LinuxQuestions.org

LinuxQuestions.org (/questions/)
-   Slackware (http://www.linuxquestions.org/questions/slackware-14/)
-   -   slackware 14.1 3.10.17 kernel question on setting up generic kernel (http://www.linuxquestions.org/questions/slackware-14/slackware-14-1-3-10-17-kernel-question-on-setting-up-generic-kernel-4175489417/)

Bertman123 12-28-2013 05:19 PM

slackware 14.1 3.10.17 kernel question on setting up generic kernel
 
I'm on 14.1 32 bit and successfully installed the generic kernel, but find that I am now unable to boot into the huge kernel. Is that normal?

I'm just happy that I can boot into my system and have my wireless working this time. :D But am curious about the huge kernel now.

TobiSGD 12-28-2013 05:53 PM

The huge and the generic kernel are independent from each other, so if you have your Lilo setup correctly (and generated an initrd for the generic kernel) there shouldn't be a problem with any of them. Please give us a more specific error description.

Bertman123 12-28-2013 06:04 PM

Quote:

Originally Posted by TobiSGD (Post 5088243)
The huge and the generic kernel are independent from each other, so if you have your Lilo setup correctly (and generated an initrd for the generic kernel) there shouldn't be a problem with any of them. Please give us a more specific error description.

Okay, give me a little bit to write down the exact error message. It looks like a kernel panic message, but I'll write it down when I reboot.

Bertman123 12-28-2013 06:08 PM

Quote:

Originally Posted by TobiSGD (Post 5088243)
The huge and the generic kernel are independent from each other, so if you have your Lilo setup correctly (and generated an initrd for the generic kernel) there shouldn't be a problem with any of them. Please give us a more specific error description.


Here's my lilo.conf to start. I'm not sure if this is how it's supposed to look.

# Linux bootable partition config begins
image = /boot/vmlinuz
root = /dev/sda2
label = Linux
read-only
# Linux bootable partition config ends
# Linux bootable partition config begins
image = /boot/vmlinuz-generic-smp-3.10.17-smp
initrd = /boot/initrd.gz
root = /dev/sda2
label = Lnx31017
read-only
# Linux bootable partition config ends

gengisdave 12-28-2013 06:22 PM

/boot/vmlinuz is a symlink, try 'ls /boot/vmlinuz' to see which kernel version is pointing; you can add another boot option too

Code:

image = /boot/vmlinuz-huge-smp-3.10.17-smp
root = /dev/sda2
label = Lnx31017
read-only

because huge version has fs driver compiled built-in, you don't need initrd, unless you have strange hardware

allend 12-28-2013 06:24 PM

My bet is that the /boot/vmlinuz symlink now points to the generic kernel vmlinuz-generic-smp-3.10.17-smp.

Change 'image = /boot/vmlinuz' to point to the huge kernel image and rerun lilo.

Bertman123 12-28-2013 08:31 PM

Quote:

Originally Posted by allend (Post 5088250)
My bet is that the /boot/vmlinuz symlink now points to the generic kernel vmlinuz-generic-smp-3.10.17-smp.

Change 'image = /boot/vmlinuz' to point to the huge kernel image and rerun lilo.

Is it either the generic kernel or the huge? I found the instructions to install the generic kernel, but does that mean I can not boot the huge? What happens if an update for the kernel comes out, or if I upgrade to another kernel? Do I need to stay to the generic kernel? I just want to know so as to not mess up my system.

TobiSGD 12-28-2013 08:51 PM

You don't need that symlink at all (and yes, a symlink can only point to one file). So just replace it in your lilo.conf with the actual names of the kernels. Which one you boot is up to you, you can decide at boot if you set up Lilo correctly.
For clarification, kernels in a release version of Slackware are very rarely updated (usually only if there is a serious security problem or a significant bug). But even if they are, you usually never update them, you install the new kernel side-by-side with the old one, so that you can actually test it before switching over.

Bertman123 12-29-2013 06:55 AM

Thanks, adding the name of the huge kernel worked and I can boot into both. More importantly I learned about inatalling generic kernels. Thanks for all of your help.

