Skip to content

Latest commit

 

History

History
644 lines (474 loc) · 25.4 KB

kernels.md

File metadata and controls

644 lines (474 loc) · 25.4 KB
Raw

Linux kernels

LinuxKit kernel images are distributed as hub images which contain the kernel, kernel modules, kernel config file, and optionally, kernel headers to compile kernel modules against. The repository containing the official LinuxKit kernels is at linuxkit/kernels.

The LinuxKit kernels are based on the latest stable releases and are updated frequently to include bug and security fixes. For some kernels we do carry additional patches, which are mostly back-ported fixes from newer kernels. The full kernel source with patches can be found on github.

Kernel Image Naming and Tags

We publish the following kernel images:

  • primary kernel
  • debug kernel
  • tools for the specific kernel build - bcc and perf
  • builder image for the specific kernel build, useful for compiling compatible kernel modules

Primary Kernel Images

Each kernel image is tagged with:

  • the full kernel version, e.g. linuxkit/kernel:6.6.13. This is a multi-arch index, and should be used whenever possible.
  • the full kernel version plus hash of the files it was created from (git tree hash of the ./kernel directory), e.g. 6.6.13-c0d96951e9892a7447a8e7965d2d6bd7e621c3fd. This is a multi-arch index.
  • the full kernel version plus architecture, e.g. linuxkit/kernel:6.6.13-amd64 or linuxkit/kernel:6.6.13-arm64. Each of these is architecture specific.
  • the full kernel version plus hash of the files it was created from (git tree hash of the ./kernel directory) plus architecture, e.g. 6.6.13-c0d96951e9892a7447a8e7965d2d6bd7e621c3fd-arm64.

Debug Kernel Images

With each kernel image, we also publish kernels with additional debugging enabled. These have the same image name and the same tags as the primary kernel, with the -dbg suffix added immediately after the version. E.g.

  • linuxkit/kernel:6.6.13-dbg
  • linuxkit/kernel:6.6.13-dbg-c0d96951e9892a7447a8e7965d2d6bd7e621c3fd
  • linuxkit/kernel:6.6.13-dbg-amd64
  • linuxkit/kernel:6.6.13-dbg-c0d96951e9892a7447a8e7965d2d6bd7e621c3fd-amd64

Tools

With each kernel image, we also publish images with various tools. As of this writing, those tools are perf and bcc.

The tools images are named linuxkit/kernel-<tool>, followed by the same tags as the primary kernel. For example:

  • linuxkit/kernel-perf:6.6.13
  • linuxkit/kernel-perf:6.6.13-c0d96951e9892a7447a8e7965d2d6bd7e621c3fd
  • linuxkit/kernel-perf:6.6.13-amd64
  • linuxkit/kernel-perf:6.6.13-c0d96951e9892a7447a8e7965d2d6bd7e621c3fd-amd64

Additional Contributions

In addition to the official images, there are also some scripts which repackage kernels packages from some Linux distributions into LinuxKit kernel packages. These are mostly provided for testing purposes.

Note now linuxkit also embraces Preempt-RT Linux kernel to support more use cases for the promising IoT scenarios. All -rt patches are grabbed from https://www.kernel.org/pub/linux/kernel/projects/rt/. But so far we just enable it over 4.14.x.

Loading kernel modules

Most kernel modules are autoloaded with mdev but if you need to modprobe a module manually you can use the modprobe package in the onboot section like this:

  - name: modprobe
    image: linuxkit/modprobe:<hash>
    command: ["modprobe", "-a", "iscsi_tcp", "dm_multipath"]

Compiling external kernel modules

This section describes how to build external (out-of-tree) kernel modules. You need the following to build external modules. All of these are to be built for a specific version of the kernel. For the examples, we will assume 5.10.104; replace with your desired version.

  • source available to your modules - you need to get those on your own
  • kernel development headers - available in the linuxkit/kernel image as kernel-dev.tar, e.g. linuxkit/kernel:5.10.104
  • OS with sources and compiler - this must be the exact same version as that used to compile the kernel

As described above, the linuxkit/kernel images include kernel-dev.tar which contains the headers and other files required to compile kernel modules against the specific version of the kernel. Currently, the headers are not included in the initial RAM disk, but it is possible to compile custom modules offline and then include the modules in the initial RAM disk.

