On Fri, Nov 2, 2018 at 11:44 AM Shuah Khan <shuah@xxxxxxxxxx> wrote:
>
> On 10/23/2018 05:57 PM, Brendan Higgins wrote:
<snip>
> > + * Example:
> > + *
> > + * .. code-block:: c
> > + *
> > + * void add_test_basic(struct test *test)
> > + * {
> > + * TEST_EXPECT_EQ(test, 1, add(1, 0));
> > + * TEST_EXPECT_EQ(test, 2, add(1, 1));
> > + * TEST_EXPECT_EQ(test, 0, add(-1, 1));
> > + * TEST_EXPECT_EQ(test, INT_MAX, add(0, INT_MAX));
> > + * TEST_EXPECT_EQ(test, -1, add(INT_MAX, INT_MIN));
> > + * }
> > + *
> > + * static struct test_case example_test_cases[] = {
> > + * TEST_CASE(add_test_basic),
> > + * {},
> > + * };
> > + *
> > + */
> > +struct test_case {
> > + void (*run_case)(struct test *test);
> > + const char name[256];
> > +
> > + /* private: internal use only. */
> > + bool success;
> > +};
> > +
>
> Introducing a prefix kunit_* might be a good idea for the API.
> This comment applies to the rest of patches as well.
What about kunit_* instead of test_* and kmock_* instead of mock_*?
Does that seem reasonable?
