From mboxrd@z Thu Jan 1 00:00:00 1970 From: Sowmini Varadhan Subject: Re: [PATCH V2 net-next 5/7] rds: zerocopy Tx support. Date: Wed, 14 Feb 2018 19:09:33 -0500 Message-ID: <20180215000933.GU11528@oracle.com> References: <97bec2535a23b8dc976f958dc27256731de9127e.1517843755.git.sowmini.varadhan@oracle.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Network Development , David Miller , rds-devel@oss.oracle.com, Santosh Shilimkar To: Willem de Bruijn Return-path: Received: from aserp2120.oracle.com ([141.146.126.78]:44710 "EHLO aserp2120.oracle.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1032091AbeBOAKI (ORCPT ); Wed, 14 Feb 2018 19:10:08 -0500 Content-Disposition: inline In-Reply-To: Sender: netdev-owner@vger.kernel.org List-ID: On (02/14/18 18:48), Willem de Bruijn wrote: > > If the missing break is intentional, no need to respin just for the other > minor comments. yes the missing break is intentional- the function returns the size of the scatterlist needed for RDMA, and RDS_CMSG_ZCOPY_COOKIE (like RDMA_DEST and RDMA_MAP) is meta-data that does not change that size. I expect to be in the neighborhood of this code pretty soon, to get the additional opimization of passing up the zcopy completion as part of recvmsg (see the discussion in https://www.mail-archive.com/netdev@vger.kernel.org/msg212788.html) I can take care of the other code-cleanup comment suggestions in here at that time.. --Sowmini