[PATCH v1] kunit: add tests for using current KUnit test field

[Rae Moar]

Create test suite called "kunit_current" to add test coverage for the use
of current->kunit_test, which returns the current KUnit test.

Add three test cases:
- kunit_current_kunit_test_field to test the use of current->kunit_test.

- kunit_current_get_current_test to test the method
kunit_get_current_test(), which utilizes current->kunit_test.

- kunit_current_fail_current_test to test the method
kunit_fail_current_test(), which utilizes current->kunit_test.

Signed-off-by: Rae Moar <rm...@google.com>
---
lib/kunit/kunit-test.c | 61 +++++++++++++++++++++++++++++++++++++++++-
1 file changed, 60 insertions(+), 1 deletion(-)

diff --git a/lib/kunit/kunit-test.c b/lib/kunit/kunit-test.c
index b63595d3e241..91984b92c916 100644
--- a/lib/kunit/kunit-test.c
+++ b/lib/kunit/kunit-test.c
@@ -6,6 +6,7 @@
* Author: Brendan Higgins <brendan...@google.com>
*/
#include <kunit/test.h>
+#include <kunit/test-bug.h>

#include "try-catch-impl.h"

@@ -532,7 +533,65 @@ static struct kunit_suite kunit_status_test_suite = {
.test_cases = kunit_status_test_cases,
};

+static void kunit_current_kunit_test_field(struct kunit *test)
+{
+ struct kunit *current_test;
+
+ /* Check to ensure the result of current->kunit_test
+ * is equivalent to current test.
+ */
+ current_test = current->kunit_test;
+ KUNIT_EXPECT_PTR_EQ(test, test, current_test);
+}
+
+static void kunit_current_get_current_test(struct kunit *test)
+{
+ struct kunit *current_test1, *current_test2;
+
+ /* Check to ensure the result of kunit_get_current_test()
+ * is equivalent to current test.
+ */
+ current_test1 = kunit_get_current_test();
+ KUNIT_EXPECT_PTR_EQ(test, test, current_test1);
+
+ /* Check to ensure the result of kunit_get_current_test()
+ * is equivalent to current->kunit_test.
+ */
+ current_test2 = current->kunit_test;
+ KUNIT_EXPECT_PTR_EQ(test, current_test1, current_test2);
+}
+
+static void kunit_current_fail_current_test(struct kunit *test)
+{
+ struct kunit fake;
+
+ /* Initialize fake test and set as current->kunit_test. */
+ kunit_init_test(&fake, "fake test", NULL);
+ KUNIT_EXPECT_EQ(test, fake.status, KUNIT_SUCCESS);
+ current->kunit_test = &fake;
+
+ /* Fail current test and expect status of fake test to be failed. */
+ kunit_fail_current_test("This test is supposed to fail.");
+ KUNIT_EXPECT_EQ(test, fake.status, (enum kunit_status)KUNIT_FAILURE);
+
+ /* Reset current->kunit_test to current test. */
+ current->kunit_test = test;
+}
+
+static struct kunit_case kunit_current_test_cases[] = {
+ KUNIT_CASE(kunit_current_kunit_test_field),
+ KUNIT_CASE(kunit_current_get_current_test),
+ KUNIT_CASE(kunit_current_fail_current_test),
+ {}
+};
+
+static struct kunit_suite kunit_current_test_suite = {
+ .name = "kunit_current",
+ .test_cases = kunit_current_test_cases,
+};
+
kunit_test_suites(&kunit_try_catch_test_suite, &kunit_resource_test_suite,
- &kunit_log_test_suite, &kunit_status_test_suite);
+ &kunit_log_test_suite, &kunit_status_test_suite,
+ &kunit_current_test_suite);

MODULE_LICENSE("GPL v2");

base-commit: 7232282dd47cce6a780c9414bd9baccf232c7686
--
2.40.0.348.gf938b09366-goog

[Daniel Latypov]

I've got a few minor comments below, but this otherwise looks good.
I like the idea of testing knuit_fail_current_test().

On Thu, Mar 30, 2023 at 3:05 PM Rae Moar <rm...@google.com> wrote:
>
> +static void kunit_current_kunit_test_field(struct kunit *test)
> +{
> + struct kunit *current_test;
> +
> + /* Check to ensure the result of current->kunit_test
> + * is equivalent to current test.
> + */
> + current_test = current->kunit_test;
> + KUNIT_EXPECT_PTR_EQ(test, test, current_test);

Perhaps we can combine this and the next test case down to
static void kunit_current_test(struct kunit *test) {
/* There are two different ways of getting the current test */
KUNIT_EXPECT_PTR_EQ(test, test, current->kunit_test);
KUNIT_EXPECT_PTR_EQ(test, test, kunit_get_current_test());
}
?

