> In the absence of any further discussions about the IPoIB CM without SRQ > patches, I will incorporate Sean Hefty's comments and plan to resubmit > the patches, unless I hear something soon.
Sorry for not devoting enough time to this, but something always seems to come up, and I really want to be able to focus a concentrated chunk of time on this, and I never seem to be able to. Anyway, I would prefer to find a solution that everyone can agree on, without me having to rule by decree. I think updating the patch is a good idea. Although I didn't get a chance to review it carefully there were a number of obvious messy parts that should be cleaned up. I am beginning to think that your basic approach is probably right, but I also still think it should be possible to handle both SRQ and non-SRQ without any overhead on the fast path. I don't understand the "maintainability" argument against doing this. Can you expand on your position a little? Thanks, Roland _______________________________________________ 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
