> so, if and only if the out dev is bond to a bridge dev would this side
> effect happen,
> right?

To be completely clear:
if and only if the out dev is a bridge dev _and_ br-nf is compiled into the 
kernel this side effect will happen.
This won't happen if the routing table sends the packet directly to a physical 
dev enslaved to a bridge dev instead of to the bridge dev itself.

> as source mac address match is a basic function of netfilter(not in
> patch-o-matic),
> it would be better to have a note in the man page.

It should be written down somewhere, but we don't have a br-nf man page. I'll 
put it down in the "how it works" doc.

cheers,
Bart

_______________________________________________
Bridge mailing list
[EMAIL PROTECTED]
http://www.math.leidenuniv.nl/mailman/listinfo/bridge

Reply via email to