| Message ID | 20240710170643.1401358-1-ericchancf@google.com (mailing list archive) |
|---|---|
| State | New |
| Headers | show |
| Series | kunit: Improve the readability and functionality of macro. | expand |
On Thu, 11 Jul 2024 at 01:06, Eric Chan <ericchancf@google.com> wrote: > > Both KUNIT_FAIL and KUNIT_ASSERT_FAILURE defined to KUNIT_FAIL_ASSERTION > with different tpye of kunit_assert_type. The current naming of > KUNIT_ASSERT_FAILURE and KUNIT_FAIL_ASSERTION is confusing due to their > similarities. To improve readability and symmetry, renames > KUNIT_ASSERT_FAILURE to KUNIT_ASSERT. Makes the naming consistent, > with KUNIT_FAIL and KUNIT_ASSERT being symmetrical. > Additionally, an explanation for KUNIT_ASSERT has been added to clarify > its usage. > > Signed-off-by: Eric Chan <ericchancf@google.com> > --- I personally am not a fan of KUNIT_ASSERT() as a name here: to me it implies that we're checking a boolean (like KUNIT_ASSERT_TRUE()). Does making this 'KUNIT_FAIL_AND_EXIT()' / 'KUNIT_FAIL_AND_ABORT()' or similar seem clearer to you? (Or possibly we could make this KUNIT_FAIL(), and make the existing KUNIT_FAIL() become KUNIT_MARK_FAILED(), though I think it's not worth the churn personally.) -- David > drivers/input/tests/input_test.c | 2 +- > include/kunit/assert.h | 2 +- > include/kunit/test.h | 13 ++++++++++++- > 3 files changed, 14 insertions(+), 3 deletions(-) > > diff --git a/drivers/input/tests/input_test.c b/drivers/input/tests/input_test.c > index 2fa5b725ae0a..cbab24a265fa 100644 > --- a/drivers/input/tests/input_test.c > +++ b/drivers/input/tests/input_test.c > @@ -31,7 +31,7 @@ static int input_test_init(struct kunit *test) > ret = input_register_device(input_dev); > if (ret) { > input_free_device(input_dev); > - KUNIT_ASSERT_FAILURE(test, "Register device failed: %d", ret); > + KUNIT_ASSERT(test, "Register device failed: %d", ret); > } > > test->priv = input_dev; > diff --git a/include/kunit/assert.h b/include/kunit/assert.h > index 24c2b9fa61e8..02c6f7bb1d26 100644 > --- a/include/kunit/assert.h > +++ b/include/kunit/assert.h > @@ -60,7 +60,7 @@ void kunit_assert_prologue(const struct kunit_loc *loc, > * struct kunit_fail_assert - Represents a plain fail expectation/assertion. > * @assert: The parent of this type. > * > - * Represents a simple KUNIT_FAIL/KUNIT_ASSERT_FAILURE that always fails. > + * Represents a simple KUNIT_FAIL/KUNIT_ASSERT that always fails. > */ > struct kunit_fail_assert { > struct kunit_assert assert; > diff --git a/include/kunit/test.h b/include/kunit/test.h > index 87a232421089..d1b085fd5dc3 100644 > --- a/include/kunit/test.h > +++ b/include/kunit/test.h > @@ -1193,7 +1193,18 @@ do { \ > fmt, \ > ##__VA_ARGS__) > > -#define KUNIT_ASSERT_FAILURE(test, fmt, ...) \ > +/** > + * KUNIT_ASSERT() - Always causes a test to assert when evaluated. > + * @test: The test context object. > + * @fmt: an informational message to be printed when the assertion is made. > + * @...: string format arguments. > + * > + * The opposite of KUNIT_SUCCEED(), it is an assertion that always fails. In > + * other words, it always results in a failed assertion, and consequently > + * always causes the test case to assert when evaluated. See KUNIT_ASSERT_TRUE() > + * for more information. > + */ > +#define KUNIT_ASSERT(test, fmt, ...) \ > KUNIT_FAIL_ASSERTION(test, KUNIT_ASSERTION, fmt, ##__VA_ARGS__) > > /** > -- > 2.45.2.803.g4e1b14247a-goog >
