[openib-general] [PATCH] IPOIB: Use a GRH when appropriate for unicast packets

[Sean Hefty]

>> If name service resolution gives me an IPv6 address that's off of the local
>> subnet, but the ARP response gives me an address that's on the local
>> subnet, then I think we can assume that ARP was unsuccessful is resolving
>> the address to the remote GID.  (I.e. the GID should be for a router.)  If
>> this is true, then we need some other way to acquire the DGID.
>
>This is where I think you have problems... Why would you ARP for an
>off-subnet address? Why would the router answer?  You push the address
>through the route table and ARP the router address that results.

I'm confusing myself.  I was considering different IB subnets, and trying to
determine whether they shared the same IP subnet.  The GIDs may have different
subnet prefixes, but the IP addresses may not, and I'm not sure how to relate
this back to using DNS.

>All of that is why I think another netdevice is a tidy
>solution. ping6/tcp/etc using this device would generate packets that
>follow the same path as RMDA connections would. No special rules about
>broadcast groups are required. The route table is used to instruct the
>kernel what IPv6 prefixes are IB GIDs and which are not by associating
>the output of the route with the ib0 device. The admins can use any
>means to set that up. Something that looks like:

At first glance, this seems like a decent approach to explore.

>But yes, it is a bit outside what the current framework envisions..

I'm fine with that.  My short-term objective is to enable basic router support
within the host stack, and I think I have an idea of what that takes.  I'd just
also like to have an idea of how an application could transfer data between
routed IB subnets, including providing a way for the application to locate a
given remote node.

- Sean