> > [ Sorry for the delay in following up.  As in the past, I've elided issues
 > >   that seemed to require no further discussion. ]
 > 
 > It looks like we're in sync.  The big issue was that I tried to review
 > the kernel ARP code without reviewing the IP code, and there are some
 > significant places where they have interactions.

Indeed -- I think a functional API could've made things a lot clearer.

-- 
meem

Reply via email to