qemu-user-static (qus) is a compilation of utilities, examples and references to build and execute OCI images (aka docker images) for foreign architectures using QEMU's user-mode emulation.
- Ready-to-use docker images are provided for each of seven host architectures officially supported by Docker, Inc. or built by official images (docker-library/official-images: Architectures other than amd64?):
amd64
,i386
,arm64v8
,arm32v7
,arm32v6
,s390x
andppc64le
. - Multiple minimal working setups to build and execute
arm64v8
containers onamd64
are configured and tested on GitHub Actions. The full flow is public, for other users to learn and adapt these resources to their needs. - These resources are tested on GNU/Linux and Windows 10 (Docker Desktop). Contributions to test them on other host OSs are welcome!
Find further details at dbhi.github.io/qus.
NOTE: Although
docker
is used in these examples, users have reported that other engines such as podman can also be used. See also kata-containers/runtime#1280.
Run the Action without arguments for registering all the supported interpreters:
- uses: dbhi/qus/action@main
Optionally, provide a space-separated list of target architectures:
- uses: dbhi/qus/action@main
with:
targets: arm aarch64
Then, execute foreign binaries and/or containers straightaway!
NOTE: yaml-multiline.info
The recommended approach is to run the following container:
docker run --rm --privileged aptman/qus -s -- -p [TARGET_ARCH]
NOTE: since
aptman/qus
is a manifest, this command works onamd64
,arm64v8
,arm32v7
,arm32v6
,i386
,s390x
orppc64le
hosts.
The required qemu-*-static
binaries (which are all included in the image) will be loaded and registered. The container will then exit. From there on, images and/or binaries for foreign architectures can be executed.
Optional argument TARGET_ARCH
is the target architecture to be translated through QEMU. If it is omitted, all available targets will be registered and loaded. The supported values are the following:
i386 i486 alpha arm armeb sparc32plus ppc ppc64 ppc64le m68k mips mipsel mipsn32 mipsn32el mips64 mips64el sh4 sh4eb s390x aarch64 aarch64_be hppa riscv32 riscv64 xtensa xtensaeb microblaze microblazeel or1k x86_64
NOTE: sudo privileges on the host are required in order to register
binfmt
formats. On GNU/Linux, it is possible to executeregister.sh
directly. On Windows, a container must be used, so that changes are applied to the underlying VM, since no kernel is available on the host (i.e., from the test list, onlyC
,V
,I
orD
will work on Windows).
In order to unset the registered formats, and unload the binaries, run:
docker run --rm --privileged aptman/qus -- -r
# docker run --rm --privileged aptman/qus -h
Usage: register.sh [--help][--interactive][--list][--static][-- ARGS]
Wrapper around qemu-binfmt-conf.sh, to configure binfmt_misc to use qemu interpreter
-h|--help|-help:
display this usage
-i|--interactive|-interactive:
execute all the remaining args with 'sh -c', then exit
-l|--list|-list:
list currently registered interpreters
-s|--static|-static:
add '--qemu-suffix -static' to ARGS
-- ARGS:
arguments for qemu-binfmt-conf.sh
To register a single static target persistently, use e.g.:
register.sh -s -- -p aarch64
To remove all register interpreters and exit, use:
register.sh -- -r
In contexts such as CI pipelines it might be desirable to reduce the required bandwidth. Hence, instead of using aptman/qus
images —which include all the binaries for all the supported target architectures—, individual tarballs are available through GitHub Releases. These can be used along with aptman/qus:register
images or with register.sh
(without an OCI runtime). See either f
, F
, c
, C
, v
or V
in Tests for examples of these use cases.