# Re: \[PATCH\] lib: use of kunit in test\_parman.c

- From: David Gow \<hidden\>
- Date: 2021-07-28 01:37:54 
- List: netdev
- Page: https://ratatoskr.run/netdev/2021/07/1460273
- Thread: https://ratatoskr.run/netdev/2021/07/1460185/t.md

```text
On Wed, Jul 28, 2021 at 7:04 AM José Aquiles Guedes de Rezende
[ref] wrote:
>
> Convert the parman test module to use the KUnit test framework.
> This makes thetest clearer by leveraging KUnit's assertion macros
> and test case definitions,as well as helps standardize on a testing framework.
>
> Co-developed-by: Matheus Henrique de Souza Silva <redacted>
> Signed-off-by: Matheus Henrique de Souza Silva <redacted>
> Signed-off-by: José Aquiles Guedes de Rezende <redacted>
> ---

Thanks for porting this! A few more notes from the KUnit side.

>  lib/test_parman.c | 145 +++++++++++++++++++---------------------------
>  1 file changed, 60 insertions(+), 85 deletions(-)

This seems to be missing changes to lib/Kconfig.debug: you'll want to
modify the TEST_PARMAN config item to
- depend on KUNIT
- only appear if !KUNIT_ALL_TESTS
- default KUNIT_ALL_TESTS

It might also be nice to:
- select PARMAN (it's otherwise extremely unlikely a config will
actually pick up the test)
- maybe rename TEST_PARMAN to PARMAN_KUNIT_TEST (this is optional,
since this is a port of an existing test, but does make it clearer, so
it really depends on what existing usage looks like)

>
> diff --git a/lib/test_parman.c b/lib/test_parman.c
> index 35e32243693c..bd5010f0a412 100644
> --- a/lib/test_parman.c
> +++ b/lib/test_parman.c

The rest of this looks okay to me, but you should check out Daniel's
comments in his email, too.

Cheers,
-- David
```
