Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752640AbdF0IZ6 (ORCPT ); Tue, 27 Jun 2017 04:25:58 -0400 Received: from sci-ig2.spreadtrum.com ([222.66.158.135]:20965 "EHLO SHSQR01.spreadtrum.com" rhost-flags-OK-FAIL-OK-OK) by vger.kernel.org with ESMTP id S1752498AbdF0IZO (ORCPT ); Tue, 27 Jun 2017 04:25:14 -0400 Date: Tue, 27 Jun 2017 16:21:33 +0800 From: Baolin Wang To: Rob Herring CC: , , , , , , Subject: Re: [PATCH v4 1/3] pinctrl: Add sleep related configuration Message-ID: <20170627082132.GB32447@spreadtrum.com> Mail-Followup-To: Rob Herring , linus.walleij@linaro.org, mark.rutland@arm.com, linux-gpio@vger.kernel.org, devicetree@vger.kernel.org, linux-kernel@vger.kernel.org, broonie@kernel.org, baolin.wang@linaro.org References: <62a67526907157d41235403bd05af309fd4db16b.1498045588.git.baolin.wang@spreadtrum.com> <20170626161348.vrtlpat5eof5aa2c@rob-hp-laptop> MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Disposition: inline Content-Transfer-Encoding: 8bit In-Reply-To: <20170626161348.vrtlpat5eof5aa2c@rob-hp-laptop> User-Agent: Mutt/1.5.21 (2010-09-15) X-MAIL: SHSQR01.spreadtrum.com v5R8OLF7030919 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 3235 Lines: 64 On 一, 6月 26, 2017 at 11:13:48上午 -0500, Rob Herring wrote: > On Wed, Jun 21, 2017 at 07:55:37PM +0800, Baolin Wang wrote: > > In some scenarios, we should set some pins as input/output/pullup/pulldown > > when the specified system goes into deep sleep mode, then when the system > > goes into deep sleep mode, these pins will be set automatically by hardware. > > > > Usually we can set the "sleep" state to set sleep related config, but one SoC > > usually has not only one system (especially for mobile SoC), some systems on > > the SoC which did not run linux kernel, they can not select the "sleep" state > > when they go into deep sleep mode. > > The wording here is not very clear. I think what you mean is some pins > are not controlled by any specific driver in the OS, but need to be > controlled when entering sleep mode. Yes, that is what I meaning, sorry for confusing. > > > Thus we introduce some sleep related config into pinconf-generic for users to > > configure. > > > > Signed-off-by: Baolin Wang > > --- > > - Add this patch since v4. > > --- > > .../bindings/pinctrl/pinctrl-bindings.txt | 12 ++++++++++++ > > drivers/pinctrl/pinconf-generic.c | 10 ++++++++++ > > include/linux/pinctrl/pinconf-generic.h | 14 ++++++++++++++ > > 3 files changed, 36 insertions(+) > > > > diff --git a/Documentation/devicetree/bindings/pinctrl/pinctrl-bindings.txt b/Documentation/devicetree/bindings/pinctrl/pinctrl-bindings.txt > > index bf3f7b0..e098059 100644 > > --- a/Documentation/devicetree/bindings/pinctrl/pinctrl-bindings.txt > > +++ b/Documentation/devicetree/bindings/pinctrl/pinctrl-bindings.txt > > @@ -236,6 +236,18 @@ low-power-enable - enable low power mode > > low-power-disable - disable low power mode > > output-low - set the pin to output mode with low level > > output-high - set the pin to output mode with high level > > +sleep-bias-pull-up - pull up the pin when the specified system goes into > > + deep sleep mode > > +sleep-bias-pull-down - pull down the pin when the specified system goes into > > + deep sleep mode > > +sleep-input-enable - enable input on pin when the specified system goes > > + into deep sleep mode (no effect on output) > > +sleep-intput-disable - disable input on pin when the specified system goes > > + into deep sleep mode (no effect on output) > > +sleep-output-low - set the pin to output mode with low level when the > > + specified system goes into deep sleep mode > > +sleep-output-high - set the pin to output mode with high level when the > > + specified system goes into deep sleep mode > > slew-rate - set the slew rate > > I don't really like having 2 ways to define pin setup and this doesn't > scale if I need to define 3 states. Couldn't we create pin state > definitions and have a pinctrl-n property within the pin controller > node to handle all the unhandled pins? As LinusW also suggest we can create one "sleep" state and program them into registers at early point (like: after probing pinctrl driver). So I think I can introduce one called "early-sleep" state which need select it after initializing pinctrl driver. > > Rob