From mboxrd@z Thu Jan 1 00:00:00 1970 From: Christoph Hellwig Subject: Re: SRPt oops with 4.5-rc3-ish Date: Tue, 1 Mar 2016 08:16:57 +0100 Message-ID: <20160301071657.GA32453@lst.de> References: <56C0A6C3.3010903@redhat.com> <1456630639.19657.47.camel@haakon3.risingtidesystems.com> <1456647963.19657.135.camel@haakon3.risingtidesystems.com> <56D31CC9.7000609@sandisk.com> <1456692183.19657.151.camel@haakon3.risingtidesystems.com> <56D392D4.2000105@sandisk.com> <20160229060504.GA6024@lst.de> <1456814998.19657.223.camel@haakon3.risingtidesystems.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Content-Disposition: inline In-Reply-To: <1456814998.19657.223.camel@haakon3.risingtidesystems.com> Sender: target-devel-owner@vger.kernel.org To: "Nicholas A. Bellinger" Cc: Christoph Hellwig , Bart Van Assche , Doug Ledford , linux-rdma , target-devel List-Id: linux-rdma@vger.kernel.org On Mon, Feb 29, 2016 at 10:49:58PM -0800, Nicholas A. Bellinger wrote: > Btw, is original ib_create_cq() usage a incompatible with chain RDMA > READ/WRITE requests, or was that an extra improvement..? Old-style CQs can be used for chained requests, see the iser target for an example. It's just a lot more painful.