The source is available as the same name as the linuxkit/kernel image, with the addition of -builder on the tag. For example:

  • linuxkit/kernel:5.10.92 has builder linuxkit/kernel:5.10.92-builder
  • linuxkit/kernel:5.15.15 has builder linuxkit/kernel:5.15.15-builder

With the above in hand, you can create a multi-stage Dockerfile build to compile your modules. There is an example, but basically one can use a multi-stage build to compile the kernel modules:

FROM linuxkit/kernel:5.10.104 AS ksrc
FROM linuxkit/kernel:5.10.104-builder AS build

RUN apk add build-base

COPY --from=ksrc /kernel-dev.tar /
RUN tar xf kernel-dev.tar

# copy module source code and compile

To use the kernel module, we recommend adding a final stage to the Dockerfile above, which copies the kernel module from the build stage and performs a insmod as the entry point. You can add this package to the onboot section in your YAML file. test.yml contains an example for the configuration.

Builder Backups

As described above, the OS builder is referenced via <kernel-image>-builder, e.g. linuxkit/kernel:5.15.15-builder.

As a fallback, in case the -builder image is not available or you cannot access it from your development environment, you have 3 total places to determine the correct version of the OS image with sources and compiler:

  • -builder tag added to the kernel version, e.g. linuxkit/kernel:5.10.104-builder
  • labels on the kernel image, e.g. docker inspect linuxkit/kernel:5.10.104 | jq -r '.[].Config.Labels["org.mobyproject.linuxkit.kernel.buildimage"]'
  • /kernel-builder file in the kernel image

You should use -builder tag as the AS build in your Dockerfile, but you can use the direct source, extracted from the labels or /kernel-builder file in the kernel image, in the AS build.

For example, in the case of 5.10.104, the label and /kernel-builder file show linuxkit/alpine:2be490394653b7967c250e86fd42cef88de428ba, so you can use either linuxkit/alpine:2be490394653b7967c250e86fd42cef88de428ba or linuxkit/kernel:5.10.104-builder to build the modules.

Thus, the following are equivalent:

FROM linuxkit/kernel:5.10.104 AS ksrc
FROM linuxkit/kernel:5.10.104-builder AS build
FROM linuxkit/kernel:5.10.104 AS ksrc
FROM linuxkit/alpine:2be490394653b7967c250e86fd42cef88de428ba AS build

Building and Modifying

This section describes how to build kernels, and how to modify existing ones.

Throughout the document, the terms used are:

  • kernel version: actual semver version of a kernel, e.g. 6.6.13 or 5.15.27
  • kernel series: major.minor version of a kernel, e.g. 6.6.x or 5.15.x

Throughout this document, the architecture used is the kernel-recognized one, available on most systems as uname -m, e.g. aarch64 or x86_64. You may be familiar with the alpine or golang one, e.g. amd64 or amd64, which are not used here.

Note: After changing and committing any changes to the kernel directory or any subdirectories, you must update tests, examples and other dependencies. This is done via:

make update-kernel-yamls

Each series of kernels has a dedicated directory in ../kernel/, e.g. 6.6.x or 5.15.x. Variants, like rt kernels, have their own directory as well, e.g. 5.11.x-rt. However, for variants, the patches from both the common kernel, e.g. 5.11.x, and the variant, e.g. 5.11.x-rt, are applied, and the configs from both are combined.

Within the series-dedicated directory, there are:

  • kernel config file for each architecture named config-<arch>, e.g. 6.6.13/config-x86_64, one per target architecture.
  • optional patches directory, e.g. 6.6.13/patches, which contains patches to apply to the kernel source

The config file and patches are applied during the kernel build process.

Note: We try to keep the differences between kernel versions and architectures to a minimum, so if you make changes to one configuration also try to apply it to the others. The script kconfig-split.py can be used to compare kernel config files. For example:

../scripts/kconfig-split.py 5.15.x/config-aarch64 5.15.x/config-x86_64

creates a file with the common and the x86_64 and arm64 specific config options for the 5.15.x kernel series.

Note: The CI pipeline does not push out kernel images. Anyone modifying a kernel should:

  1. Follow the steps below for the desired changes and commit them.
  2. Run appropriate make build or variants to ensure that it works.
  3. Open a PR with the changes. This may fail, as the CI pipeline may not have access to the modified kernels.
  4. A maintainer should run make push to push out the images.
  5. Run (or rerun) the tests.

Build options

The targets and variants for building are as follows:

  • make build - make all kernels in the version list and their variants
  • make build-<version> - make all variants of a specific kernel version
  • make buildkernel-<version> - make all variants of a specific kernel version
  • make buildplainkernel-<version> - make just the provided version's kernel
  • make builddebugkernel-<version> - make just the provided version's debug kernel
  • make buildtools-<version> - make just the provided version's tools

To push:

  • make push - push all kernels in the version list and their variants
  • make push-<version> - push all variants of a specific kernel version

Finally, for convenience:

  • make list - list all kernels in the version list

By default, it builds for all supported architectures. To build just for a specific architecture:

make build ARCH=amd64

The variable ARCH should use the golang variants only, i.e. amd64 and arm64.

