[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20210614064205.GA29220@lst.de>
Date: Mon, 14 Jun 2021 08:42:05 +0200
From: Christoph Hellwig <hch@....de>
To: André Almeida <andrealmeid@...labora.com>
Cc: Christoph Hellwig <hch@....de>,
Andy Shevchenko <andriy.shevchenko@...ux.intel.com>,
linux-kernel@...r.kernel.org,
Brendan Higgins <brendanhiggins@...gle.com>,
linux-kselftest@...r.kernel.org, kunit-dev@...glegroups.com,
Shuah Khan <shuah@...nel.org>, ~lkcamp/patches@...ts.sr.ht,
nfraprado@...labora.com, leandro.ribeiro@...labora.com,
Vitor Massaru Iha <vitor@...saru.org>, lucmaga@...il.com,
David Gow <davidgow@...gle.com>,
Daniel Latypov <dlatypov@...gle.com>, tales.aparecida@...il.com
Subject: Re: [PATCH v3 1/1] lib: Convert UUID runtime test to KUnit
> +config UUID_KUNIT_TEST
> + tristate "Unit test for UUID" if !KUNIT_ALL_TESTS
> + depends on KUNIT
> + default KUNIT_ALL_TESTS
> + help
> + This builds the UUID unit test.
Does this first help line really add any value if we have this second
line:
> + Tests parsing functions for UUID/GUID strings.
?
> + If unsure, say N.
Not specific to this case, but IMHO we can drop this line for all kunit
tests as it is completely obvious.
> @@ -354,5 +353,6 @@ obj-$(CONFIG_LIST_KUNIT_TEST) += list-test.o
> obj-$(CONFIG_LINEAR_RANGES_TEST) += test_linear_ranges.o
> obj-$(CONFIG_BITS_TEST) += test_bits.o
> obj-$(CONFIG_CMDLINE_KUNIT_TEST) += cmdline_kunit.o
> +obj-$(CONFIG_UUID_KUNIT_TEST) += test_uuid.o
Another meta-comment on the kunit tests: Wouldn't it make more sense
to name them all as CONFIG_KUNIT_TEST_FOO to allow for easier grepping?
> -struct test_uuid_data {
> +struct test_data {
> const char *uuid;
> guid_t le;
> uuid_t be;
> };
>
> -static const struct test_uuid_data test_uuid_test_data[] = {
> +static const struct test_data correct_data[] = {
What is the reason for these renames? Is this a pattern used for
other kunit tests?
> +static void uuid_correct_le(struct kunit *test)
> {
> + guid_t le;
> + const struct test_data *data = (const struct test_data *)(test->param_value);
Overly long line. But as far as I can tell there is no need for the
case that causes this mess anyway given that param_value is a
"const void *".
Same for all the other instances of this.
> +static void uuid_wrong_le(struct kunit *test)
> {
> guid_t le;
> + const char **data = (const char **)(test->param_value);
No need for the second pair of braces. Same for various other instances.
Powered by blists - more mailing lists