On Mon, Jun 4, 2012 at 8:06 PM, Hefty, Sean <[email protected]> wrote:
> > Still, the 1st and most thing to handle here is feedback on the QP > > groups concept suggested by this patch set to support TSS/RSS over > > verbs. The plan is for this concept to (with little help from a > > framework for verbs extension) apply to user space RSS as well, > > for both UD and RAW QPs. > Conceptually, it seems that there should be more alignment with XRC, which > also > ends up with an associated group of QPs. > Is the use of these QPs restricted to TCP/UDP/IP traffic? Yep, sure, if its possible to align/squeez/unite these QP groups into the XRC interfaces could be nice to go that way. As for you question on the nature of the traffic, for example, RX wise, its one QP (the parent) which dispatches traffic to set of (child) QPs where with RSS the dispatching is based on hashing which is based on the assumptions that the packets are TCP/UDP/IP, below the RSS hood, packets for which the HW fails to apply hash on are dispatched to a default QP in that set. The suggested scheme matches well IB/UD QPs and Ethernet RAW packet QPs. Or. -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html
