From mboxrd@z Thu Jan 1 00:00:00 1970 From: aisheng.dong@freescale.com (Dong Aisheng) Date: Mon, 16 Apr 2012 11:35:20 +0800 Subject: [PATCH 3/3] ARM: imx6q: switch to use pinctrl driver In-Reply-To: <20120415073751.GF24997@S2101-09.ap.freescale.net> References: <1334333915-1174-1-git-send-email-b29396@freescale.com> <1334333915-1174-3-git-send-email-b29396@freescale.com> <20120414135331.GB3852@pengutronix.de> <20120415073751.GF24997@S2101-09.ap.freescale.net> Message-ID: <20120416033519.GA17173@shlinux2.ap.freescale.net> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org On Sun, Apr 15, 2012 at 03:37:54PM +0800, Shawn Guo wrote: > On Sun, Apr 15, 2012 at 12:03:16PM +0800, Dong Aisheng wrote: > ... > > For MX3 and MX5, before their pinctrl driver are available, user may > > need to define a dummy in dt to avoid such an error. > > > That is insane, and will not work. > > 1) We do not have device tree support for MX3 yet. > 2) It's not practical to add dummy entries in dt to avoid the error. > 3) There are still so many non-dt board files support in the tree. > How will they work with that? > > My suggestion is to not make it an error but a dev_dbg, assuming that > the pins are already set up by platform code if we do not find the > entry from pinctrl system. We can take it as a transition solution, > and turn the dev_dbg back to error when all the users of the device > driver are converted to device tree. > My thinking is let pinctrl core handle dummy state as regulator subsystem and does not depend on dt. Linus, Stephen, any comment? Regards Dong Aisheng From mboxrd@z Thu Jan 1 00:00:00 1970 From: Dong Aisheng Subject: Re: [PATCH 3/3] ARM: imx6q: switch to use pinctrl driver Date: Mon, 16 Apr 2012 11:35:20 +0800 Message-ID: <20120416033519.GA17173@shlinux2.ap.freescale.net> References: <1334333915-1174-1-git-send-email-b29396@freescale.com> <1334333915-1174-3-git-send-email-b29396@freescale.com> <20120414135331.GB3852@pengutronix.de> <20120415073751.GF24997@S2101-09.ap.freescale.net> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Return-path: Content-Disposition: inline In-Reply-To: <20120415073751.GF24997@S2101-09.ap.freescale.net> Sender: linux-kernel-owner@vger.kernel.org To: Shawn Guo Cc: Dong Aisheng , Sascha Hauer , linus.walleij@stericsson.com, devicetree-discuss@lists.ozlabs.org, linux-kernel@vger.kernel.org, rob.herring@calxeda.com, kernel@pengutronix.de, linux-arm-kernel@lists.infradead.org List-Id: devicetree@vger.kernel.org On Sun, Apr 15, 2012 at 03:37:54PM +0800, Shawn Guo wrote: > On Sun, Apr 15, 2012 at 12:03:16PM +0800, Dong Aisheng wrote: > ... > > For MX3 and MX5, before their pinctrl driver are available, user may > > need to define a dummy in dt to avoid such an error. > > > That is insane, and will not work. > > 1) We do not have device tree support for MX3 yet. > 2) It's not practical to add dummy entries in dt to avoid the error. > 3) There are still so many non-dt board files support in the tree. > How will they work with that? > > My suggestion is to not make it an error but a dev_dbg, assuming that > the pins are already set up by platform code if we do not find the > entry from pinctrl system. We can take it as a transition solution, > and turn the dev_dbg back to error when all the users of the device > driver are converted to device tree. > My thinking is let pinctrl core handle dummy state as regulator subsystem and does not depend on dt. Linus, Stephen, any comment? Regards Dong Aisheng From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751537Ab2DPDYr (ORCPT ); Sun, 15 Apr 2012 23:24:47 -0400 Received: from db3ehsobe006.messaging.microsoft.com ([213.199.154.144]:24000 "EHLO db3outboundpool.messaging.microsoft.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751078Ab2DPDYq (ORCPT ); Sun, 15 Apr 2012 23:24:46 -0400 X-SpamScore: -6 X-BigFish: VS-6(zcb8kz1432N98dKzz1202hzzz2dh2a8h668h839h944hd25h) X-Forefront-Antispam-Report: CIP:70.37.183.190;KIP:(null);UIP:(null);IPV:NLI;H:mail.freescale.net;RD:none;EFVD:NLI Date: Mon, 16 Apr 2012 11:35:20 +0800 From: Dong Aisheng To: Shawn Guo CC: Dong Aisheng , Sascha Hauer , , , , , , Subject: Re: [PATCH 3/3] ARM: imx6q: switch to use pinctrl driver Message-ID: <20120416033519.GA17173@shlinux2.ap.freescale.net> References: <1334333915-1174-1-git-send-email-b29396@freescale.com> <1334333915-1174-3-git-send-email-b29396@freescale.com> <20120414135331.GB3852@pengutronix.de> <20120415073751.GF24997@S2101-09.ap.freescale.net> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Disposition: inline In-Reply-To: <20120415073751.GF24997@S2101-09.ap.freescale.net> User-Agent: Mutt/1.5.20 (2009-06-14) X-OriginatorOrg: freescale.com Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Sun, Apr 15, 2012 at 03:37:54PM +0800, Shawn Guo wrote: > On Sun, Apr 15, 2012 at 12:03:16PM +0800, Dong Aisheng wrote: > ... > > For MX3 and MX5, before their pinctrl driver are available, user may > > need to define a dummy in dt to avoid such an error. > > > That is insane, and will not work. > > 1) We do not have device tree support for MX3 yet. > 2) It's not practical to add dummy entries in dt to avoid the error. > 3) There are still so many non-dt board files support in the tree. > How will they work with that? > > My suggestion is to not make it an error but a dev_dbg, assuming that > the pins are already set up by platform code if we do not find the > entry from pinctrl system. We can take it as a transition solution, > and turn the dev_dbg back to error when all the users of the device > driver are converted to device tree. > My thinking is let pinctrl core handle dummy state as regulator subsystem and does not depend on dt. Linus, Stephen, any comment? Regards Dong Aisheng