Errors were encountered while processing:
upower
Running prelink, please wait...
E: Sub-process /usr/bin/dpkg returned an error code (1)

[TheAnonymousx]

So, I've compiled the new kernel with no real problems and all and then I restarted my machine.
However, upon restarting I notice an error message and then get dumped back to the command line. I try "startx" but get an error message, so I peruse the forums here and install the libqt3 packages and then run "make clean" and "make xconfig." Both (eventually) worked with no problems but I still can't start X. Anyone got ideas?
I'm using an OLD amptron motherboard with everything on board if that helps.
**edit**
Now doing startx gives me the error message:
fatal IO error 104 (connection reset by peer) on xserver 0:0
after 0 request...

[Rob2687]

What is the error message?

[TheAnonymousx]

What is the error message?

Same as the one above, but I think that error message stems from this one...
I do
sudo make clean (from /usr/src/linux)
sudo make xconfig

I get the error message

qconf cannot connect to Xserver
make[1]: [xconfig] Error1
make: ***[xconfig] Error2

The other error I was getting is stated in the earlier comment.
w00t!
Man is this always this involved or am I this dumbtarded?

[Gandalf]

upower doesn't uninstall :(

Removing upower ...
Searching for GRUB installation directory ... found: /boot/grub .
Testing for an existing GRUB menu.list file... found: /boot/grub/menu.lst .
Searching for splash image... none found, skipping...
Found kernel: /boot/vmlinuz-2.6.15-ck1
Found kernel: /boot/vmlinuz-2.6.12-10-686
Found kernel: /boot/memtest86+.bin
Updating /boot/grub/menu.lst ... done

update-rc.d: /etc/init.d/upower exists during rc.d purge (use -f to force)
update-rc.d: /etc/init.d/upower10 exists during rc.d purge (use -f to force)
update-rc.d: /etc/init.d/upower20 exists during rc.d purge (use -f to force)
update-rc.d: /etc/init.d/upower30 exists during rc.d purge (use -f to force)
update-rc.d: /etc/init.d/upower40 exists during rc.d purge (use -f to force)
update-rc.d: /etc/init.d/upower50 exists during rc.d purge (use -f to force)
update-rc.d: /etc/init.d/upower60 exists during rc.d purge (use -f to force)
dpkg: error processing upower (--remove):
subprocess post-removal script returned error exit status 1
Searching for GRUB installation directory ... found: /boot/grub .
Testing for an existing GRUB menu.list file... found: /boot/grub/menu.lst .
Searching for splash image... none found, skipping...
Found kernel: /boot/vmlinuz-2.6.15-ck1
Found kernel: /boot/vmlinuz-2.6.12-10-686
Found kernel: /boot/memtest86+.bin
Updating /boot/grub/menu.lst ... done

Errors were encountered while processing:
upower
Running prelink, please wait...
E: Sub-process /usr/bin/dpkg returned an error code (1)

that's not a problem, do the following

sudo rm /etc/init.d/upower*
sudo apt-get remove --purge upower


@limit223 actually me neither i've been using 2.6.14 since ck2 was out, and i didn't get what was said in that post, but vga=number work for me as suse boot like this so my card support it

[newuser111]

Same as the one above, but I think that error message stems from this one...
I do
sudo make clean (from /usr/src/linux)
sudo make xconfig

I get the error message

qconf cannot connect to Xserver
make[1]: [xconfig] Error1
make: ***[xconfig] Error2

The other error I was getting is stated in the earlier comment.
w00t!
Man is this always this involved or am I this dumbtarded?


i had the same problem i have no idea what causes it but use

sudo make menuconfig instead of sudo make xconfig which worked for me

[TheAnonymousx]

i had the same problem i have no idea what causes it but use

sudo make menuconfig instead of sudo make xconfig which worked for me

Okay so that seems to have worked (had to install some base fonts for X), but now GNOME won't freakin' launch (even though I was allowed to start Xwindows).
I get the login screen, enter my password, and even have the system ATTEMPT to load gnome, then I get this error message:

Error restarting gnome settings daemon
last error message:
Failed to execute /usr/lib/control-center/gnome-settings : daemon 2 (no such file or directory)

Then I'm informed the system will attempt to attempt to reload gnome on the next logon (this is the third attempt though).