Bertman123 01-11-2014 03:17 PM

Update:

I did successfully install the generic kernel on my laptop. And only have one generic kernel installed. I then went to work on my HP tower that has an NVIDIA 6150 SE Video card. It was running slackware 14.0 64 bit and I also successfully installed the generic kernel on that too, but did not have the proprietary NVIDIA drivers installed. I was using the nouveau drivers.

I blacklisted the kernel in slackpkg and proceeded to upgrade to 14.1, which went successfully and everything is running well with the 3.2.45 kernel from 14.0. I proceeded to then install the generic kernel for 14.1 following the directions from here: http://mirrors.slackware.com/slackwa.../README.initrd and here: http://docs.slackware.com/slackware:beginners_guide.

I can successfully log in with both kernels and tried installing the proprietary NVIDIA legacy driver, but that failed to install. When I'm logged in with the 3.10.17 kernel it looks like the proprietary NVIDIA driver is trying to install it on the 3.2.45 kernel.

Could I have missed installing something? Everything seems to be working right now, but I'd like to know for the future.

These were the packages that I installed: kernel-generic-3.10.17-x86_64-3.txz, kernel-modules-3.10.17-x86_64-3.txz, & mkinitrd-1.4.8-x86_64-1.txz.

Didier Spaier 01-11-2014 03:41 PM

Whenever you change the kernel you have to build new nvidia driver for it, you can't use the old one.

Bertman123 01-11-2014 03:43 PM

Quote:

Originally Posted by Didier Spaier (Post 5096294)
Whenever you change the kernel you have to build new nvidia driver for it, you can't use the old one.

That's the strange thing, I didn't have it installed prior to the upgrade.

Didier Spaier 01-11-2014 03:59 PM

Maybe I misunderstood what you wrote:
Quote:

I can successfully log in with both kernels and tried installing the proprietary NVIDIA legacy driver, but that failed to install. When I'm logged in with the 3.10.17 kernel it looks like the proprietary NVIDIA driver is trying to install it on the 3.2.45 kernel
How do you know that? And what exact steps did you follow to install the proprietary NVIDIA legacy driver?

enorbet 01-11-2014 05:49 PM

Quote:

Originally Posted by Bertman123 (Post 5096279)
Update:
<snip>
I can successfully log in with both kernels and tried installing the proprietary NVIDIA legacy driver, but that failed to install. When I'm logged in with the 3.10.17 kernel it looks like the proprietary NVIDIA driver is trying to install it on the 3.2.45 kernel.
<snip>
These were the packages that I installed: kernel-generic-3.10.17-x86_64-3.txz, kernel-modules-3.10.17-x86_64-3.txz, & mkinitrd-1.4.8-x86_64-1.txz.

There is really good documentation for each nVidia linux driver. For me I find it easiest to execute the NVIDIA-xxxxx.run file with the "--extract-only" switch, so I can get at all the info in one known place.

Example - This is from the Readme.txt

Quote:

Originally Posted by nvidia readme.txt
4C. INSTALLING THE KERNEL INTERFACE

The NVIDIA kernel module has a kernel interface layer that must be compiled
specifically for each kernel. NVIDIA distributes the source code to this
kernel interface layer.

When the installer is run, it will check your system for the required kernel
sources and compile the kernel interface. You must have the source code for
your kernel installed for compilation to work. On most systems, this means
that you will need to locate and install the correct kernel-source,
kernel-headers, or kernel-devel package; on some distributions, no additional
packages are required.


After the correct kernel interface has been compiled, the kernel interface
will be linked with the closed-source portion of the NVIDIA kernel module.
This requires that you have a linker installed on your system. The linker,
usually '/usr/bin/ld', is part of the binutils package. You must have a linker
installed prior to installing the NVIDIA driver.

Actually I had thought that one cannot build the nvidia driver without fully qualified "/usr/src/linux/.config" meaning iirc at least "make modules&&make modules_install" must be actually run, and I always have gone through the full process of building a custom kernel. But the above quotes (bold part) seem to imply one might only need to add the "kernel-devel" package.

However I am confident that if you read that Readme, the "lights will go on for you" as to how much you need to do to prepare a proper environment on your box.

Richard Cranium 01-11-2014 09:53 PM

How did you did you do the upgrade from 14.0 to 14.1?


All times are GMT -5. The time now is 08:45 AM.