Michael S. Tsirkin wrote:
But now we only do the list walk if the ip route resolution returns
a loopback device, so we are ok, right?
That's the way it seems to me (at least for SDP but not CMA (addr)).
OK, Sean, do you want to cook up a patch for CMA based on this code of mine?
I could do it too but I dont have a way to test cma yet.
I need to read back over this thread; I didn't follow it the first time. (At
this point, I think that it makes more sense to convert SDP to use the CMA or
ib_addr, rather than duplicating their functionality.)
There are both kernel and userspace test programs (cmatose) for the CMA checked
into the tree.
ip_addr can use a different approach than ip_dev_find, but I think it makes
sense to use existing kernel functionality wherever possible. If ip_dev_find
cannot be modified to support IPv6 addresses, then how about adding a API that
is similar, but takes a pointer to an address, along with an indication of the
address family that's used?
I can work on a patch for this.
- Sean
_______________________________________________
openib-general mailing list
[email protected]
http://openib.org/mailman/listinfo/openib-general
To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
