drm_analog_tv_mode() and its variants return a drm_display_mode that needs to be destroyed later one. The drm_test_modes_analog_tv_mono_576i() test never does however, which leads to a memory leak.
Let's make sure it's freed. Closes: https://lore.kernel.org/dri-devel/a7655158a6367ac46194d57f4b7433ef0772a73e.ca...@mailbox.org/ Fixes: e31538489908 ("drm/tests: Add tests for the new Monochrome value of tv_mode") Signed-off-by: Maxime Ripard <mrip...@kernel.org> --- drivers/gpu/drm/tests/drm_modes_test.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/drivers/gpu/drm/tests/drm_modes_test.c b/drivers/gpu/drm/tests/drm_modes_test.c index 7ba646d87856f52f99b0f9bcfed0c31d848c798d..f5b20f92df8be73ad9ef46aedae8b855e637e4dd 100644 --- a/drivers/gpu/drm/tests/drm_modes_test.c +++ b/drivers/gpu/drm/tests/drm_modes_test.c @@ -154,17 +154,21 @@ static void drm_test_modes_analog_tv_pal_576i_inlined(struct kunit *test) static void drm_test_modes_analog_tv_mono_576i(struct kunit *test) { struct drm_test_modes_priv *priv = test->priv; struct drm_display_mode *mode; + int ret; mode = drm_analog_tv_mode(priv->drm, DRM_MODE_TV_MODE_MONOCHROME, 13500 * HZ_PER_KHZ, 720, 576, true); KUNIT_ASSERT_NOT_NULL(test, mode); + ret = drm_kunit_add_mode_destroy_action(test, mode); + KUNIT_ASSERT_EQ(test, ret, 0); + KUNIT_EXPECT_EQ(test, drm_mode_vrefresh(mode), 50); KUNIT_EXPECT_EQ(test, mode->hdisplay, 720); /* BT.601 defines hsync_start at 732 for 576i */ KUNIT_EXPECT_EQ(test, mode->hsync_start, 732); -- 2.49.0
