[email protected] wrote:
> On (09/17/09 14:28), James Carlson wrote:
>> [email protected] wrote:
>>> I've added code that will ignore the RTM_LOSING if there's
>>> no alternate default router available.
>> It'd be nice if in.routed were smart enough to ignore RTM_LOSING when
>> there are generally a lack of alternatives, and not just for RDISC.  Any
>> time we're about to do accelerated (or even perhaps _normal_) age-out of
>> a route, and removal would mean that there's no longer a path to that
>> subnet, then making sure it lingers as metric 15 instead would be a bit
>> smarter.
>>
>> table.c
>>   1240: consider printing the symbolic address rather than raw hex;
>>      use naddr_ntoa().
>>
>>   2265,2296,2450: is a sowmini stop like a bus stop or a command?  ;-}
>>
> 
> See 
> 
>    http://cr.opensolaris.org/~sowmini/routed_fixes/
> 
> which now handles both rip and rdisc, for all prefixes (including
> default)

Nice!

Bug at line 1262: that should be "%s", not "%x".

Nit at line 1256 -- the message should say something about why it's
reporting a lack of a route.  (got RTM_LOSING, but ...)  Similar issue
with 1270.  Nits at 1263 and 1270: \n and trailing spaces probably not
needed.

-- 
James Carlson         42.703N 71.076W         <[email protected]>
_______________________________________________
networking-discuss mailing list
[email protected]

Reply via email to