On Tue, Apr 11, 2006 at 11:35:44AM -0700, Sangeeta Misra wrote:
> >>I got the M_BREAK cleanup ( justification and the code cleanup) checked 
> >>out by THiru, but yet we will test this as you point out above.
> >
> >
> >You may be able to get rid of the b_next massaging too.  If you can't, please
> >restate why you need to massage b_next in ip_rput().
> 
> Yes. We should get rid of the b_next assigmentas well.

Ahhh, cool!  Glad to help reduce lines of code.  (Meem would be so proud!)

> Also you state:
> --------
> This part of the block comment:
> 
>       * M_BREAK are also used to pass back in multicast packets
>       * that are encapsulated with a source route.
> 
> Bothers me *just* enough so that we should probably make sure we aren't
> hitting this codepath.  A simple test:
> 
>       ping -g <Surya-router-1> -g <Surya-router-2> 224.0.0.1
> 
> Would check if you get an M_BREAK on your Surya box.  Assuming that's 
> kosher,you're in good shape.
> ----------
> Can you explain how the above ping command would trigger the M_BREAK 
> logic? Also I plan to run this with onnv first to see if it works there.

The high-probability target is Surya-router-2.  It will process the
source-route packet it receives, swap it's own destination for the 224.0.0.1,
and then transmit it as if it originated a packet destined for 224.0.0.1.

If there's any M_BREAK left relating to source routing and a multicast
destination, Surya-router-2 in my example above will have to deal with it.
So put onnv on Surya-router-2 (and you can even eliminate Surya-router-1 or
make it onnv also), and see if it works at all.

> What should surya-router-1/Surya-router-2 config be?

Enable IP source route forwarding, IP forwarding, and possibly multicast
forwarding on Surya-router-2.

> Should it be running as multicast routers ( ie mrouted daemon).

It doesn't need to run mrouted, AFAIK.

> I hope this does not require multicast tunnels, as that is already broken
> in onnv( I filed a bug against this).

My suggested ping does NOT require multicast tunnels, fortunately.

BTW, I'd suggest *eliminating* multicast tunnels in favor of our existing
generic ip-in-ip tunnels instead, but that's another discussion for another
time.

Thanks,
Dan
_______________________________________________
networking-discuss mailing list
[email protected]

Reply via email to