[ofa-general] rdma_resolve_route() returning -EINVAL

[Talpey, Thomas]

At 09:09 PM 10/2/2008, Roland Dreier wrote:
>It was added in:
>
>commit 38ca83a588662f0af684ba2567dd910a564268ab
>Author: Amir Vadai <amirv at mellanox.co.il>
>Date:   Tue Jul 22 14:14:23 2008 -0700
>
>    RDMA/cma: Add RDMA_CM_EVENT_TIMEWAIT_EXIT event
>    
>    Consumers that want to re-use their QPs in new connections need to
>    know when the QP has exited the timewait state.  Report the timewait
>    event through the rdma_cm.
>
>basically you can't really free an IB QP until it fully leaves the CM
>state machine, ie when it finishes with timewait.

Ok,. I guess that's not obvious from the name, but in particular if it's
so important, then why doesn't anyone implement it? It happens pretty
much iimmediately on my IB adapter QPs, under what conditions would
it be delayed? And since rdma_destroy_qp() is void, how would I know
if I call it too early?

I'm think maybe I should be waiting for it, and not the disconnect event,
before beginning my connection recovery...

Tom.