In message <[EMAIL PROTECTED]>,Eli Cohen writes: >> > On the receive side the code is changed to post to receive queue once in >> > 16 completions. This is done in for both UD and and CM. >> >> Ohmm, have you tested latency? I think it will increase latency for >> small messages. > >I see also improvement in latency. I used netperf -t UDP_RR and -t >TCP_RR and in both cases I improvement.
this is just going to affect the resupply of rx buffers if i understand the code correctly. this wont delay the arrival of packets but perhaps it should be some fraction of the rq length instead of just 16. _______________________________________________ general mailing list [email protected] http://lists.openfabrics.org/cgi-bin/mailman/listinfo/general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
