[openib-general] Re: [PATCH] CMA and iWARP

[Tom Tucker]

Ok, you got me ;-) Busted.  Steve pointed out another one -- the
connection is already established, then the filter rule is added, the
connection continues (bypassing the rule) until the ARP cache entry
times out...ga!

BTW, the same holds true for RSO (a netdev sanctioned non-offload,
offload). Albeit, the period of surprise is somewhat shorter, i.e. a
dozen or so frames.

The netdev people need to be involved. I think it will be briefly
painful, but ultimately worthwhile to do this "right". 

On Tue, 2006-01-24 at 09:13 -0800, Roland Dreier wrote:
>     Tom> The intended behavior is to provide "full coordination". For
>     Tom> the example you give, I would expect that rdma_resolve_addr
>     Tom> would fail due to to a timeout waiting for an ARP reply.
> 
> OK, now I'm going off into crazy-land, but I could have a rule that
> filters on source MAC and ethertype, and lets ARPs but no other
> packets through.
> 
>  - R.