Linux Kernel Modules Installation HOWTO

1. Purpose of this Document

My experience with Linux and modules has been that the existing documents fail to provide a satisfactory explanation as to how to successfully set up Linux with modules configured and working. The procedure explained in this document has been successfully used several times, both on my own system and over the Internet to give directions to somebody trying to get some feature to work which requires a driver supplied only in module form.

My own system runs from a RedHat 4.1 distribution of Linux, and it was on this setup that I developed the procedure. I have since successfully installed it on systems running from various Slackware distributions, and on one system running from a Debian distribution, and the necessary procedure to correctly configure modules under Linux in all three is documented herein.

I have recently used the same procedure with RedHat 4.2, but with inconsistent results on apparently identical systems. I have not yet determined what the problem is, so can make no guarantees at this stage as to whether or not it will work on your system.

This document is distributed under the terms of the GNU Free Documentation License. You should have received a copy along with it. If not, it is available from http://www.fsf.org/licenses/fdl.html.

2. Pre-requisites

• Before the steps in this document can be applied, the reader must have a working Linux installation in which one can get to the Linux prompt as user root since the majority of the steps involved can only be undertaken by the said user.

• The existing kernel may be compiled either to use modules or not to use modules, and can even display error messages during the boot-up procedure as a result of modules being configured which aren't available at the moment, providing the above condition is met.

• The source tree for the current kernel is assumed to be found rooted at /usr/src/linux and that is also assumed to be the current directory throughout this document at the start of any sequence of commands to be issued.

3. Compiler Speed-up

If your machine has 16 or more Megabytes of RAM, there is a useful speed-up that can be done, which is to permit the kernel to compile two or modules in parallel. This will increase the load on the machine whilst the kernel is being recompiled, but will reduce the time during which the compilation will be taking place.

Before you can use this method, you need to check the amount of RAM present in your machine, as if you set this too high, the compilation will actually slow down. Experience has shown that the optimum value depends on the amount of RAM in your system according to the following formula, at least for systems with up to 32 Megabytes of RAM, although it may be a little conservative for systems with larger amounts of RAM:

N = [RAM in Megabytes] / 8 + 1

For the benefit of those with a dislike of math, the values for the common amounts of RAM are as follows:

When you have decided on the correct number, edit the file /usr/src/linux/Makefile and find the line that currently reads:

MAKE=make
    

Replace it with one reading:

MAKE=make -j N
    

where N is the number determined above.

4. Recompiling the Kernel for Modules

The kernel can be reconfigured to use modules for everything other than the file system mounted as root (in most cases, this is the ext2 file system).

However, there are certain items that appear to be difficult to set up properly as modules, so I would recommend the following be compiled into the kernel:

• Ethernet hardware drivers.

• SCSI CD-ROM drivers.

On the other hand, there are certain driver combinations that ONLY work as modules, especially combinations of two or more of the following group:

• A Parallel Printer,

• A Parallel Port drive, such as the IOMEGA ZipDrive or JazzDrive, or the BackPack CD-ROM drive, and

• The PLIP Daemon.

You will need to decide what you are compiling into the kernel, and what as modules, but should take the above points into consideration. The actual choices are made during the compilation, by the second of the following sequence of instructions:

cd /usr/src/linux
make menuconfig
make dep clean modules modules_install zImage
    

Having done that, the module dependencies need to be mapped out. This is done with the following command:

depmod -a
    

The new kernel now needs to be inserted in the boot chain. I am assuming the reader is using LILO for this purpose, since this is the only loader I have any experience with.

I recommend that one does NOT automatically insert the newly compiled kernel as the default Linux kernel since if it should fail, it is then extremely difficult to recover one's Linux setup without doing a complete reinstallation, which is not to be recommended. For this reason, I have the following entry in my /etc/lilo.conf file:

image=/usr/src/linux/arch/i386/boot/zImage
  label=new
  alias=n
  read-only
  vga=ask
  optional
    

This entry says that there is an OPTIONAL boot option (which will be ignored if the image in question does not exist) which boots the file /boot/newlinux if selected, and allows one to select the video mode it is to be booted in.

Assuming the existence of the above entry in /etc/lilo.conf the revised kernel is already correctly located at the end of compilation, and it can be installed via the following command:

lilo
    

Having done that, the reader needs to follow the further steps relevant to their selected distribution, as follows:

cd /etc/rc.d
ls -l *.d rc.*