To build for multiple architectures, call it multiple times:

make build ARCH=amd64
make build ARCH=arm64

When building for a specific architecture, the build process will use your local Docker, passing it --platforms for the architecture. If you have a builder on a different architecture, e.g. you are running on an Apple Silicon Mac (arm64) and want to build for x86_64 without emulating (which can be very slow), you can use the BUILDER variable:

make build ARCH=x86_64 BUILDER=remote-amd64-builder

Builder also supports a builder pattern. If BUILDER contains the string {{.Arch}}, it will be replaced with the architecture being built.

For example:

make build ARCH=x86_64 BUILDER=remote-{{.Arch}}-builder
make build ARCH=aarch64 BUILDER=remote-{{.Arch}}-builder

will build x86_64 on remote-amd64-builder and aarch64 on remote-arm64-builder.

Finally, if no BUILDER is specified, the build will look for a builder named linuxkit-linux-{{.Arch}}-builder, e.g. linuxkit-linux-amd64-builder or linuxkit-linux-arm64-builder. If that builder does not exist, it will fall back to your local Docker setup.

Modifying the kernel config

The process of modifying the kernel configuration is as follows:

  1. Create a linuxkit/kconfig container image: make kconfig. This is not pushed out.
  2. Run a container based on linuxkit/kconfig.
  3. In the container, modify the config to suit your needs using normal kernel tools like make defconfig or make menuconfig.
  4. Save the config from the image.

The linuxkit/kconfig image contains the patched sources for all support kernels and architectures in /linux-<major>.<minor>.<rev>. The kernel source also has the kernel config copied to the default kernel config location, so that make menuconfig and make defconfig work correctly.

Run the container as follows:

docker run --rm -ti -v $(pwd):/src linuxkit/kconfig

This will give you a interactive shell where you can modify the kernel configuration you want, while mounting the directory, so that you can save the modified config.

To create or modify the config, you must cd to the correct directory, e.g.

cd /linux-6.6.13
# or
cd /linux-5.15.27

Now you can build the config.

When make defconfig or make menuconfig is done, the modified config file will be in .config; save the file back to /src, e.g.

cp .config /src/6.6.x/config-x86_64

You can also configure other architectures other than the native one. For example to configure the arm64 kernel on x86_64, use:

make ARCH=arm64 defconfig
make ARCH=arm64 oldconfig # or menuconfig

Note that the generated file must be final. When you actually build the kernel, it will check that running make defconfig will have no changes. If there are changes, the build will fail.

The easiest way to check it is to rerun make defconfig inside the kconfig container.

  1. Finish your creation of the config file, as above.
  2. Copy the .config file to the target location, as above.
  3. Copy the .config file to the source location for defconfig, e.g. cp .config arch/x86/configs/x86_64_config or cp. config /linux/arch/arm64/configs/defconfig
  4. Run make defconfig again, and check that there are no changes, e.g. diff .config arch/x86/configs/x86_64_config or diff .config /linux/arch/arm64/configs/defconfig

If there are no differences, then you can commit the new config file.

Finally, test that you can build the kernel with that config as make build-<version>, e.g. make build-5.15.148.

Adding a new kernel version

If you want to add a new kernel version within an existing series, e.g. 5.15.27 already exists and you want to add (or replace it with) 5.15.148, apply the following process.

  1. Determine the series, i.e. the kernel major.minor version, followed by x. E.g. for 5.15.148, the series is 5.15.x.
  2. Modify the KERNEL_VERSION in the build-args file in the series directory to the new version. E.g. 5.15.x/build-args.
  3. Create a new linuxkit/kconfig container image: make kconfig. This is not pushed out.
  4. Run a container based on linuxkit/kconfig.
docker run --rm -ti -v $(pwd):/src linuxkit/kconfig
  1. In the container, change directory to the kernel source directory for the new version, e.g. cd /linux-5.15.148.
  2. Run make defconfig to create the default config file.
  3. If the config file has changed, copy it out of the container and check it in, e.g. cp .config /src/5.15.x/config-x86_64.
  4. Repeat for other architectures.
  5. Commit the changed config files.
  6. Test that you can build the kernel with that config as make build-<version>, e.g. make build-5.15.148.

