On Mon, Jan 09, 2012 at 11:03:22AM -0800, Pravin Shelar wrote:
> On Mon, Jan 9, 2012 at 9:04 AM, Ben Pfaff <b...@nicira.com> wrote:
> > On Mon, Jan 09, 2012 at 10:10:41AM -0500, Brian Haley wrote:
> >> On 01/06/2012 07:01 PM, Pravin B Shelar wrote:
> >> > --- a/lib/packets.h
> >> > +++ b/lib/packets.h
> >> > @@ -464,4 +464,14 @@ void *snap_compose(struct ofpbuf *, const uint8_t
> >> > eth_dst[ETH_ADDR_LEN],
> >> > ? ? ? ? ? ? ? ? ? ? const uint8_t eth_src[ETH_ADDR_LEN],
> >> > ? ? ? ? ? ? ? ? ? ? unsigned int oui, uint16_t snap_type, size_t size);
> >> >
> >> > +static inline int ipv6_addr_is_zero(const struct in6_addr *addr)
> >> > +{
> >> > + ? ?const uint32_t *a = (const uint32_t *)&addr->s6_addr[0];
> >> > + ? ?int i, ret = 0;
> >> > +
> >> > + ? ?for (i = 0; i < 4; i++) {
> >> > + ? ? ? ?ret += a[i];
> >> > + ? ?}
> >> > + ? ?return !ret;
> >> > +}
> >>
> >> There's already an IN6_IS_ADDR_UNSPECIFIED macro defined in in.h, which
> >> every
> >> platform should have for RFC compliance. ? [...]
> >
> > Thanks. ?I didn't know about that one.
> >
> > Pravin, let's use IN6_IS_ADDR_UNSPECIFIED instead of adding a new
> > function.
>
> As suggested, I am checking nw_proto to validate ip header. So I do
> not need this function anymore.
Even better, thanks.
_______________________________________________
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
