Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
MIME-Version: 1.0
References: <20201019141008.871177-1-daniel@0x0f.com> <20201019141008.871177-2-daniel@0x0f.com>
 <20201026134604.GA43947@bogus>
In-Reply-To: <20201026134604.GA43947@bogus>
From:   Daniel Palmer <daniel@0x0f.com>
Date:   Tue, 27 Oct 2020 00:15:45 +0900
Message-ID: <CAFr9PXmuOCRoUiqNAdMi=xX142U46=Vnk-EHp9H02L8nut0esQ@mail.gmail.com>
Subject: Re: [PATCH v2 1/5] dt-bindings: gpio: Add a binding header for the
 MSC313 GPIO driver
To:     Rob Herring <robh@kernel.org>
Cc:     "open list:GPIO SUBSYSTEM" <linux-gpio@vger.kernel.org>,
        DTML <devicetree@vger.kernel.org>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
        Linus Walleij <linus.walleij@linaro.org>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

Hi Rob,

On Mon, 26 Oct 2020 at 22:46, Rob Herring <robh@kernel.org> wrote:
>
> On Mon, Oct 19, 2020 at 11:10:04PM +0900, Daniel Palmer wrote:
> > Header adds defines for the gpio number of each pin
> > from the driver view. The gpio block seems to support 128 lines
> > but what line is mapped to a physical pin depends on the chip.
> > The driver itself uses the index of a pin's offset in an array
> > of the possible offsets for a chip as the gpio number.
> >
> > The defines remove the need to work out that index to consume
> > a pin in the device tree.
>
> I'd expect the DT to have 0-127 numbering... If you need to map that to
> another number, then an array property in DT could handle that.
>

Thank you for the comments on this header and the binding description.

Thinking about this again I'm thinking about having the GPIO numbers
be 0-127 like you say but supplying the valid offsets for that
specific chip and the pad/pin names to make visible to the user via an
array/arrays that contains the pin register offsets and the pin names.
Basically my per-chip table moves out of the driver and into the DT.
Does that sound acceptable? The main thing I want to avoid is
presenting the user with 128 gpios when the actually chip only has <10
of them wired up.

Thanks,

Daniel