[Spice-devel] [PATCH] macros: make SPICE_CONTAINEROF more typesafe

[Frediano Ziglio]

> 
> Hey,
> 
> I just realized that this patch breaks the build of spice 0.12.6 (and
> earlier). I don't think upgrading spice-protocol should cause such
> breakage, if we can't turn this to a warning, I would revert the patch
> (or wrap it in some #ifdef SPICE_STRICT_TYPE_CHECK)
> 
> Christophe
> 

Won't be better to fix the code?
And why old versions should use a newer spice-protocol ?

Frediano

> On Thu, Dec 03, 2015 at 03:28:09PM +0000, Frediano Ziglio wrote:
> > Check the pointer given is the same type as member pointer.
> > 
> > Signed-off-by: Frediano Ziglio <fziglio at redhat.com>
> > ---
> >  spice/macros.h | 6 ++++++
> >  1 file changed, 6 insertions(+)
> > 
> > 
> > Note that some spice-server need some update in order to compile after
> > this macro change.
> > 
> > 
> > diff --git a/spice/macros.h b/spice/macros.h
> > index cd63528..2efbaff 100644
> > --- a/spice/macros.h
> > +++ b/spice/macros.h
> > @@ -143,8 +143,14 @@
> >      ((long) ((uint8_t*) &((struct_type*) 0)->member))
> >  #endif
> >  
> > +#if    __GNUC__ > 3 || (__GNUC__ == 3 && __GNUC_MINOR__ >= 4)
> > +#define SPICE_CONTAINEROF(ptr, struct_type, member) ({ \
> > +    const typeof( ((struct_type *)0)->member ) *__mptr = (ptr);    \
> > +    ((struct_type *)(void *)((uint8_t *)(__mptr) -
> > SPICE_OFFSETOF(struct_type, member))); })
> > +#else
> >  #define SPICE_CONTAINEROF(ptr, struct_type, member) \
> >      ((struct_type *)(void *)((uint8_t *)(ptr) -
> >      SPICE_OFFSETOF(struct_type, member)))
> > +#endif
> >  
> >  #define SPICE_MEMBER_P(struct_p, struct_offset)   \
> >      ((gpointer) ((guint8*) (struct_p) + (glong) (struct_offset)))
> > --
> > 2.4.3
> >