On Sat, May 07, 2016 at 09:38:07AM -0400, Frediano Ziglio wrote:
> >
> > Silence compile-time warning with clang:
> > comparison of constant 16 with expression of type 'SpiceROP' is always
> > true [-Wtautological-constant-out-of-range-compare]
> > ---
> > common/pixman_utils.c | 4 ++--
> > common/pixman_utils.h | 3 ++-
> > 2 files changed, 4 insertions(+), 3 deletions(-)
> >
> > diff --git a/common/pixman_utils.c b/common/pixman_utils.c
> > index 81806c9..5a03ff2 100644
> > --- a/common/pixman_utils.c
> > +++ b/common/pixman_utils.c
> > @@ -311,7 +311,7 @@ void spice_pixman_fill_rect_rop(pixman_image_t *dest,
> > spice_assert(height > 0);
> > spice_assert(x + width <= pixman_image_get_width(dest));
> > spice_assert(y + height <= pixman_image_get_height(dest));
> > - spice_assert(rop < 16);
> > + spice_assert(rop < SPICE_ROP_END);
> >
> > if (depth == 8) {
> > solid_rop_8_func_t rop_func = solid_rops_8[rop];
>
> What?
> Honestly I think either we are enabling too much warning or clang is wrong!
> This is C, not C++, enumerators can be different quite easily so there is no
> reason to assume that (I would open a bug to clang if this is inside -Wall).If you read the C standard literally, enum SpiceROP rop can never have 16 as a value, as 16 is not one of the values listed in enum SpiceROP definition. Thus, 'rop' being outside of the [0, 15] range is undefined behaviour/should never happen. So I would not say clang is wrong. But maybe a bit overzealous with the warning ;) Christophe
signature.asc
Description: PGP signature
_______________________________________________ Spice-devel mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/spice-devel
