linux.git/tools/testing/kunit, branch master

kunit: tool: copy caller args in run_kernel to prevent mutation

2026-03-02T17:02:06+00:00

run_kernel() appended KUnit flags directly to the caller-provided args
list. When exec_tests() calls run_kernel() repeatedly (e.g. with
--run_isolated), each call mutated the same list, causing later runs
to inherit stale filter_glob values and duplicate kunit.enable flags.

Fix this by copying args at the start of run_kernel(). Add a regression
test that calls run_kernel() twice with the same list and verifies the
original remains unchanged.

Fixes: ff9e09a3762f ("kunit: tool: support running each suite/test separately")
Signed-off-by: Shuvam Pandey 
Reviewed-by: David Gow 
Signed-off-by: Shuah Khan

kunit: add bash completion

2026-01-20T18:17:52+00:00

Currently, kunit.py has many subcommands and options, making it difficult
to remember them without checking the help message.

Add --list-cmds and --list-opts to kunit.py to get available commands and
options, use those outputs in kunit-completion.sh to show completion.

This implementation is similar to perf and tools/perf/perf-completion.sh.

Example output:
  $ source tools/testing/kunit/kunit-completion.sh
  $ ./tools/testing/kunit/kunit.py [TAB][TAB]
  build   config  exec    parse   run
  $ ./tools/testing/kunit/kunit.py run --k[TAB][TAB]
  --kconfig_add  --kernel_args  --kunitconfig

Link: https://lore.kernel.org/r/20260117-kunit-completion-v2-1-cabd127d0801@gmail.com
Reviewed-by: David Gow 
Signed-off-by: Ryota Sakamoto 
Signed-off-by: Shuah Khan

kunit: tool: test: Don't rely on implicit working directory change

2026-01-07T17:00:47+00:00

If no kunitconfig_paths are passed to LinuxSourceTree() it falls back to
DEFAULT_KUNITCONFIG_PATH. This resolution only works when the current
working directory is the root of the source tree. This works by chance
when running the full testsuite through the default unittest runner, as
some tests will change the current working directory as a side-effect of
'kunit.main()'. When running a single testcase or using pytest, which
resets the working directory for each test, this assumption breaks.

Explicitly specify an empty kunitconfig for the affected tests.

Link: https://lore.kernel.org/r/20260107015936.2316047-2-davidgow@google.com
Signed-off-by: Thomas Weißschuh 
Reviewed-by: David Gow 
Signed-off-by: David Gow 
Signed-off-by: Shuah Khan

kunit: tool: test: Rename test_data_path() to _test_data_path()

2026-01-07T17:00:41+00:00

Running the KUnit testsuite through pytest fails, as the function
test_data_path() is recognized as a test function. Its execution fails
as pytest tries to resolve the 'path' argument as a fixture which does
not exist.

Rename the function, so the helper function is not incorrectly
recognized as a test function.

Link: https://lore.kernel.org/r/20260107015936.2316047-1-davidgow@google.com
Signed-off-by: Thomas Weißschuh 
Reviewed-by: David Gow 
Signed-off-by: David Gow 
Signed-off-by: Shuah Khan

kunit: qemu_configs: Add 32-bit big endian ARM configuration

2026-01-05T22:40:32+00:00

Add a basic config to run kunit tests on 32-bit big endian ARM.

Link: https://lore.kernel.org/r/20260102-kunit-armeb-v1-1-e8e5475d735c@linutronix.de
Signed-off-by: Thomas Weißschuh 
Reviewed-by: David Gow 
Signed-off-by: Shuah Khan

kunit: tool: Don't overwrite test status based on subtest counts

2026-01-05T22:36:08+00:00

If a subtest itself reports success, but the outer testcase fails,
the whole testcase should be reported as a failure. However the status
is recalculated based on the test counts, overwriting the outer test
result. Synthesize a failed test in this case to make sure the failure
is not swallowed.

Link: https://lore.kernel.org/r/20251230-kunit-nested-failure-v1-2-98cfbeb87823@linutronix.de
Signed-off-by: Thomas Weißschuh 
Reviewed-by: David Gow 
Signed-off-by: Shuah Khan

kunit: tool: Add test for nested test result reporting

2026-01-05T22:35:49+00:00

Currently there is a lack of tests validating the result reporting from
nested tests. Add one, it will also be used to validate upcoming changes
to the nested test parsing.

Link: https://lore.kernel.org/r/20251230-kunit-nested-failure-v1-1-98cfbeb87823@linutronix.de
Signed-off-by: Thomas Weißschuh 
Reviewed-by: Rae Moar 
Reviewed-by: David Gow 
Signed-off-by: Shuah Khan

kunit: respect KBUILD_OUTPUT env variable by default

2026-01-05T22:35:22+00:00

Currently, kunit.py ignores the KBUILD_OUTPUT env variable and always
defaults to .kunit in the working directory. This behavior is inconsistent
with standard Kbuild behavior, where KBUILD_OUTPUT defines the build
artifact location.

This patch modifies kunit.py to respect KBUILD_OUTPUT if set.  A .kunit
subdirectory is created inside KBUILD_OUTPUT to avoid polluting the build
directory.

Link: https://lore.kernel.org/r/20260106-kunit-kbuild_output-v2-1-582281797343@gmail.com
Reviewed-by: David Gow 
Signed-off-by: Ryota Sakamoto 
Signed-off-by: Shuah Khan

kunit: qemu_configs: Add MIPS configurations

2025-09-15T16:30:07+00:00

Add basic support to run various MIPS variants via kunit_tool using the
virtualized malta platform.

Link: https://lore.kernel.org/r/20250908-kunit-mips-v5-1-d9f0632d1854@linutronix.de
Signed-off-by: Thomas Weißschuh 
Reviewed-by: David Gow 
Signed-off-by: Shuah Khan

kunit: Enable PCI on UML without triggering WARN()

2025-09-15T16:29:59+00:00

Various KUnit tests require PCI infrastructure to work. All normal
platforms enable PCI by default, but UML does not. Enabling PCI from
.kunitconfig files is problematic as it would not be portable. So in
commit 6fc3a8636a7b ("kunit: tool: Enable virtio/PCI by default on UML")
PCI was enabled by way of CONFIG_UML_PCI_OVER_VIRTIO=y. However
CONFIG_UML_PCI_OVER_VIRTIO requires additional configuration of
CONFIG_UML_PCI_OVER_VIRTIO_DEVICE_ID or will otherwise trigger a WARN() in
virtio_pcidev_init(). However there is no one correct value for
UML_PCI_OVER_VIRTIO_DEVICE_ID which could be used by default.

This warning is confusing when debugging test failures.

On the other hand, the functionality of CONFIG_UML_PCI_OVER_VIRTIO is not
used at all, given that it is completely non-functional as indicated by
the WARN() in question. Instead it is only used as a way to enable
CONFIG_UML_PCI which itself is not directly configurable.

Instead of going through CONFIG_UML_PCI_OVER_VIRTIO, introduce a custom
configuration option which enables CONFIG_UML_PCI without triggering
warnings or building dead code.

Link: https://lore.kernel.org/r/20250908-kunit-uml-pci-v2-1-d8eba5f73c9d@linutronix.de
Signed-off-by: Thomas Weißschuh 
Reviewed-by: Johannes Berg 
Reviewed-by: David Gow 
Signed-off-by: Shuah Khan