http://defect.opensolaris.org/bz/show_bug.cgi?id=13416



--- Comment #5 from Anurag S. Maskey <Anurag.Maskey at Sun.COM> 2009-12-24 
15:56:58 UTC ---
(In reply to comment #4)
> However, one major issue still exists.  The Refactor project changed
> ill_dls_info() in ip_if.c calling ill_get_name() instead of ipif_get_name(). 
> So, now the routing socket messages don't include the lifnum.  Interfaces 
> names
> in the routing socket messages should appear as "bge0:1", but appear as 
> "bge0".
>  This creates a problem in nwamd_ncu_handle_if_state_event().  We get the 
> flags
> on the interface to determine if we have stateless, v4dhcp, v6dhcp or static
> address configured.  Since the lifnum is not included, we end up getting the
> flags for the 0th lifnum.  The v6 test cases in the nwam test suite are 
> failing
> because of this.

Erik Nordmark's response to the change in the routing socket messages:

> It was a fundamental change due to the decoupling of the IREs from source 
> address selection. The result of this is that there is no ire_ipif pointer; 
> we have an ire_ill instead.
>
> Note that previously the ire_ipif was overloaded to mean different things - 
> in some cases it was the origin of the source address, but other code used it 
> as an indication of where packets were sent out. Thus I suspect that if you 
> used usesrc or a deprecated address (something which affected ire_ipif) the 
> ifname you would have seen in a routing socket message would have been a tad 
> > confusing. 

So, when we receive RTM_CHGADDR and RTM_FREEADDR, we will walk the interfaces
list to checking the address on each interface to the address in the routing
socket message to figure out the interface name with lifnum.

-- 
Configure bugmail: http://defect.opensolaris.org/bz/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the QA contact for the bug.

Reply via email to