> +}
> +
> +static void kunit_current_get_current_test(struct kunit *test)
> +{
> + struct kunit *current_test1, *current_test2;
> +
> + /* Check to ensure the result of kunit_get_current_test()
> + * is equivalent to current test.
> + */
> + current_test1 = kunit_get_current_test();
> + KUNIT_EXPECT_PTR_EQ(test, test, current_test1);
> +
> + /* Check to ensure the result of kunit_get_current_test()
> + * is equivalent to current->kunit_test.
> + */
> + current_test2 = current->kunit_test;
> + KUNIT_EXPECT_PTR_EQ(test, current_test1, current_test2);

> +}
> +
> +static void kunit_current_fail_current_test(struct kunit *test)
> +{
> + struct kunit fake;
> +
> + /* Initialize fake test and set as current->kunit_test. */

Nit: I think the code is self-explanatory enough that we can drop this comment.

> + kunit_init_test(&fake, "fake test", NULL);
> + KUNIT_EXPECT_EQ(test, fake.status, KUNIT_SUCCESS);
> + current->kunit_test = &fake;
> +
> + /* Fail current test and expect status of fake test to be failed. */

Nit: I think this comment could also be dropped or maybe shortened to
kunit_fail_current_test("This should make `fake` fail");

or
/* Now kunit_fail_current_test() should modify `fake`, not `test` */
kunit_fail_current_test("This should make `fake` fail");

> + kunit_fail_current_test("This test is supposed to fail.");
> + KUNIT_EXPECT_EQ(test, fake.status, (enum kunit_status)KUNIT_FAILURE);
> +

Hmm, should we try calling
kunit_cleanup(&fake);
?

Right now this does resource cleanups, but we might have other state
to cleanup for our `fake` test object in the future.

Daniel

[David Gow]

On Fri, 31 Mar 2023 at 06:21, 'Daniel Latypov' via KUnit Development
<kuni...@googlegroups.com> wrote:
>
> I've got a few minor comments below, but this otherwise looks good.
> I like the idea of testing knuit_fail_current_test().
>
>
> On Thu, Mar 30, 2023 at 3:05 PM Rae Moar <rm...@google.com> wrote:
> >
> > +static void kunit_current_kunit_test_field(struct kunit *test)
> > +{
> > + struct kunit *current_test;
> > +
> > + /* Check to ensure the result of current->kunit_test
> > + * is equivalent to current test.
> > + */
> > + current_test = current->kunit_test;
> > + KUNIT_EXPECT_PTR_EQ(test, test, current_test);
>
> Perhaps we can combine this and the next test case down to
> static void kunit_current_test(struct kunit *test) {
> /* There are two different ways of getting the current test */
> KUNIT_EXPECT_PTR_EQ(test, test, current->kunit_test);
> KUNIT_EXPECT_PTR_EQ(test, test, kunit_get_current_test());
> }
> ?
>

Agreed: checking current->kunit_test twice feels a bit odd.

I could go either way here. We currently don't do this with the other
status tests (kunit_status), only with the resource ones.
But it doesn't hurt to add it...

>
> Daniel
>
> --
> You received this message because you are subscribed to the Google Groups "KUnit Development" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to kunit-dev+...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/kunit-dev/CAGS_qxqNwVcymkG6-8Kv72oZc9aDqjFjBBmjr%2Bf%2BmOVKT1bGvA%40mail.gmail.com.

[Rae Moar]

On Thu, Mar 30, 2023 at 6:21 PM 'Daniel Latypov' via KUnit Development
<kuni...@googlegroups.com> wrote:
>
> I've got a few minor comments below, but this otherwise looks good.
> I like the idea of testing knuit_fail_current_test().
>
>
> On Thu, Mar 30, 2023 at 3:05 PM Rae Moar <rm...@google.com> wrote:
> >
> > +static void kunit_current_kunit_test_field(struct kunit *test)
> > +{
> > + struct kunit *current_test;
> > +
> > + /* Check to ensure the result of current->kunit_test
> > + * is equivalent to current test.
> > + */
> > + current_test = current->kunit_test;
> > + KUNIT_EXPECT_PTR_EQ(test, test, current_test);
>
> Perhaps we can combine this and the next test case down to
> static void kunit_current_test(struct kunit *test) {
> /* There are two different ways of getting the current test */
> KUNIT_EXPECT_PTR_EQ(test, test, current->kunit_test);
> KUNIT_EXPECT_PTR_EQ(test, test, kunit_get_current_test());
> }
> ?

Hi Daniel!

Yes, I would be happy to combine these for v2. I might want to alter
that proposed comment slightly. "Two different ways" seems a bit
unclear to me. Maybe: Check results of both current->kunit_test and
kunit_get_current_test() are equivalent to current test. What do you
think? I might send out a v2 with a proposed comment.

I agree the "initialize fake test" comment is self-explanatory. But if
we keep the comment regarding resetting the current test, I think we
should mark when we set the test as a fake with a comment as well.

> > + kunit_init_test(&fake, "fake test", NULL);
> > + KUNIT_EXPECT_EQ(test, fake.status, KUNIT_SUCCESS);
> > + current->kunit_test = &fake;
> > +
> > + /* Fail current test and expect status of fake test to be failed. */
>
> Nit: I think this comment could also be dropped or maybe shortened to
> kunit_fail_current_test("This should make `fake` fail");
>

This first option seems good to me.

