From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from pasmtpB.tele.dk (pasmtpb.tele.dk [80.160.77.98]) by ozlabs.org (Postfix) with ESMTP id A4ABDDE297 for ; Wed, 18 Jun 2008 07:00:12 +1000 (EST) Date: Tue, 17 Jun 2008 23:00:55 +0200 From: Sam Ravnborg To: Scott Wood Subject: Re: [PATCH] [Rev2] MPC5121 FEC support Message-ID: <20080617210055.GA13468@uranus.ravnborg.org> References: <1213729717-26688-1-git-send-email-jrigby@freescale.com> <48581188.90102@freescale.com> <20080617195739.GB13147@uranus.ravnborg.org> <485818AA.8090701@freescale.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii In-Reply-To: <485818AA.8090701@freescale.com> Cc: linuxppc-dev@ozlabs.org, jeff@garzik.org, John Rigby , netdev@vger.kernel.org List-Id: Linux on PowerPC Developers Mail List List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , On Tue, Jun 17, 2008 at 03:03:54PM -0500, Scott Wood wrote: > Sam Ravnborg wrote: > >On Tue, Jun 17, 2008 at 02:33:28PM -0500, Scott Wood wrote: > >>John Rigby wrote: > >>>config FS_ENET > >>> tristate "Freescale Ethernet Driver" > >>>- depends on CPM1 || CPM2 > >>>+ depends on CPM1 || CPM2 || FS_ENET_MPC5121_FEC > >>> select MII > >>> select PHYLIB > >>> > >>>+config FS_ENET_MPC5121_FEC > >>>+ bool "Freescale MPC512x FEC driver" > >>>+ depends on PPC_MPC512x > >>>+ select FS_ENET > >>>+ select PPC_CPM_NEW_BINDING > >>>+ default y > >>No default y. > >I by the way do not see the need for the prompt of FS_ENET. > > Agreed, especially since it's overly broad (there is Freescale ethernet > hardware that this driver doesn't support). We'd need to change depends > into selects in the more specific entries. > > >Do you ever want to change it if one of the dependencies > >are selected? > > Do you mean if CPM1 or CPM2 is selected? Yes, it's quite possible that > the user has no need for the CPM ethernet and would rather reclaim the > memory (especially on CPM1, which has boards as small as 8MiB). I took a closer look now. And I can see that FS_ENET is indeed a driver and CPM1, CPM2, FS_ENET_MPC5121_FEC are all booleans that say that it may make sense to use this driver. But I was misguided by: >+config FS_ENET_MPC5121_FEC >+ select FS_ENET This is not good. In general when you select a symbol that has dependencies you are almost always on the wrong track. Use a dependency here with a sane default - then people can set it to 'n' if they really do not want this driver. Spreading selects too much is just causing you pain in the long run. Sam