apt-get kernel 2.6

[hijinks 40]

want to know how to install 2.6 via apt. Well its easy. Add the following line to your /etc/apt/sources.list file

 

rpm http://people.redhat.com arjanv/2.6 kernel

 

then run the following commands

 

apt-get update
apt-get install kernel#2.6.1-1.131

 

that was when 2.6.1-1.131 was the newest. You can get a complete listing by running

 

apt-get install kernel

 

NOTE: This was done on a fedora core1 box. RH 9 might have some depandancy issues

[hijinks 40]

Ok I just installed it using RH 9.

 

You need to do an upgrade because it needs a version of mkinitrd and a few other rpms that aren't shipped with RH 9 but that apt-get site has them for download

 

apt-get update
apt-get upgrade
apt-get install kernel#2.6.1-1.131

[anyweb 490]

ive tried this in fedora core release 1

and while the apt-get part worked perfectly, the kernel itself was a total disasater on my laptop (dell latitude c400)

 

it completelyhung during boot (after kudzu and some other services were started)

 

a reboot started fsck which FAILED and forced another reboot,

 

needless to say i tried on and off about 6 times and i could not get the kernel to work

 

(2.6.1.141)

 

so for me it was a disaster,

 

be careful !

 

cheers

 

anyweb

[grep420 2]

No problems at all for me. I have done this on 9 and fedora.

 

[root@xwing root]# date && cat /etc/redhat-release && uname -a

Thu Jan 22 11:11:19 CST 2004

Fedora Core release 1 (Yarrow)

Linux xwing.dal2.mindspring.net 2.6.1-1.141 #1 Wed Jan 21 04:33:51 EST 2004 i686 i686 i386 GNU/Linux

[deraj 0]

that's why you should just use gentoo, it's as easy as ...

 

emerge genkernel
emerge gentoo-dev-sources
genkernel --config

[grep420 2]

I can do it in a single line with redhat:

apt-get install kernel-smp#2.6.1-1.141

[deraj 0]

apt-get schmat-get, but does you being able to do it in one command make you special?

[grep420 2]

no, just pointing out that redhat is easier than gentoo, and much less of pain to set up.

[robbin 1]

Hi there, I wonder if someone could say why this has happened, I have updated Ubuntu Ibex to the Ubuntu Jaunty version via the update manager only the kernel says its still Ubuntu 8.10, kernel 2.6.27-14-generic, can I change this and how would l do that, feel like I read my eyes out! now I'm near to Boot and Nuke....thought I'd ask first.

 

Also when I start the terminal after the new install I get these two lines first.

 

bash: /home/china/.bashrc: line 102: unexpected EOF while looking for matching `"'

bash: /home/china/.bashrc: line 103: syntax error: unexpected end of file

 

robbins@desktop:~$

 

I have took some screen shots of other errors on this install, still abit new but if you got time I'm all ears, thanks again

[anyweb 490]

did you reboot yet ? kernel updates require a reboot

[znx 24]

Hi there, I wonder if someone could say why this has happened, I have updated Ubuntu Ibex to the Ubuntu Jaunty version via the update manager only the kernel says its still Ubuntu 8.10, kernel 2.6.27-14-generic, can I change this and how would l do that, feel like I read my eyes out! now I'm near to Boot and Nuke....thought I'd ask first.

 

The kernel saying that it isn't updated could be that the kernel is restricted from updating (not uncommon). Therefore you need to specify an upgrade to the kernel manually:

apt-get update kernel

 

Then reboot into the new kernel.

 

The unrelated problem to your kernel upgrade and is more likely a personal alteration gone wrong. The /home/china/.bashrc is a personal configuration file for the user "china".

 

I suggest that you either edit the file and fix the problem or simply move the file out of the way:

mv /home/china/.bashrc /home/china/bashrc-old

 

As suggested by the problem, the line 102 (end of the file) is where the problem is detected. You are missing a " quote somewhere inside the file.

[robbin 1]

Thanks man, I'm kinda getting into this "he says"

 

I still ain't got much to add! if you could see past how little I know about Linux and help like this I'm sure we'll have it all sorted out in no time.

I'm am interested in building my own kernel and a new OS is only a format away! Only why do we do this apart from its new. I say my PCs online before it goes to the BIOS call and boots or are they all like this ? Put me right where I'm wrong.

 

My question! will making my own kernel this side of the fence make the pc mine or will I forever be connected to an unknowing force from the

udi:/org/freedesktop/hal/devices/net and still getting messed with I wonder :-P

 

I pay for a 10mb line from Virgin but it may aswell be on 28k sometimes I don't mind sharing if thats what they call it but their not having it all! Its not the PC I don't believe yet as I've had better off SKYs free 1mb phone line.

 

Also how would I make the mouse tell me its not a mac mouse, I would really like to learn how to edit the system only I need someone to say look open such and such and faint in there dude. If I don't know soon I'm off to learn to juggle....

 

Thanks again for the help.

[Dungeon-Dave 293]

Just relating my experiences...

I recently build a Ubuntu 10.10 box (logan) from a unetbootin image (boot from USB pen), then used apt-get to install apps I was missing as well as updating current packages to their newer counterparts, but came across something interesting: one of my other boxes (neptune), updated to the hilt, had kernel version 2.6.35-31-generic... however after an "apt-get update" and "apt-get upgrade" to bring all the packages up to their latest, logan was still on 2.6.35-22. Weird.

 

I could run "apt-get install linux-headers-2.6.35-31" (and also linux-headers-2.6.35-31-generic) which would bring the newer kernel down, but after a reboot I was still on 2.6.35-22.

 

Eventually I fathomed out the issue, which required two things to be done:

1. Install linux-image-2.6.35-31-generic
   
2. Run "update-grub".
   

The second command rebuilds your grub.cfg, but the "10_linux" script which adds kernel images to the boot loader only finds them if the "linux-image" package is installed, so consequently any manually-installed kernels were being overlooked previously.

 

Once I'd put the right package in place (the install of which re-ran "update-grub" automatically) a reboot showed the newer kernel in use, and the older ones could be cleared out with "apt-get remove linux-headers-2.6.35-22" etc.

 

Just surprised that an "apt-get upgrade" didn't bring down newer kernels, but hope those are pointers for anyone who wants to do kernel updates.

[hybrid 189]

Sometimes I have had problems with Ubuntu boxes that won't update the kernels via apt-get update/upgrade commands. I often see the kernel packages with the message "the following packages have been kept back". Oddly, running the Update Manager graphically does update the packages, but apt-get over the command line doesn't. It's odd, and I haven't yet found a solution other than manually installing the new kernel packages.

[Dungeon-Dave 293]

okay.. interesting.. I may try the graphical updater next, see what that does.

[hybrid 189]

You might be interested to know that I ran into this issue again recently on 11.10 x86 -- most of the updates ran fine via apt-get update, then apt-get upgrade, but the kernel (and in this instance, Evolution packages) were 'kept back' and were only offered properly by the graphical updater.

 

It's a really odd issue, and I don't know a way, from the command line, to unmark the packages to be 'kept back' without additional software (I seem to recall that wajig unhold would work, but it's an extra package to have installed).

 

[Dungeon-Dave 293]

I found that apt-cache search kernel would show me the newer kernels, and apt-get install of the later "linux-header" package would force the others as dependencies. But you're right that, as a manual process, an upgrade from the command line ought to have fixed this.

 

Weird.