Any ideas guys? Maybe I have to interrupt gnome's loading and jump to a command-line interface (however that's done) when it's loading to grab some package or change some setting?

[madchicken]

Sorry guys, I'm lost in this thread :confused:
Anyone has resolved the pmount issue?
I just compiled a 2.6.15+ck+suspend2 with success, but I can't get pmount working.

Any help is appreciated.

Bye

[limit223]

Ok..so far was this case: my compilation was fine with upower except I couldn't get in console, meaning that the splash with the image and the bar from upower theme loaded but F2 and Ctrl+Alt+F2/7 didn't...and of course I went forward to see the what happens.
I recompiled another vanilla 2.6.15-ck1 kernel with using different .config file and upower didn't work at all.
I tried a test:
I installed linux-686-smp with the linux-headers-2.6.12-10 and Upower worked perfectly in my computer with all things: switch in console, splash image, splash bar.... I found some differences of modules loaded and used in lsmod between this kernels(mine and the from the repo) especially regarding those modules suffixed by fb: *fb - tileblit, fbcon, cfbcopyarea... I tried to identify them in repo's kernel .config file to see what I am missing in my compiled kernels.
So far I've got 3 kernels in test: one with fully upower function, one with partial function, one without any response...
Conclusion:I'm pretty sure that usplash or upower are bounded to be supported by some of this modules and they differ from video card to video card Now probably the question are: New kernel 2.6.15 contains all this modules? Do we need extra patches in it? Which are actually they?

My mind of 2 cents...:p

[juantxorena]

Sorry guys, I'm lost in this thread :confused:
Anyone has resolved the pmount issue?
I just compiled a 2.6.15+ck+suspend2 with success, but I can't get pmount working.

Any help is appreciated.

Bye
I've found, in 4 or 5 compilations I've made, that the resolution of pmount is completly aleatory. Upper in this post I attached a conf file with the pmount solved. You can try it, modifying it the lesser as possible.

[rjwood]

I compiled the new 2.6.15-ck kernel and when I reboot I get
"kernel panic-not syncing: VFS: Unable to mount root fs on unknown-block (0,0)" I know it's not mounting the root file system--but- I don't know what I did wrong. I thought I was very careful as this was my second go-round with it. Thanks in advance.

BTW I want to thank you and tselliott for the great 'How-To's"

[fergus.b]

Yay, now my partitons on my other drive will mount! Now I can play planet penguin racer at fullspeed AND access my other drive with my data. Thank you so much! :)
Did you manage to get your Unichrome working mate? If so what was the procedure or did you just do the new kernel + patch?

[madchicken]

I've found, in 4 or 5 compilations I've made, that the resolution of pmount is completly aleatory. Upper in this post I attached a conf file with the pmount solved. You can try it, modifying it the lesser as possible.

Thaks, I already tryed your config file, but using it with suspend2 patch (I need it because I can't suspend with suspsw) I get an error at compile time. I'll try to diff my config with your's to see if there are "big" differences.

Thanks

[tseliot]

I compiled the new 2.6.15-ck kernel and when I reboot I get
"kernel panic-not syncing: VFS: Unable to mount root fs on unknown-block (0,0)" I know it's not mounting the root file system--but- I don't know what I did wrong. I thought I was very careful as this was my second go-round with it. Thanks in advance.

BTW I want to thank you and tselliott for the great 'How-To's"
I have some questions for you:
1) What's the model of your motherboard?
2) Did you compile the support for ext3 filesystem in the kernel (NOT as a module)?

