Re: [PATCH 0/2] MR2573: dinput: Set guidType = GUID_Unknown for HID collections.

[Florian Will (＠w-flo)]

Florian Will (@w-flo) commented about dlls/dinput/tests/joystick8.c:
> - if (FAILED(hr = dinput_test_create_device( DIRECTINPUT_VERSION, &devinst, &device ))) goto done;
> -
> - check_dinput_devices( DIRECTINPUT_VERSION, &devinst );
> -
> - hr = IDirectInputDevice8_EnumObjects( device, check_objects, &check_objects_params, DIDFT_ALL );
> - ok( hr == DI_OK, "EnumObjects returned %#lx\n", hr );
> - ok( check_objects_params.index >= check_objects_params.expect_count, "missing %u objects\n",
> -
>    check_objects_params.expect_count - check_objects_params.index );
> -
> - ref = IDirectInputDevice8_Release( device );
> - ok( ref == 0, "Release returned %ld\n", ref );
> -
>
> +done:
> - hid_device_stop( &desc, 1 );
> - cleanup_registry_keys();
> - winetest_pop_context();

I just noticed this: I copied the test_driving_wheel_axes() function for my test, and I believe both test_driving_wheel_axes() and my new test should drop the winetest_pop_context() because they don't push anything. Or maybe they should push something at the start of the test? Same with the test_many_axes_joystick() test and maybe others. It seems off-topic for this MR though.