From mboxrd@z Thu Jan 1 00:00:00 1970 From: Hal Rosenstock Subject: Re: [PATCH V1 7/9] ib/pma: add include file for IBA performance counters definitions Date: Wed, 06 Jul 2011 12:57:46 -0400 Message-ID: <4E14940A.9080700@dev.mellanox.co.il> References: <4DFA10CC.5080002@mellanox.com> <35AAF1E4A771E142979F27B51793A4888838E71F03@AVEXMB1.qlogic.org> <35AAF1E4A771E142979F27B51793A4888838FEE740@AVEXMB1.qlogic.org> <4E133575.8040105@mellanox.com> <35AAF1E4A771E142979F27B51793A4888838FEE894@AVEXMB1.qlogic.org> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: Sender: linux-rdma-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org To: Roland Dreier Cc: Mike Marciniszyn , Or Gerlitz , Or Gerlitz , linux-rdma , Ira Weiny , Hal Rosenstock , Sean Hefty List-Id: linux-rdma@vger.kernel.org On 7/6/2011 11:51 AM, Roland Dreier wrote: >> The patch that is not yet approved is https://patchwork.kernel.org/patch/523041/ >> >> The objections to date for the patch center on the use of reserved bits to key the existence of some vendor specific queries. > > Hmm, I'm not a huge fan of that patch since as you say it uses bits > that the IB spec says are reserved. I'd be more inclined to take Or's > patches which have gotten positive feedback from Sean, if we have to > choose. Hal (and/or anyone else who has more insight into the IBTA > intentions for management architecture), what do you think of the > patch linked above? When the patch was originally issued, I had commented on this issue (http://www.spinics.net/lists/linux-rdma/msg07020.html). There were two issues I saw: 1. Two CapabilityMask2 bits are set for congestion statistics which appears to be a proprietary feature. #define IB_PMA_CLASS_CAP_PORT_CONGS cpu_to_be32(3 << 30) In general, it's dangerous to use IBA reserved bits as they may be claimed by errata to support new features. Should this feature be taken forward to the IBTA ? 2. A reserved field in some structure was being set for xmit wait rather than updating the structure for the spec errata for this. I think this has been fixed in the latest version of the patch. You had originally commented on this issue in http://www.spinics.net/lists/linux-rdma/msg07058.html I can rereview the patch intoto if this is needed. -- Hal > > - R. -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org More majordomo info at http://vger.kernel.org/majordomo-info.html