From mboxrd@z Thu Jan 1 00:00:00 1970 From: Dmitry Torokhov Subject: Re: [PATCH 2/6] Input: pm8xxx-vib: sync device tree bindings doc with the driver Date: Sat, 1 Apr 2017 11:06:07 -0700 Message-ID: <20170401180607.GM17130@dtor-ws> References: <20170331161538.11657-1-damien.riegel@savoirfairelinux.com> <20170331161538.11657-2-damien.riegel@savoirfairelinux.com> <20170401165409.GE17130@dtor-ws> <20170401175127.3nt5gujkvgyoh2tc@workotop.localdomain> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Received: from mail-pf0-f196.google.com ([209.85.192.196]:36605 "EHLO mail-pf0-f196.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751957AbdDASGL (ORCPT ); Sat, 1 Apr 2017 14:06:11 -0400 Received: by mail-pf0-f196.google.com with SMTP id r137so4571731pfr.3 for ; Sat, 01 Apr 2017 11:06:10 -0700 (PDT) Content-Disposition: inline In-Reply-To: <20170401175127.3nt5gujkvgyoh2tc@workotop.localdomain> Sender: linux-input-owner@vger.kernel.org List-Id: linux-input@vger.kernel.org To: Damien Riegel Cc: linux-input@vger.kernel.org, Rob Herring , Mark Rutland , kernel@savoirfairelinux.com On Sat, Apr 01, 2017 at 01:51:27PM -0400, Damien Riegel wrote: > On Sat, Apr 01, 2017 at 09:54:09AM -0700, Dmitry Torokhov wrote: > > On Fri, Mar 31, 2017 at 12:15:34PM -0400, Damien Riegel wrote: > > > The driver uses a hardcoded value for the register, so the parameter set > > > in the device tree is not actually used. > > > > > > Cc: Rob Herring > > > Cc: Mark Rutland > > > Signed-off-by: Damien Riegel > > > --- > > > Documentation/devicetree/bindings/input/qcom,pm8xxx-vib.txt | 3 ++- > > > 1 file changed, 2 insertions(+), 1 deletion(-) > > > > > > diff --git a/Documentation/devicetree/bindings/input/qcom,pm8xxx-vib.txt b/Documentation/devicetree/bindings/input/qcom,pm8xxx-vib.txt > > > index 4ed467b1e402..86ce95fc6cf8 100644 > > > --- a/Documentation/devicetree/bindings/input/qcom,pm8xxx-vib.txt > > > +++ b/Documentation/devicetree/bindings/input/qcom,pm8xxx-vib.txt > > > @@ -12,7 +12,8 @@ PROPERTIES > > > - reg: > > > Usage: required > > > Value type: > > > - Definition: address of vibration control register > > > + Definition: address of vibration control register. This value is > > > + actually ignored and hardcoded in the driver to value 0x4a > > > > I do not think we need to commit that the value is hard coded, it is > > implementation detail of current version of Linux driver, whereas DT > > bindings should be independent of OS as much as reasonably possible. > > > > Also, I think you can change the code to actually read and use it from > > DT to support your other device use case. > > I was hesitant to do that because that might break stuff for people who > use a device tree with reg != 0x4a, but if you tell me that's okay I'll > send a v2 that reads device tree for all pm8xxx vibrators. Actually, I was looking at the rest of the code and I now I wonder if we should be doing this for any of the devices. The registers are chip-specific and we get chip data from compatible string, so the driver is fine to simply use static mappings. It is only if we start using a common compatible string for different chips we would need to start parsing DT data. Thanks. -- Dmitry