Adding a new kernel series

To add a new kernel series, you need to:

  1. Create new directory for the series, e.g. 6.7.x
  2. Create config files for each architecture in that directory
  3. Optionally, create a patches/ subdirectory in that directory with any patches to add
  4. Create a build-args file in that directory with at least the following settings:
KERNEL_VERSION=<version>
KERNEL_SERIES=<series>
BUILD_IMAGE=linuxkit/alpine:<builder>

Since the last major series likely is the best basis for the new one, subject to additional modifications, you can use the previous one as a starting point.

  1. Make the directory for the new series, e.g. mkdir 7.0.x
  2. Create a new linuxkit/kconfig container image: make kconfig. This is not pushed out.
  3. Run a container based on linuxkit/kconfig.
docker run --rm -ti -v $(pwd):/src linuxkit/kconfig
  1. In the container, change directory to the kernel source directory for the new version, e.g. cd /linux-7.0.5.
  2. Copy the existing config file for the previous series, e.g. cp /src/6.6.x/config-x86_64 .config.
  3. Run make oldconfig to create the config file for the new series from the old one. Answer any questions.
  4. Save the newly generated config file .config to the source directory, e.g. cp .config /src/7.0.x/config-x86_64.
  5. Repeat for other architectures.
  6. Commit the new config files.
  7. Test that you can build the kernel with that config as make build-<version>, e.g. make build-7.0.5.

In addition, there are tests that are applied to a specific kernel version, notably the tests in 020_kernel. You will need to add a new test case for the new series, copying an existing one and modifying it as needed.

Building and using custom kernels

To build and test locally modified kernels, e.g., to try a different kernel config or new patches, the existing kernel build system in the kernel directory can be re-used. For example, assuming the current 4.9 kernel is 4.9.33, you can build a local kernel with:

make build_4.9.x

This will create a local kernel image called linuxkit/kernel:4.9.33-<hash>-dirty assuming you haven't committed you local changes. You can then use this in your YAML file as:

kernel:
  image: linuxkit/kernel:4.9.33-<hash>-dirty

If you have committed your local changes, the -dirty will not be appended. Then you can also override the Hub organisation to use the image elsewhere with (and also disable image signing):

make ORG=<your hub org>

The image will be uploaded to Hub and can be use in a YAML file as <your hub org>/kernel:4.9.33 or as <your hub org>/kernel:4.9.33-<hash>.

The kernel build system has some provision to allow local customisation to the build.

If you want to override/add some kernel config options, you can add a file called config-4.9.x-x86_64-foo and then invoke the build with make EXTRA=-foo build_4.9.x-foo and this will build an image with the additional kernel config options enabled.

If you want additional patches being applied, just copy them to the patches-4.X.x and the build process will pick them up.

Working with Linux kernel patches for LinuxKit

We may apply patches to the Linux kernel used in LinuxKit, primarily to cherry-pick some upstream patches or to add some additional functionality, not yet accepted upstream.

Patches are located in kernel/patches-<kernel version> and should follow these rules:

  • Patches must be in git am format, i.e. they should contain a complete and sensible commit message.
  • Patches must contain a Developer's Certificate of Origin.
  • Patch files must have a numeric prefix to ensure the ordering in which they are applied.
  • If patches are cherry-picked, they must be cherry-picked with -x to contain the original commit ID.
  • If patches are from a different git tree (other than the stable tree), or from a mailing list posting they should contain an Origin: line with a link to the source.

This document outlines the recommended procedure to handle patches. The general process is to apply them to a branch of the Linux stable tree and then export them with git format-patch.

If you want to add or remove patches currently used, please also ping @rneugeba on the PR so that we can update our internal Linux tree to ensure that patches are carried forward if we update the kernel in the future.

Preparation

Patches are applied to point releases of the linux stable tree. You need an up-to-date copy of that tree:

git clone git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable.git

Add it as a remote to a clone of the LinuxKit clone.

We use the following variables:

  • KITSRC: Base directory of LinuxKit repository
  • LINUXSRC: Base directory of Linux stable kernel repository e.g.:
KITSRC=~/src/linuxkit/linuxkit
LINUXSRC=~/src/linuxkit/linux

to refer to the location of the LinuxKit and Linux kernel trees.

Updating the patches to a new kernel version

There are different ways to do this, but we recommend applying the patches to the current version and then rebase to the new version. We define the following variables to refer to the current base tag and the new tag you want to rebase the patches to:

CURTAG=v4.9.14
NEWTAG=v4.9.15

If you don't already have a branch, it's best to import the current patch set and then rebase:

cd $LINUXSRC
git checkout -b ${NEWTAG}-linuxkit ${CURTAG}
git am ${KITSRC}/kernel/patches/*.patch
git rebase ${NEWTAG}-linuxkit ${NEWTAG}

The git am should not have any conflicts and if the rebase has conflicts resolve them, then git add <files> and git rebase --continue.

