[openib-general] CM and REP handling

[Sean Hefty]

Viswanath Krishnamurthy wrote:
> In the current communication manager (CM) implementation how is the REP MAD
> getting lost handled. When the REP gets lost, the cm_dup_req_handler 
> gets called
> which currently enters the default condition and does nothing.  The 
> client retries
> the number of timers it is configured to and fails.  If the first REP 
> gets lost, the connection
> never gets established. So what should be the behavior ?

The REP will be resent until an RTU is received.  Repeated REQs can be dropped 
in cm_dup_req_handler() because the initial REQ has been received and a REP 
generated.  That is, cm_dup_req_handler() is called on the side sending the REP.

Are you seeing an issue with the code when the first REP is lost?

- Sean