Return-Path: Received: from mail-wm0-f66.google.com ([74.125.82.66]:36558 "EHLO mail-wm0-f66.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751101AbcDZUuL (ORCPT ); Tue, 26 Apr 2016 16:50:11 -0400 Subject: Re: [PATCH v2 17/18] xprtrdma: Remove ro_unmap() from all registration modes To: Chuck Lever References: <20160425185956.3566.64142.stgit@manet.1015granger.net> <20160425192307.3566.96703.stgit@manet.1015granger.net> <571FCF8E.1060304@grimberg.me> <2AB2B747-1D14-4885-9CEA-23104AC2A379@oracle.com> Cc: linux-rdma@vger.kernel.org, Linux NFS Mailing List From: Sagi Grimberg Message-ID: <571FD480.80100@grimberg.me> Date: Tue, 26 Apr 2016 23:50:08 +0300 MIME-Version: 1.0 In-Reply-To: <2AB2B747-1D14-4885-9CEA-23104AC2A379@oracle.com> Content-Type: text/plain; charset=windows-1252; format=flowed Sender: linux-nfs-owner@vger.kernel.org List-ID: >> Now we have ro_unmap_safe() which sorta implies that there >> is a non-safe unmap? Why not just keep it ro_umap? > > Er, I would find that confusing too. > > How about name it ro_unmap_slow(), but it's for asynchronous > cases too. Heh, if a future nfs-rdma developer sees ro_unmap_slow() I assume his first action is to grep ro_unmap_fast... Anyway, it's not critical, your call..