I'm asking you those questions because I can't compile any vanilla kernel (well I can but it doesn't boot). Even Dapper's kernels won't boot. It says it doesn't find the kernel. I have filed a bug in the Ubuntu Bugzilla. I hope they solve the bug otherwise there will be no Dapper for me :(

I hope your case is different from mine.

[rjwood]

I have some questions for you:
1) What's the model of your motherboard?
2) Did you compile the support for ext3 filesystem in the kernel (NOT as a module)?

I'm asking you those questions because I can't compile any vanilla kernel (well I can but it doesn't boot). Even Dapper's kernels won't boot. It says it doesn't find the kernel. I have filed a bug in the Ubuntu Bugzilla. I hope they solve the bug otherwise there will be no Dapper for me :(

I hope your case is different from mine.

1) * Manufacturer's name - ASUS A7N8X-LA
2)I thought I was real careful. I am almost positve I compiled it for ext3. Is there a way for me to check and be sure? I hope it was my mistake and my motherboard is not too old.

[tseliot]

1) * Manufacturer's name - ASUS A7N8X-LA
2)I thought I was real careful. I am almost positve I compiled it for ext3. Is there a way for me to check and be sure? I hope it was my mistake and my motherboard is not too old.

you can find the config file of your bios under /boot
you will find files such as config-name_of_your_kernel , etc.

Open the file as a text and use the Find function. Search "ext3".

BTW my motherboard is not old it's just an Ubuntu bug (I wonder if it's the same in Debian but I guess not since Kanotix works). It works great in Fedora Core 4 which uses kernel 2.6.14.1.

[rjwood]

Thanks for the help tseliot do this look correct?

#
# File systems
#
# CONFIG_EXT2_FS is not set
CONFIG_EXT3_FS=y
CONFIG_EXT3_FS_XATTR=y
CONFIG_EXT3_FS_POSIX_ACL=y
CONFIG_EXT3_FS_SECURITY=y
CONFIG_JBD=y
CONFIG_JBD_DEBUG=y
CONFIG_FS_MBCACHE=y
CONFIG_REISERFS_FS=m
# CONFIG_REISERFS_CHECK is not set
# CONFIG_REISERFS_PROC_INFO is not set
CONFIG_REISERFS_FS_XATTR=y
CONFIG_REISERFS_FS_POSIX_ACL=y
CONFIG_REISERFS_FS_SECURITY=y
CONFIG_JFS_FS=m
CONFIG_JFS_POSIX_ACL=y
CONFIG_JFS_SECURITY=y
# CONFIG_JFS_DEBUG is not set
CONFIG_JFS_STATISTICS=y
CONFIG_FS_POSIX_ACL=y
CONFIG_XFS_FS=m
CONFIG_XFS_EXPORT=y
CONFIG_XFS_QUOTA=y
CONFIG_XFS_SECURITY=y
CONFIG_XFS_POSIX_ACL=y
CONFIG_MINIX_FS=m
CONFIG_ROMFS_FS=m
CONFIG_INOTIFY=y
CONFIG_QUOTA=y
CONFIG_QFMT_V1=m
CONFIG_QFMT_V2=m
CONFIG_QUOTACTL=y
CONFIG_DNOTIFY=y
CONFIG_AUTOFS_FS=m
CONFIG_AUTOFS4_FS=m
# CONFIG_FUSE_FS is not set

#
# CD-ROM/DVD Filesystems
#
CONFIG_ISO9660_FS=m
CONFIG_JOLIET=y
CONFIG_ZISOFS=y
CONFIG_ZISOFS_FS=m
CONFIG_UDF_FS=m
CONFIG_UDF_NLS=y

#
# DOS/FAT/NT Filesystems
#
# CONFIG_MSDOS_FS is not set
# CONFIG_VFAT_FS is not set
# CONFIG_NTFS_FS is not set

#
# Pseudo filesystems
#
CONFIG_PROC_FS=y
CONFIG_PROC_KCORE=y
CONFIG_SYSFS=y
CONFIG_TMPFS=y
# CONFIG_HUGETLBFS is not set
# CONFIG_HUGETLB_PAGE is not set
CONFIG_RAMFS=y
# CONFIG_RELAYFS_FS is not set

