On Mon, Feb 7, 2022 at 8:01 PM David Gow davidgow@google.com wrote:
The list_del_init_careful() function was added[1] after the list KUnit test. Add a very basic test to cover it.
Note that this test only covers the single-threaded behaviour (which matches list_del_init()), as is already the case with the test for list_empty_careful().
Signed-off-by: David Gow davidgow@google.com
Changes since v1: https://lore.kernel.org/linux-kselftest/20220205061539.273330-1-davidgow@goo...
- Patch 1/3 unchanged
lib/list-test.c | 19 +++++++++++++++++++ 1 file changed, 19 insertions(+)
diff --git a/lib/list-test.c b/lib/list-test.c index ee09505df16f..976e9ae1f3c5 100644 --- a/lib/list-test.c +++ b/lib/list-test.c @@ -161,6 +161,24 @@ static void list_test_list_del_init(struct kunit *test) KUNIT_EXPECT_TRUE(test, list_empty_careful(&a)); }
+static void list_test_list_del_init_careful(struct kunit *test) +{
/* This test doesn't check correctness under concurrent access */
nit: I personally didn't read this comment in the intended way at first. I'd personally find something like
NOTE: this doesn't test for concurrency/memory ordering/however-you-want-to-word-it issues
a bit more readable.
struct list_head a, b;LIST_HEAD(list);list_add_tail(&a, &list);list_add_tail(&b, &list);/* before: [list] -> a -> b */list_del_init(&a);
Is this supposed to use list_del_init_careful(&a)? That would make it match the name of the test case.
/* after: [list] -> b, a initialised */KUNIT_EXPECT_PTR_EQ(test, list.next, &b);KUNIT_EXPECT_PTR_EQ(test, b.prev, &list);KUNIT_EXPECT_TRUE(test, list_empty_careful(&a));+}
static void list_test_list_move(struct kunit *test) { struct list_head a, b; @@ -707,6 +725,7 @@ static struct kunit_case list_test_cases[] = { KUNIT_CASE(list_test_list_replace_init), KUNIT_CASE(list_test_list_swap), KUNIT_CASE(list_test_list_del_init),
KUNIT_CASE(list_test_list_del_init_careful), KUNIT_CASE(list_test_list_move), KUNIT_CASE(list_test_list_move_tail), KUNIT_CASE(list_test_list_bulk_move_tail),-- 2.35.0.263.gb82422642f-goog