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: <20200428195651.6793-1-mani@kernel.org> <20200428195651.6793-3-mani@kernel.org>
 <CACRpkdZ3b-VLvxN06H_4cDOtUEQTVbe=Zw+NA=YjssMzK2d2sQ@mail.gmail.com>
 <20200429124918.GC6443@Mani-XPS-13-9360> <20200519085703.GB27787@localhost>
In-Reply-To: <20200519085703.GB27787@localhost>
From:   Linus Walleij <linus.walleij@linaro.org>
Date:   Mon, 25 May 2020 10:59:59 +0200
Message-ID: <CACRpkdapMuMs_mEUHheGtaKYg97=nL1bH3zq4Tc3cnX9Jbw-Ew@mail.gmail.com>
Subject: Re: [PATCH 2/2] usb: serial: xr_serial: Add gpiochip support
To:     Johan Hovold <johan@kernel.org>
Cc:     Manivannan Sadhasivam <mani@kernel.org>,
        Greg KH <gregkh@linuxfoundation.org>,
        linux-usb <linux-usb@vger.kernel.org>,
        "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
        patong.mxl@gmail.com,
        "open list:GPIO SUBSYSTEM" <linux-gpio@vger.kernel.org>
Content-Type: text/plain; charset="UTF-8"
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Tue, May 19, 2020 at 10:57 AM Johan Hovold <johan@kernel.org> wrote:
> > On Wed, Apr 29, 2020 at 02:12:24PM +0200, Linus Walleij wrote:

> > > to something that is device-unique, like "xr-gpios-<serial number>"
> > > which makes it easy to locate the GPIOs on a specific serial converter
> > > for lab use. However the USB serial maintainers know better what
> > > to use here. Whatever makes a USB-to-serial unique from a TTY
> > > point of view is probably fine with me too.
> > >
> > > My idea is that people might want to know which USB cable
> > > this is sitting on, so I have this USB cable and from this label
> > > I can always figure out which GPIO device it is.
>
> I think we've had this discussion before. First, not every device has a
> unique serial number. Second, we already have a universal way of
> distinguishing devices namely by using the bus topology. That's
> available through sysfs and shouldn't have to be be re-encoded by every
> driver in the gpiochip name.

I remember I even referred to this myself, but I've been waning a bit
on it recently, because it turns out that userspace/users aren't very
good at parsing sysfs for topology.

For userspace other than udev there seems to be a kind of agreement
gap. Dunno how best to bridge it though. Education maybe.

Yours,
Linus Walleij