#
# Miscellaneous filesystems
#
# CONFIG_HFSPLUS_FS is not set
# CONFIG_CRAMFS is not set
# CONFIG_VXFS_FS is not set
# CONFIG_HPFS_FS is not set
# CONFIG_QNX4FS_FS is not set
# CONFIG_SYSV_FS is not set
# CONFIG_UFS_FS is not set

[limit223]

I've got Upower full working with cute console in my machine !! I did, I did...happpy:)

Comparing and searching I found those modules important for console are: vesafb and fbcon. In .config file making xconfig I checked:
Graphics support:
-VGA 16-color graphics support - module (m)
-VESA VGA graphics support - build in kernel(y)
-
Console display driver support:
-VGA text console and Video mode selection support- build in kernel (y)
-MDA text console-module (m)
-Framebuffer Console and Framebuffer Console Rotation support-build in kernel (y)
-Select compiled-in fonts with VGA 8x8 fonts, VGA 8x16 font - build in kernel(y)
-Sparc console 12x12 font- build in kernel (y)


Some referrence you may find here:
http://www.icewalkers.com/Linux/Howto/Framebuffer-HOWTO-5.html
http://gentoo-wiki.com/HOWTO_Framebuffer:Bootsplash:Grubsplash#Smaller_fo nts_on_framebuffer_for_greater_area
http://ubuntuforums.org/showthread.php?t=90689

For the only thing I was afraid was enabling 3D accel side effect. Reading goodchild and heftigrat's posts from here http://ubuntuforums.org/showthread.php?t=76116&page=56 remind me of my console problems. But after this new compilation, installing fglrx-kernel*.deb generated from compilation, enabling 3D(by doing
cd /lib/modules/fglrx/build_mod
sudo ./make.sh
cd ..
sudo ./make_install.sh
Reboot)

I was amazed to see that I've got them both: switching in a cute console without any problem in boot splash time and 3D accel in X time.

A Simple Reflection:
Compiling a new ver or even different build of the same vanilla kernel is not really as simple to copy .config file from an old ver and pass it to new one. Every time you should stay tuned to new changes that are made especially regarding headers that support your hardware components. I do not suggest to anyone to copy and paste that .config file in every compilation without seeing exactly what headers are configured in it to be build in or as module. If things changed and there are bugs for a certain module you may go see if some of patch providers would already issue a patch that may solve the problem.
For sure I'm not an expert but I really enjoy trying new things in open source starting with first layer in my machine: the kernel.

[tseliot]

Thanks for the help tseliot do this look correct?
Yes, it is.

Do you have that problem only with kernel 2.6.15? Did you try other vanilla kernels?

I didn't have kernel panic, just a GRUB error kernel not found therefore your problem might be different from mine.
Have a look at the bug I filed: http://bugzilla.ubuntu.com/show_bug.cgi?id=21652

Another thing you could do is to try Dapper's livecd (it uses kernel 2.6.15) and see if it works.

[rjwood]

comparing file system choices in this new kernel and my other one. My 2.6.12-10k7 kernel has all the file systems as modules and the subsys choices are set to yes. Could that be a problem? how do I go back in and make changes??

[tseliot]

...My 2.6.12-10k7 kernel has all the file systems as modules...
That's absolutely normal for Ubuntu kernels. On the other hand if you use vanilla kernels if you leave the filesystem of the partition with your OS as a module the result is kernel panic.

[rjwood]

I think I may know what my problem is.
In block devices, there is box for "config_blk_dev_initrd" I am not able to check that device and the kernel message I get on boot seems to indicate a similarity to this options description. I see that other people who have shared their config file have this checked. Can anyone help me here. I tried doing it as root and with sudo.

