tuxlava 0.20.0

Python library and command line tool for generating LAVA jobs

Documentation - Repository - Issues

TuxLAVA is a command-line tool and Python library that simplifies creating LAVA job definitions for various device types. Developed by Linaro, it's part of TuxSuite and streamlines Linux kernel test automation.

• AVH
• FASTBOOT
• FVP
• NFS
• QEMU
• SSH

TuxLAVA is a part of TuxSuite, a suite of tools and services to help with Linux kernel development.

Table of Contents

• Installing TuxLAVA
• Examples
• Using TuxLAVA as a command line
• Using TuxLAVA as a library
• Contributing

Installing TuxLAVA

• From PyPI
• Debian packages
• RPM packages
• ArchLinux packages
• Run uninstalled

Examples

LAVA job to boot test a mipsel kernel at https://url/to/vmlinux:

tuxlava --device qemu-mips32el \
    --kernel https://url/to/vmlinux

Generate a LAVA job with ltp-smoke test:

tuxlava --device qemu-mips32el \
    --kernel https://url/to/vmlinux \
    --test ltp-smoke

Using TuxLAVA as a command line

Call tuxlava as follows:

tuxlava --device nfs-x86_64 \
    --kernel https://url/to/Image \
    --modules https://url/to/modules /usr/ \
    --rootfs https://url/to/rootfs \
    --tests boot

The --kernel, --modules, and --rootfs arguments can be URLs (e.g. https://...), file URLs (e.g. file:///...), or absolute file paths (e.g. /path/to/Image).

TuxLAVA will output the LAVA job to the stdout with the provided arguments for x86_64 device

The complete list of tuxlava options is available with the following command:

tuxlava --help

Using TuxLAVA as a library

TuxLAVA can be used as a python library as follows:

#!/usr/bin/env python

from tuxlava.jobs import Job

job = Job(
    device="nfs-x86_64",
    kernel="https://url/to/bzImage",
    rootfs="https://url/to/rootfs.tar.xz",
    tests=["ltp-smoke", "ltp-math"],
    modules="https://url/to/modules.tar.xz",
    parameters={"LAVA_JOB_PRIORITY": 50},
    timeouts={"deploy": 20},
)
job.initialize()
print(job.render())

Contributing

Contributions, bug reports and feature requests are welcome! Please see the issues or open a pull request.