Neil Horman wrote:
On Sat, Jan 20, 2007 at 08:05:07AM +0900, YOSHIFUJI Hideaki / 吉藤英明 wrote:
Hello.
In article <[EMAIL PROTECTED]> (at Fri, 19 Jan 2007 16:23:14 -0500), Neil Horman
<[EMAIL PROTECTED]> says:
Patch to Implement IPv6 RFC 4429 (Optimistic Duplicate Address Detection). In
Good work. We will see if this would break core and basic ipv6 code.
Dave, please hold on.
Some quick comments.
--yoshfuji
<snip>
Reposted patch, with your suggestions/corrections incorporated. The only thing
I left alone was your last comment regarding the checking of saddr for being a
unicast address. According to the RFC as I read it, its possible to receive
valid neighbor soliciations with a source address that is the unspecified
address, which I didn't think ipv6_addr_type flagged as being unicast. Now it
may be that such NS messages are discarded before arriving at that recieve
routine, but I was figuring better safe than sorry. If you're sufficiently
confident that we won't see non-unicast saddrs there, let me know and I'll
happily remove that as well.
Thanks & Regards
Neil
NTATIVE) {
+
+ if (ifp->flags & (IFA_F_TENTATIVE|IFA_F_OPTIMISTIC)) {
/* Address is tentative. If the source
- is unspecified address, it is someone
- does DAD, otherwise we ignore solicitations
- until DAD timer expires.
+ is unspecified address, someone else
+ is doing DAD, and if its not us, then
+ we need to fail our own DAD
+ RFC 4429 Sec 3.3
*/
if (!dad)
goto out;
@@ -816,8 +845,20 @@ static void ndisc_recv_ns(struct sk_buff *skb)
goto out;
}
}
- addrconf_dad_failure(ifp);
- return;
+
+ /* The one exception to the above rule about
+ optimistic addresses is that we need to always
+ respond to an NS from a unicast address if we are
+ optimistic. RFC 4429 Sec 3.3. If (unicast
+ and optimistic) are false then we can just fail
+ dad now.
+ */
+ type = ipv6_addr_type(saddr);
+ if (!((ifp->flags & IFA_F_OPTIMISTIC) &&
+ (type & IPV6_ADDR_UNICAST))) {
+ addrconf_dad_failure(ifp);
+ return;
+ }
}
I think you should remove / modify the :
if (!dad)
goto out;
which makes the rfc4429 tests not functional.
--Mika
-
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/majordomo-info.html