[macsimski]

Hello all,

i've tried to build my first kernel but i got a kernel panic during boot. must be something i set incorrect in xconfig.

so i want ot go back to my previous kernel, the default breezy kernel. boots up fine, albeit slow, but it boots. then i get to the login screen and i type my pass. screen turns black, a grey screen comes up with a watch cursor and bang, i'm back at the login screen.

but when i switch to a virtual terminal and log in with my own account and then do sudo -s and startx my session starts and am in my beloved xfce again.

I think that there must be something wrong with the permissions, but what?

-is there a program to check permissions like in Mac OS X?
-is there a way to revert to the standard kernel installed by Kunbuntu without a complete reinstall??

[edit] found the problem. somewhere during the described building process i did'nt use sudo but in stead used sudo -s -H so i actually became root. some rights in my home folder got th owner of root.root. when i changed them to to myself everything works as before. Phew!

maybe i will try builing a kernel later...

[moment]

The only problem that I had was that I had to modprobe "evdev" module in order to get my alps touchpad to work again. Thanks!

[Brando569]

Same as the one above, but I think that error message stems from this one...
I do
sudo make clean (from /usr/src/linux)
sudo make xconfig

I get the error message

qconf cannot connect to Xserver
make[1]: [xconfig] Error1
make: ***[xconfig] Error2

The other error I was getting is stated in the earlier comment.
w00t!
Man is this always this involved or am I this dumbtarded?

i believe its because in the terminal session your logged in as root but in your X session your logged in as your username and the terminal tries to connect to the root X session and cant. a simple fix it to SU back to your username and use this: sudo make xconfig

[rjwood]

Took me 5 attempts but I got going flawlessly. Thanks everyone especially tseliott (your the best).

[firecat53]

Hey...successfully compiled the 2.6.15 kernel with ck1 patch using gcc-4.0.

I initially had trouble with the cardbus PCMCIA cards being detected, but a second reboot cured that problem. Now I just have to figure out the splash screen issue and fix the touchpad scroll feature. Already recompiled madwifi and wpa_supplicant.

Good how-to :)

Thanks, Scott

Does anyone have any suggestions for fixing the touchpad vertical scroll feature? I've been looking and I'm getting stumped. All features of the touchpad work fine except the vertical scroll area. Here's the output from "cat /var/log/Xorg.0.log|grep synaptics"

(II) LoadModule: "synaptics"
(II) Loading /usr/X11R6/lib/modules/input/synaptics_drv.o
(II) Module synaptics: vendor="The XFree86 Project"
Synaptics Touchpad no synaptics event device found (checked 1 nodes)
(EE) Synaptics Touchpad no synaptics touchpad detected and no repeater device
(II) UnloadModule: "synaptics"

Hmmm...so why would the touchpad not be detected but it all works except for one feature??

[tseliot]

Took me 5 attempts but I got going flawlessly. Thanks everyone especially tseliott (your the best).
I'm happy it worked in the end :)

[TheAnonymousx]

Alright, so anyone got some ideas for the problem I posted previously? Basically, with the make menuconfig I can load far enough to get past the splash screen, but right after that I get an error message that says gnome doesn't have any default setting to load.
Anyone got a suggestion?
Also, I should know this, but how DO I bring up the command line when the machine first boots?

[nevienc]

