From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Steve Wise" Subject: RE: [PATCH] RDMA/nes: Adding queue drain functions Date: Mon, 4 Apr 2016 09:42:35 -0500 Message-ID: <03d101d18e80$3a838170$af8a8450$@opengridcomputing.com> References: <20160329175837.GA9868@TENIKOLO-MOBL2> <570141FE.4090400@grimberg.me> <20160404013152.GD5264@leon.nu> <57020C15.3000601@grimberg.me> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <57020C15.3000601-NQWnxTmZq1alnMjI0IkVqw@public.gmane.org> Content-Language: en-us Sender: linux-rdma-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org To: 'sagig' , leon-2ukJVAZIZ/Y@public.gmane.org Cc: 'Tatyana Nikolova' , 'Doug Ledford' , linux-rdma-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, faisal.latif-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org List-Id: linux-rdma@vger.kernel.org > > Sagi, > > Can you please explain why iWARP devices need explicit drain QP logic? > > Steve can explain it much better than I can (so I'd wait for him to > confirm). > > My understanding is that in iWARP, after moving the QP to error state, > all the posts should fail instead of completing with a flush error (I > think that SQ_DRAIN is the state that triggers flush errors). > > This is why iWARP devices need a special handling for draining a > queue-pair. Sorry for the delayed reply Leon, I replied to your original email today (I was on vacation). Sagi is correct, iWARP QPs need to fail post_send/recv requests synchronously and I didn't want to divert from the iWARP Verbs specification. Steve. -- 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