On Thu, Jul 11, 2024 at 11:49 AM David Gow <davidgow@google.com> wrote: > > On Thu, 11 Jul 2024 at 01:06, Eric Chan <ericchancf@google.com> wrote: > > > > Both KUNIT_FAIL and KUNIT_ASSERT_FAILURE defined to KUNIT_FAIL_ASSERTION > > with different tpye of kunit_assert_type. The current naming of > > KUNIT_ASSERT_FAILURE and KUNIT_FAIL_ASSERTION is confusing due to their > > similarities. To improve readability and symmetry, renames > > KUNIT_ASSERT_FAILURE to KUNIT_ASSERT. Makes the naming consistent, > > with KUNIT_FAIL and KUNIT_ASSERT being symmetrical. > > Additionally, an explanation for KUNIT_ASSERT has been added to clarify > > its usage. > > > > Signed-off-by: Eric Chan <ericchancf@google.com> > > --- > > I personally am not a fan of KUNIT_ASSERT() as a name here: to me it > implies that we're checking a boolean (like KUNIT_ASSERT_TRUE()). > > Does making this 'KUNIT_FAIL_AND_EXIT()' / 'KUNIT_FAIL_AND_ABORT()' or > similar seem clearer to you? > > (Or possibly we could make this KUNIT_FAIL(), and make the existing > KUNIT_FAIL() become KUNIT_MARK_FAILED(), though I think it's not worth > the churn personally.) > > -- David Hi David, Thank you very much for patiently reviewing. I understand your suggestion, indeed KUNIT_ASSERT will be misunderstood as still an assert behavior. But in fact, this macro has one extra abort behavior than KUNIT_FAIL. I think KUNIT_FAIL_AND_ABORT is a pretty good name to understand itself. I've updated patch v2 at [0]. Thanks for the reviewing and suggestions. [0] https://lore.kernel.org/lkml/20240711193729.108720-1-ericchancf@google.com/ Sincerely, Eric Chan > > > > > > drivers/input/tests/input_test.c | 2 +- > > include/kunit/assert.h | 2 +- > > include/kunit/test.h | 13 ++++++++++++- > > 3 files changed, 14 insertions(+), 3 deletions(-) > > > > diff --git a/drivers/input/tests/input_test.c b/drivers/input/tests/input_test.c > > index 2fa5b725ae0a..cbab24a265fa 100644 > > --- a/drivers/input/tests/input_test.c > > +++ b/drivers/input/tests/input_test.c > > @@ -31,7 +31,7 @@ static int input_test_init(struct kunit *test) > > ret = input_register_device(input_dev); > > if (ret) { > > input_free_device(input_dev); > > - KUNIT_ASSERT_FAILURE(test, "Register device failed: %d", ret); > > + KUNIT_ASSERT(test, "Register device failed: %d", ret); > > } > > > > test->priv = input_dev; > > diff --git a/include/kunit/assert.h b/include/kunit/assert.h > > index 24c2b9fa61e8..02c6f7bb1d26 100644 > > --- a/include/kunit/assert.h > > +++ b/include/kunit/assert.h > > @@ -60,7 +60,7 @@ void kunit_assert_prologue(const struct kunit_loc *loc, > > * struct kunit_fail_assert - Represents a plain fail expectation/assertion. > > * @assert: The parent of this type. > > * > > - * Represents a simple KUNIT_FAIL/KUNIT_ASSERT_FAILURE that always fails. > > + * Represents a simple KUNIT_FAIL/KUNIT_ASSERT that always fails. > > */ > > struct kunit_fail_assert { > > struct kunit_assert assert; > > diff --git a/include/kunit/test.h b/include/kunit/test.h > > index 87a232421089..d1b085fd5dc3 100644 > > --- a/include/kunit/test.h > > +++ b/include/kunit/test.h > > @@ -1193,7 +1193,18 @@ do { \ > > fmt, \ > > ##__VA_ARGS__) > > > > -#define