On 16:01 Wed 04 Feb , Hal Rosenstock wrote: > > I think that position is extreme. I don't think I broke anything that > wasn't already broken.
At least after fast look: (port_num > p_mon_node->redir_tbl_size) and similar tests look broken, using osm_get_node_by_guid() likely slows down existing PerfMgr. :( Both triggered by those patches. I would be fine with subsequent patches if there would no degradations. > Anyhow, if you really want, I'll produce one patch for these changes. Thanks. Sasha _______________________________________________ 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