root@nn:/usr/src/linux# sudo make-kpkg -initrd --revision=ck1 kernel_image
test -f stamp-configure || /usr/bin/make -f /usr/share/kernel-package/rules configure
/usr/bin/make ARCH=i386 \
bzImage
make[1]: Entering directory `/usr/src/linux-2.6.14.6cK1'
CHK include/linux/version.h
CHK include/linux/compile.h
CHK usr/initramfs_list
Kernel: arch/i386/boot/bzImage is ready (#1)
make[1]: Leaving directory `/usr/src/linux-2.6.14.6cK1'
/usr/bin/make ARCH=i386 \
modules
make[1]: Entering directory `/usr/src/linux-2.6.14.6cK1'
CHK include/linux/version.h
CC [M] drivers/net/bonding/bond_main.o
drivers/net/bonding/bond_main.c: In function ‘bond_compute_features’:
drivers/net/bonding/bond_main.c:1617: error: ‘struct bonding’ has no member named ‘bond_features’
drivers/net/bonding/bond_main.c: In function ‘bond_init’:
drivers/net/bonding/bond_main.c:4511: error: ‘struct bonding’ has no member named ‘bond_features’
make[4]: *** [drivers/net/bonding/bond_main.o] Error 1
make[3]: *** [drivers/net/bonding] Error 2
make[2]: *** [drivers/net] Error 2
make[1]: *** [drivers] Error 2
make[1]: Leaving directory `/usr/src/linux-2.6.14.6cK1'
make: *** [stamp-build] Error 2
help please!

[LightStruck]

help please!

Ok if its the problem I had then the ck1 patch wasn't playing nicely with my kernel. Because the patches are supposed to be applied to a simple 2.6.14 kernel. And it seems to me you're trying to compile a 2.6.14.6 kernel? Correct me if I misunderstood your message. But simply if you want to use the ck1 patch you need to use a regular 2.6.14 vanilla kernel from kernel.org

[LightStruck]

Also, I responded to his post but I've been having a really really tough time getting a working kernel config going. I used a 2.6.14 and patched it to ck3 and the results were gorgeous. I was amazed how fast it was. I was like "OK! Tune more!" And things fell apart. I was suffering from the pmount support I believe early on because it wasn't loading my local filesystems but every compile since has installed fine I choose the 2.6.14 in my grub menu and the "Uncompressing Linux" comes up and then all the font on the screen just shifts and freezes. No real loading just stops at a font shift. I've recompiled around 12 times with one success. Anyone got any tips? I've been using the .config file from the 2.6.12-10-686-smp and changed it to my p4 processor and used the tweaks that were suggested at the beginning but no go. Just seeing if anyone had this problem and cured it. Thanks for your time.

[nastya]

Did you manage to get your Unichrome working mate? If so what was the procedure or did you just do the new kernel + patch?

Yup it worked just fine. I had to recomplile my kernel to do it though. In case anyone else had trouble finding where the option was located, here's what I did. In the configuration under device drivers -> character devices, go to direct rendering manager, and under that is the option to add support for via unichrome video cards. After that, configure everything else the way you'd like, and compilie and it should work. :)

[kaamos]

Just upgraded to 2.6.15cK1. Works great otherwise, but couldn't still get pmount working. :(

[Bloot]

Just upgraded to 2.6.15cK1. Works great otherwise, but couldn't still get pmount working. :(

It's really annoying, I'd like to use new kernels but that way it's not possible. I guess we'll have to wait to a newer and fixed version of pmount...

Greetings.

[kaamos]

Just noticed that the tty terminals (ctrl+alt+F1 - F6) are empty. I have "CONFIG_FB_VESA=y" the .config, and I thought that this fixes the problem. Usplash works though.. Does anyone know what config option would fix the terminals?

[rulppa]

deleted.

[newuser111]

ive had good results compiling a kernel with the archck (based on ck) patchset

http://iphitus.loudas.com/archck.php

btw i recommend ati users stick to 2.14 kernels until ati fixes some serious system instability caused by using 2.15 with ati fglrx

[ErikTheRed]

I followed the guide to compile and install the 2.6.15.1 kernel, but I've run into a few oddities. I ended up omitting the cK patches, because I would get a hunk failure when I tried to patch (I tried both cK1 and cK2). Also for some reason the pretty kubuntu splash screen does not show up like usual, I just have a blank screen while the kernel, modules, etc load up. Everything else works fine though.

[newuser111]

try the archck patch instead, it includes the bootsplash patches

http://iphitus.loudas.com/archck.php