From mboxrd@z Thu Jan 1 00:00:00 1970 From: Parav Pandit Subject: RE: [PATCH] IB/core: Change rdma_protcol_iboe to roce Date: Mon, 18 May 2015 21:45:49 +0530 Message-ID: References: <1431630106-28829-1-git-send-email-ira.weiny@intel.com> <1431702851.4214.3.camel@opteya.com> <2807E5FD2F6FDA4886F6618EAC48510E1107F5B2@CRSMSX101.amr.corp.intel.com> <1431725566.11477.15.camel@opteya.com> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: In-Reply-To: <1431725566.11477.15.camel-RlY5vtjFyJ3QT0dZR+AlfA@public.gmane.org> Sender: linux-rdma-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org To: Yann Droneaud , "Weiny, Ira" Cc: dledford-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org, linux-rdma-u79uwXL29TY76Z2rM5mHXA@public.gmane.org List-Id: linux-rdma@vger.kernel.org Hi, At present we have chosen to call it RoCE. However I believe IBoE is mo= re generic for the reason where, If the future RoCE versions may work on non-converged Ethernet, IBoE wo= uld be more appropriate name in long haul. Parav -----Original Message----- =46rom: linux-rdma-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org [mailto:linux-rdma-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org] On Behalf Of Yann Droneaud Sent: Saturday, May 16, 2015 3:03 AM To: Weiny, Ira Cc: dledford-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org; linux-rdma-u79uwXL29TY76Z2rM5mHXA@public.gmane.org Subject: Re: [PATCH] IB/core: Change rdma_protcol_iboe to roce Hi, Le vendredi 15 mai 2015 =C3=A0 16:29 +0000, Weiny, Ira a =C3=A9crit : > > Le jeudi 14 mai 2015 =C3=A0 15:01 -0400, ira.weiny-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org a =C3=A9= crit : > > > From: Ira Weiny > > > > > > It has been decided that ROCE should be used within the kernel > > > rather than IBOE as we used before. Change iboe to roce on the > > > new > > rdma_protocol_* functions. > > > > > > > Erk ... What's the usefulness of such patch > > Currently there is inconsistency in the naming of the RoCE technology= =2E > > 12:27:35 > grep -r roce drivers/infiniband | grep -v Binary | grep -i > -v proce | wc -l > 85 > > All in the driver code. > > 12:27:48 > grep -r iboe drivers/infiniband | grep -v Binary | wc -l > 130 > > Mainly in the core and mlx4 driver. > I see thing a bit differently: $ grep -ri 'roce' include/rdma/ \ include/uapi/rdma \ include/uapi/linux/if_infiniband.h \ drivers/infiniband/core | grep -vi proce | wc -l 0 $ grep -ri 'iboe' include/rdma/ \ include/uapi/rdma \ include/uapi/linux/if_infiniband.h \ drivers/infiniband/core | grep -vi proce | wc -l 30 (On next-20150515). I believe the drivers can have the names they want, especially ocrdma, Emulex OneConnect RoCE. > This patch was to clean up the management helper function as a genera= l > move toward standardizing on roce rather than iboe. Most people in > the community refer to this as "RoCE" so that name was chosen to move > to. > It's not the first time Linux use a name not matching the "vendor" one, amd64, arm64, etc. > > > > IBoE is used throughout the IB/RDMA subsystem. > > > > Changing only these occurences is rather inconsistent. > > > > I asked about changing all the references and Doug mentioned he would > make a patch to change the other references. Personally I don't want > to see a massive rename patch but this could be done. > That's my main concern: I dislike patch that change such large portion = of code, for not well defined purpose. I think "It has been decided that ROCE should be used within" is not en= ough to justify the change. But "Most people in the community refer to this as "RoCE" so that name = was chosen to move to." sound a bit better as an explanation. And now, I'm stopping bikeshedding :) Regards. -- Yann Droneaud OPTEYA -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" i= n the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org More majordomo info at http://vger.kernel.org/majordomo-info.html -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" i= n the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org More majordomo info at http://vger.kernel.org/majordomo-info.html