KUNIT_ASSERT_FAILURE(test, fmt, ...) \ > > +/** > > + * KUNIT_ASSERT() - Always causes a test to assert when evaluated. > > + * @test: The test context object. > > + * @fmt: an informational message to be printed when the assertion is made. > > + * @...: string format arguments. > > + * > > + * The opposite of KUNIT_SUCCEED(), it is an assertion that always fails. In > > + * other words, it always results in a failed assertion, and consequently > > + * always causes the test case to assert when evaluated. See KUNIT_ASSERT_TRUE() > > + * for more information. > > + */ > > +#define KUNIT_ASSERT(test, fmt, ...) \ > > KUNIT_FAIL_ASSERTION(test, KUNIT_ASSERTION, fmt, ##__VA_ARGS__) > > > > /** > > -- > > 2.45.2.803.g4e1b14247a-goog > >
diff --git a/drivers/input/tests/input_test.c b/drivers/input/tests/input_test.c index 2fa5b725ae0a..cbab24a265fa 100644 --- a/drivers/input/tests/input_test.c +++ b/drivers/input/tests/input_test.c @@ -31,7 +31,7 @@ static int input_test_init(struct kunit *test) ret = input_register_device(input_dev); if (ret) { input_free_device(input_dev); - KUNIT_ASSERT_FAILURE(test, "Register device failed: %d", ret); + KUNIT_ASSERT(test, "Register device failed: %d", ret); } test->priv = input_dev; diff --git a/include/kunit/assert.h b/include/kunit/assert.h index 24c2b9fa61e8..02c6f7bb1d26 100644 --- a/include/kunit/assert.h +++ b/include/kunit/assert.h @@ -60,7 +60,7 @@ void kunit_assert_prologue(const struct kunit_loc *loc, * struct kunit_fail_assert - Represents a plain fail expectation/assertion. * @assert: The parent of this type. * - * Represents a simple KUNIT_FAIL/KUNIT_ASSERT_FAILURE that always fails. + * Represents a simple KUNIT_FAIL/KUNIT_ASSERT that always fails. */ struct kunit_fail_assert { struct kunit_assert assert; diff --git a/include/kunit/test.h b/include/kunit/test.h index 87a232421089..d1b085fd5dc3 100644 --- a/include/kunit/test.h +++ b/include/kunit/test.h @@ -1193,7 +1193,18 @@ do { \ fmt, \ ##__VA_ARGS__) -#define KUNIT_ASSERT_FAILURE(test, fmt, ...) \ +/** + * KUNIT_ASSERT() - Always causes a test to assert when evaluated. + * @test: The test context object. + * @fmt: an informational message to be printed when the assertion is made. + * @...: string format arguments. + * + * The opposite of KUNIT_SUCCEED(), it is an assertion that always fails. In + * other words, it always results in a failed assertion, and consequently + * always causes the test case to assert when evaluated. See KUNIT_ASSERT_TRUE() + * for more information. + */ +#define KUNIT_ASSERT(test, fmt, ...) \ KUNIT_FAIL_ASSERTION(test, KUNIT_ASSERTION, fmt, ##__VA_ARGS__) /**
Both KUNIT_FAIL and KUNIT_ASSERT_FAILURE defined to KUNIT_FAIL_ASSERTION with different tpye of kunit_assert_type. The current naming of KUNIT_ASSERT_FAILURE and KUNIT_FAIL_ASSERTION is confusing due to their similarities. To improve readability and symmetry, renames KUNIT_ASSERT_FAILURE to KUNIT_ASSERT. Makes the naming consistent, with KUNIT_FAIL and KUNIT_ASSERT being symmetrical. Additionally, an explanation for KUNIT_ASSERT has been added to clarify its usage. Signed-off-by: Eric Chan <ericchancf@google.com> --- drivers/input/tests/input_test.c | 2 +- include/kunit/assert.h | 2 +- include/kunit/test.h | 13 ++++++++++++- 3 files changed, 14 insertions(+), 3 deletions(-)