Dracut Modules

Writing modules

Some general rules for writing modules:

  • Use one of the inst family of functions to actually install files on to the initramfs. They handle mangling the pathnames and (for binaries, scripts, and kernel modules) installing dependencies as appropriate so you do not have to.

  • Scripts that end up on the initramfs should be POSIX compliant. dracut will try to use /bin/dash as /bin/sh for the initramfs if it is available, so you should install it on your system — dash aims for strict POSIX compliance to the extent possible.

  • Hooks MUST be POSIX compliant — they are sourced by the init script, and having a bashism break your user’s ability to boot really sucks.

  • Generator modules should have a two digit numeric prefix — they run in ascending sort order. Anything in the 90-99 range is stuff that dracut relies on, so try not to break those hooks.

  • Hooks must have a .sh extension.

  • Generator modules are described in more detail later on.

  • We have some breakpoints for debugging your hooks. If you pass rdbreak as a kernel parameter, the initramfs will drop to a shell just before switching to a new root. You can pass rdbreak=hookpoint, and the initramfs will break just before hooks in that hookpoint run.

Also, there is an attempt to keep things as distribution-agnostic as possible. Every distribution has their own tool here and it’s not something which is really interesting to have separate across them. So contributions to help decrease the distro-dependencies are welcome.

Module format

Most of the functionality that dracut implements are actually implemented by dracut modules. dracut modules live in modules.d, and have the following structure:

dracut_install_dir/modules.d/
	00modname/
		module-setup.sh
		<other files as needed by the hook>
00modname

The name of the module prefixed by a two-digit numeric sort code.

The numeric code must be present and in the range of 00 - 99.

Range 50 - 59 is reserved for out of tree (3rd party) dracut modules.

Modules with lower numbers are installed first. This is important because the dracut install functions (which install files onto the initrd) refuse to overwrite already installed files. This makes it easy for an earlier module to override the functionality of a later module, so that you can have a distro or system specific module override or modify the functionality of a generic module without having to patch the more generic module.

module-setup.sh

dracut sources this script to install the functionality that a module implements onto the initrd. For the most part, this amounts to copying files from the host system onto the initrd in a controlled manner.

module-setup.sh function API

install()

This function of module-setup.sh is called to install all non-kernel files.

dracut supplies several install functions that are specialized for different file types. Browse through dracut-functions for more details.

dracut also provides a $moddir variable if you need to install a file from the module directory, such as an initrd hook, a udev rule, or a specialized executable.

installkernel()

This function of module-setup.sh is called to install all kernel related files.

check()

dracut calls this function to check and see if a module can be installed on the initrd.

When called without options, check should check to make sure that any files it needs to install into the initrd from the host system are present. It should exit with a 0 if they are, and a 1 if they are not.

When called with $hostonly set, it should perform the same check that it would without it set, and it should also check to see if the functionality the module implements is being used on the host system.

For example, if this module handles installing support for LUKS encrypted volumes, it should return 0 if all the tools to handle encrypted volumes are available and the host system has the root partition on an encrypted volume, 1 otherwise.

depends()

This function should output a list of dracut modules that it relies upon.

An example would be the nfs and iscsi modules, which rely on the network module to detect and configure network interfaces.

Any other files in the module will not be touched by dracut directly.

You are encouraged to provide a README that describes what the module is for.

Hooks

init has the following hook points to inject scripts:

/lib/dracut/hooks/cmdline/*.sh

scripts for command line parsing

/lib/dracut/hooks/pre-udev/*.sh

scripts to run before udev is started

/lib/dracut/hooks/pre-trigger/*.sh

scripts to run before the main udev trigger is pulled

/lib/dracut/hooks/initqueue/*.sh

runs in parallel to the udev trigger

Udev events can add scripts here with /sbin/initqueue.

If /sbin/initqueue is called with the --onetime option, the script will be removed after it was run.

If /lib/dracut/hooks/initqueue/work is created and udev >= 143 then this loop can process the jobs in parallel to the udevtrigger.

If the udev queue is empty and no root device is found or no root filesystem was mounted, the user will be dropped to a shell after a timeout.

Scripts can remove themselves from the initqueue by rm $job.

/lib/dracut/hooks/pre-mount/*.sh

scripts to run before the root filesystem is mounted

Network filesystems like NFS that do not use device files are an exception. Root can be mounted already at this point.

/lib/dracut/hooks/mount/*.sh

scripts to mount the root filesystem

If the udev queue is empty and no root device is found or no root filesystem was mounted, the user will be dropped to a shell after a timeout.

/lib/dracut/hooks/pre-pivot/*.sh

scripts to run before latter initramfs cleanups

/lib/dracut/hooks/cleanup/*.sh

scripts to run before the real init is executed and the initramfs disappears

All processes started before should be killed here.