Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
MIME-Version: 1.0
References: <20220728091712.13395-1-m.zatovic1@gmail.com> <CACRpkdbD1qzJujhq-U0UN0tWam9CaoLvVuAQfafq4XNaEdZ2QA@mail.gmail.com>
 <CAPGNi94R6bW80Yc=B5iesW=ihxy_bD1p-iPhLCASh_=82mjT9w@mail.gmail.com>
In-Reply-To: <CAPGNi94R6bW80Yc=B5iesW=ihxy_bD1p-iPhLCASh_=82mjT9w@mail.gmail.com>
From:   Linus Walleij <linus.walleij@linaro.org>
Date:   Tue, 20 Sep 2022 11:44:19 +0200
Message-ID: <CACRpkdY2pGOV=PFKEaPAnKpt16HNLomRuQKRgH8A+cKFUotRXw@mail.gmail.com>
Subject: Re: [PATCH RFC v1 1/2] bus: add Wiegand write-only GPIO driver
To:     =?UTF-8?B?TWFydGluIFphxaVvdmnEjQ==?= <m.zatovic1@gmail.com>
Cc:     linux-kernel@vger.kernel.org, devicetree@vger.kernel.org,
        mani@kernel.org, hemantk@codeaurora.org, elder@linaro.org,
        f.fainelli@gmail.com, Michael.Srba@seznam.cz,
        jeffrey.l.hugo@gmail.com, gregkh@linuxfoundation.org,
        bjorn.andersson@linaro.org, saravanak@google.com,
        krzysztof.kozlowski+dt@linaro.org, robh+dt@kernel.org
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On Mon, Sep 19, 2022 at 2:57 PM Martin Za=C5=A5ovi=C4=8D <m.zatovic1@gmail.=
com> wrote:

> 1. drivers/bus/wiegand.c
> 2. include/linux/wiegand.h
> 3. drivers/gpio/gpio-wiegand.h
>
> I have looked for drivers with a similar division and tried to figure out=
 what the file drivers/bus/wiegand.c should contain.
> So far I am only sure that it will contain the parity calculation functio=
ns.

That is a question for people who know wiegand... Maybe just yourself!

You also clearly need some bus infrastructure to populate and spawn devices=
.
This mechanism depends on the specifics of the bus.

> The code for sysfs attribute files may also be a part of this file, but I=
 am not sure about it.
> If hypothetically there was another Wiegand driver besides the GPIO one a=
nd a user would
> like to connect both of these Wiegand interfaces, he would loose the abil=
ity to set attributes
> for individual Wiegand interfaces, am I right?

You should spawn individual devices on a wiegand bus (the kernel concept of
bus). Each new device becomes something like a struct wiegand_device that
contains a struct device, and the struct device is the anchor in sysfs.
https://www.kernel.org/doc/html/latest/driver-api/driver-model/bus.html

> Should the drivers/bus/wiegand.c be a full-scale driver or should it just=
 be a file that
> contains the Wiegand-specific functions(such as parity calculation)?

It should be what we call a bus driver, OK a bit self-referential, but anyt=
hing
that is common for the bus and the code to spawn the devices.

Yours,
Linus Walleij