From: Chuck Lever Subject: Re: [RFC,PATCH 21/38] svc: Change svc_sock_received to svc_xprt_received and export it Date: Fri, 30 Nov 2007 18:23:08 -0500 Message-ID: <3E6C8687-0A0C-489F-B7F8-1EEFA2E01919@oracle.com> References: <20071129223917.14563.77633.stgit@dell3.ogc.int> <20071129224037.14563.69171.stgit@dell3.ogc.int> <7E88BE3B-F35C-44B1-AE84-E5DE62E4EFA5@oracle.com> <1196464634.5432.68.camel@trinity.ogc.int> Mime-Version: 1.0 (Apple Message framework v752.2) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Cc: "J. Bruce Fields" , linux-nfs@vger.kernel.org To: Tom Tucker Return-path: Received: from agminet01.oracle.com ([141.146.126.228]:39485 "EHLO agminet01.oracle.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756010AbXK3XXW (ORCPT ); Fri, 30 Nov 2007 18:23:22 -0500 In-Reply-To: <1196464634.5432.68.camel-SMNkleLxa3ZimH42XvhXlA@public.gmane.org> Sender: linux-nfs-owner@vger.kernel.org List-ID: On Nov 30, 2007, at 6:17 PM, Tom Tucker wrote: > On Fri, 2007-11-30 at 16:33 -0500, Chuck Lever wrote: >> On Nov 29, 2007, at 5:40 PM, Tom Tucker wrote: >>> { >>> - svsk->sk_xprt.xpt_pool = NULL; >>> - clear_bit(XPT_BUSY, &svsk->sk_xprt.xpt_flags); >>> - svc_xprt_enqueue(&svsk->sk_xprt); >>> + xprt->xpt_pool = NULL; >>> + clear_bit(XPT_BUSY, &xprt->xpt_flags); >>> + svc_xprt_enqueue(xprt); >>> } >>> - >>> +EXPORT_SYMBOL_GPL(svc_xprt_received); >> >> When I submitted the RPC client-side transport switch, Trond >> suggested we add the EXPORTs later when it was clear why they are >> needed. This may be a personal preference of the server maintainer, >> but I just thought I'd mention the possibility; it seems to make >> sense here too. > > Sure, but we already have a server side provider that helps accelerate > the proving process. The svcrdma module won't build without this being > exported. That's fine. You just add the EXPORTs when you introduce the svcrdma module. No biggie. -- Chuck Lever chuck[dot]lever[at]oracle[dot]com