On Tue, Jun 25, 2024 at 10:17:58PM +0100, Matthew Wilcox (Oracle) wrote:
> This is not a very sophisticated test suite yet, but it helped find
> a few bugs and provides a framework for adding more tests as more
> bugs are found.
>
> Signed-off-by: Matthew Wilcox (Oracle) <willy@xxxxxxxxxxxxx>
> ---
> lib/Kconfig.debug | 3 +
> lib/Makefile | 1 +
> lib/test_rosebush.c | 140 ++++++++++++++++++++++++++++++++++++++++++++
> 3 files changed, 144 insertions(+)
> create mode 100644 lib/test_rosebush.c
>
> diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug
> index 59b6765d86b8..f3cfd79d8dbd 100644
> --- a/lib/Kconfig.debug
> +++ b/lib/Kconfig.debug
> @@ -2447,6 +2447,9 @@ config TEST_RHASHTABLE
>
> If unsure, say N.
>
> +config TEST_ROSEBUSH
> + tristate "Test the Rosebush data structure"
> +
This needs a `depends on KUNIT`.
And compiling this test as module results in rbh_destroy not found. I
think you missed EXPORT_SYMBOL of rbh_destroy in the previous patch.
> config TEST_IDA
> tristate "Perform selftest on IDA functions"
>
> +
> +static void check_empty_rbh(struct kunit *test, struct rbh *rbh)
> +{
> + iter_rbh(test, rbh, 0, NULL);
> + iter_rbh(test, rbh, 1, NULL);
> + iter_rbh(test, rbh, 17, NULL);
> + iter_rbh(test, rbh, 42, NULL);
> +}
Do these hashes hold any significance that you test them often after an
insert?
--
Pankaj
