Drop one "i" from the subject :) On Wed, Jun 10, 2015 at 04:09:14PM +0200, Andreas Pokorny wrote: > --- > test/litest.h | 7 +++++++ > 1 file changed, 7 insertions(+) > > diff --git a/test/litest.h b/test/litest.h > index 3a08b6a..0d39dd3 100644 > --- a/test/litest.h > +++ b/test/litest.h > @@ -428,4 +428,11 @@ void litest_semi_mt_touch_up(struct litest_device *d, > #define ck_assert_notnull(ptr) ck_assert_ptr_ne(ptr, NULL) > #endif > > +#define litest_assert_double_eq(X,Y) do { \ > + double _litest_x = (X); \ > + double _litest_y = (Y); \
fwiw, you don't need () here, there's no danger of wrong expansion. otherwise please use the same naming convention that all the other helpers have. > + ck_assert_msg(fabs(_litest_x - _litest_y) < 1E-5, "Assertion '" #X " == " > #Y "' failed: "#X"==%f, "#Y"==%f", _litest_x, _litest_y); \ > +} while (0) > + nope, that's not right. the litest_assert_ macros have a specific behaviour (see 84e6321f515f0683), using ck_assert_msg will make it inconsistent. #define this as ck_assert_double and we'll deal with conflicts if check ever adds the equivalent. and for good measure, add ge, gt, le, lt and ne as well. Cheers, Peter > + > #endif /* LITEST_H */ > -- > 2.1.4 > > _______________________________________________ > wayland-devel mailing list > wayland-devel@lists.freedesktop.org > http://lists.freedesktop.org/mailman/listinfo/wayland-devel > _______________________________________________ wayland-devel mailing list wayland-devel@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/wayland-devel