> or
> /* Now kunit_fail_current_test() should modify `fake`, not `test` */
> kunit_fail_current_test("This should make `fake` fail");
>
> > + kunit_fail_current_test("This test is supposed to fail.");
> > + KUNIT_EXPECT_EQ(test, fake.status, (enum kunit_status)KUNIT_FAILURE);
> > +
>
> Hmm, should we try calling
> kunit_cleanup(&fake);
> ?
>
> Right now this does resource cleanups, but we might have other state
> to cleanup for our `fake` test object in the future.
>

I would be fine to add this here if it is wanted.

Thanks Daniel for the comments!

Rae

[Rae Moar]

Create test suite called "kunit_current" to add test coverage for the use
of current->kunit_test, which returns the current KUnit test.

Add two test cases:
- kunit_current_test to test current->kunit_test and the method

kunit_get_current_test(), which utilizes current->kunit_test.

- kunit_current_fail_test to test the method

kunit_fail_current_test(), which utilizes current->kunit_test.

Signed-off-by: Rae Moar <rm...@google.com>
---

Changes from v1->v2:
- Combine two test cases to test both ways of getting current test in
kunit_current_test.
- Changes to comments.
- Add kunit_cleanup to kunit_current_fail_test.

lib/kunit/kunit-test.c | 42 +++++++++++++++++++++++++++++++++++++++++-
1 file changed, 41 insertions(+), 1 deletion(-)

diff --git a/lib/kunit/kunit-test.c b/lib/kunit/kunit-test.c
index b63595d3e241..42e44caa1bdd 100644

--- a/lib/kunit/kunit-test.c
+++ b/lib/kunit/kunit-test.c
@@ -6,6 +6,7 @@
* Author: Brendan Higgins <brendan...@google.com>
*/
#include <kunit/test.h>
+#include <kunit/test-bug.h>

#include "try-catch-impl.h"

@@ -532,7 +533,46 @@ static struct kunit_suite kunit_status_test_suite = {
.test_cases = kunit_status_test_cases,
};

+static void kunit_current_test(struct kunit *test)
+{
+ /* Check results of both current->kunit_test and
+ * kunit_get_current_test() are equivalent to current test.
+ */
+ KUNIT_EXPECT_PTR_EQ(test, test, current->kunit_test);
+ KUNIT_EXPECT_PTR_EQ(test, test, kunit_get_current_test());
+}
+
+static void kunit_current_fail_test(struct kunit *test)

+{
+ struct kunit fake;
+

+ kunit_init_test(&fake, "fake test", NULL);
+ KUNIT_EXPECT_EQ(test, fake.status, KUNIT_SUCCESS);
+

+ /* Set current->kunit_test to fake test. */

+ current->kunit_test = &fake;
+

+ kunit_fail_current_test("This should make `fake` test fail.");

+ KUNIT_EXPECT_EQ(test, fake.status, (enum kunit_status)KUNIT_FAILURE);

+ kunit_cleanup(&fake);

+
+ /* Reset current->kunit_test to current test. */
+ current->kunit_test = test;
+}
+
+static struct kunit_case kunit_current_test_cases[] = {

+ KUNIT_CASE(kunit_current_test),
+ KUNIT_CASE(kunit_current_fail_test),

[Daniel Latypov]

What you went with in v2 works for me.
I'll take a look at the other changes in v2.

Thanks!
Daniel

[Daniel Latypov]

On Mon, Apr 3, 2023 at 1:19 PM Rae Moar <rm...@google.com> wrote:
>
> Create test suite called "kunit_current" to add test coverage for the use
> of current->kunit_test, which returns the current KUnit test.
>
> Add two test cases:
> - kunit_current_test to test current->kunit_test and the method
> kunit_get_current_test(), which utilizes current->kunit_test.
>
> - kunit_current_fail_test to test the method
> kunit_fail_current_test(), which utilizes current->kunit_test.
>
> Signed-off-by: Rae Moar <rm...@google.com>

Reviewed-by: Daniel Latypov <dlat...@google.com>

Looks good and runs well here.
It's nice to have a test for this given kunit_fail_current_test() went
from a simple function call to now using `kunit_hooks` to do an
indirect call and relying on the `kunit_running` static key.

[David Gow]

On Tue, 4 Apr 2023 at 04:19, Rae Moar <rm...@google.com> wrote:
>
> Create test suite called "kunit_current" to add test coverage for the use
> of current->kunit_test, which returns the current KUnit test.
>
> Add two test cases:
> - kunit_current_test to test current->kunit_test and the method
> kunit_get_current_test(), which utilizes current->kunit_test.
>
> - kunit_current_fail_test to test the method
> kunit_fail_current_test(), which utilizes current->kunit_test.
>
> Signed-off-by: Rae Moar <rm...@google.com>
> ---

Looks good to me, thanks!

Reviewed-by: David Gow <davi...@google.com>

Cheers,
-- David