If you already have linux tree with a ${CURTAG}-linuxkit branch, you can rebase by creating a new branch from the current branch and then rebase:

cd $LINUXSRC
git checkout ${CURTAG}-linuxkit
git branch ${NEWTAG}-linuxkit ${CURTAG}-linuxkit
git rebase --onto ${NEWTAG} ${NEWTAG} ${NEWTAG}-linuxkit

Again, resolve any conflicts as described above.

Adding/Removing patches

If you want to add or remove patches make sure you have an up-to-date branch with the currently applied patches (see above). Then either any normal means (git cherry-pick -x, git am, or git commit, etc) to add new patches. For cherry-picked patches also please add a Origin: line after the DCO lines with a reference the git tree the patch was cherry-picked from.

If the patch is not cherry-picked try to include as much information in the commit message as possible as to where the patch originated from. The canonical form would be to add a Origin: line after the DCO lines, e.g.:

Origin: https://patchwork.ozlabs.org/patch/622404/

Export patches to LinuxKit

To export patches to LinuxKit, you should use git format-patch from the Linux tree, e.g., something along these lines:

cd $LINUXSRC
rm $KITSRC/kernel/patches-4.9.x/*
git format-patch -o $KITSRC/kernel/patches-4.9.x v4.9.15..HEAD

Then, create a PR for LinuxKit.

Using perf

The kernel-perf package contains a statically linked perf binary under /usr/bin which is matched with the kernel of the same tag. The simplest way to use the perf utility is to add the package to the init section in the YAML file. This adds the binary to the root filesystem.

To use the binary, you can either bind mount it into the getty or ssh service container or you can access the root filesystem from the getty container via nsenter:

nsenter -m/proc/1/ns/mnt ash

Alternatively, you can add the kernel-perf package as stage in a multi-stage build to add it to a custom package.

ZFS

The kernel build Makefile has support for building the ZFS kernel modules. Note, the modules are currently not distributed as standard LinuxKit packages and if you wish to use them you have to compile them yourself:

cd kernel
make ORG=<foo> push_zfs_4.9.x # or different kernel version

will build and push a zfs-kmod-4.9.<version> image to Docker Hub under the ORG specified. This package contains the all the standard kernel modules from the kernel specified plus the spl and zfs kernel modules, with depmod run over them, so they can be modprobeed. To use the modules do something like this in your YAML file:

kernel:
  image: linuxkit/kernel:4.9.<version>
  cmdline: "console=tty0 console=ttyS0 console=ttyAMA0"
init:
  - <foo>/zfs-kmod:4.9.<version>
  ...

Then, you also need to make sure the Alpine zfs utilities are available in the container where your want to run zfs commands. The Alpine zfs utilities are available in linuxkit/alpine and the version of the kernel module should match the version of the tools. The container where you run the zfs tools might also need CAP_SYS_MODULE to be able to load the kernel modules.

Kernels in examples and tests

All of the linuxkit .yml files use the images from linuxkit/kernel:<tag>.

When updating the kernel, you run commands to update the tests. The updates to any file that contains references to linuxkit/kernel in this repository work as follows:

  • Semver tags are replaced by the most recent kernel version. For example, linuxkit/kernel:5.10.104 will become 6.6.13 when available, and then 6.6.15, and then 7.0.1, etc. The highest semver always is used.
  • Semver+hash tags are replaced by the most recent hash and patch version for that series. For example, linuxkit/kernel:5.10.104-abcdef1234 will become 5.10.104-aaaa54232 (same semver, newer hash), and then 5.10.105-bbbb12345 (newer semver, newer hash), etc. The highest semver+hash always is used.

This is not an inherent characteristic of linuxkit tool, which never will change your .yml files. It is part of the update process for yml files in this repository.

The net of the above is the following rule:

  • If you want a reference to a specific kernel series, e.g. a test or example that works only with 5.10.x, then use a specific hash, e.g. linuxkit/kernel:5.10.104-abcdef1234. The hash and patch version will update, but not more. The most common use case for this is kernel version-specific tests.
  • If you want a reference to the most recent kernel, whatever version it is, then use a semver tag, e.g. linuxkit/kernel:6.6.13. The most common use case for this is examples that work with any kernel version, which is the vast majority of cases.

You can get the current hash by executing the following:

$ cd kernel
$ make tag-plain-kernel-<version>
# for example:
$ make tag-plain-kernel-6.6.13
linuxkit/kernel:6.6.13-3a8b3faf92390265b1fbee792b9a3fe14d14c26e