Re: [PATCH 1/7] gpiolib: gpiolib-of: Implement device tree gpio-names based lookup - Chen-Yu Tsai

From: Chen-Yu Tsai <wens@csie.org>
To: Alexandre Courbot <gnurou@gmail.com>
Cc: Maxime Ripard <maxime.ripard@free-electrons.com>,
	Linus Walleij <linus.walleij@linaro.org>,
	Johannes Berg <johannes@sipsolutions.net>,
	"John W. Linville" <linville@tuxdriver.com>,
	Arnd Bergmann <arnd@arndb.de>,
	Heikki Krogerus <heikki.krogerus@linux.intel.com>,
	Mika Westerberg <mika.westerberg@linux.intel.com>,
	Stephen Warren <swarren@wwwdotorg.org>,
	"linux-gpio@vger.kernel.org" <linux-gpio@vger.kernel.org>,
	linux-wireless <linux-wireless@vger.kernel.org>,
	netdev <netdev@vger.kernel.org>,
	"linux-arm-kernel@lists.infradead.org" 
	<linux-arm-kernel@lists.infradead.org>,
	"devicetree@vger.kernel.org" <devicetree@vger.kernel.org>,
	Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
	linux-sunxi <linux-sunxi@googlegroups.com>
Subject: Re: [PATCH 1/7] gpiolib: gpiolib-of: Implement device tree gpio-names based lookup
Date: Wed, 16 Apr 2014 17:56:54 +0800	[thread overview]
Message-ID: <CAGb2v65PF=i3pL58JycJLyp-TjsEeVxO-uDuZ6xrEob5oJLj5g@mail.gmail.com> (raw)
In-Reply-To: <CAAVeFu+rJXH=Uat5_qWPu6OfmDAFxHj2b+n=OSmwtSw4s+fv6w@mail.gmail.com>

Hi,

On Wed, Apr 16, 2014 at 3:06 PM, Alexandre Courbot <gnurou@gmail.com> wrote:
> On Wed, Apr 16, 2014 at 3:12 PM, Alexandre Courbot <gnurou@gmail.com> wrote:
>> On Tue, Apr 15, 2014 at 11:20 PM, Maxime Ripard
>> <maxime.ripard@free-electrons.com> wrote:
>>> Hi Chen-Yu,
>>>
>>> On Tue, Apr 15, 2014 at 02:41:35PM +0800, Chen-Yu Tsai wrote:
>>>> This patch provides of_get_gpiod_flags_by_name(), which looks up GPIO
>>>> phandles by name only, through gpios/gpio-names, and not by index.
>>>
>>> IIRC, gpios only uses the *-gpios properties, and not gpios/gpio-names
>>> pattern seen on various other things.
>>>
>>> Is it some new property you introduce? If so, please add it to the
>>> documentation.
>>>
>>> Now, I'm not sure that having two distinct representations of GPIOs in
>>> the DT is a good thing. Yes, it's looking odd compared to other
>>> similar bindings, but it's what we have to deal with.
>>
>> Mmmm I *think* I somehow remember a discussion about this topic
>> recently, but I cannot find it. Maybe Chen-yu could point us to the
>> conclusion of this discussion and the rationale for (re)implementing
>> named GPIOs this way?
>
> Aha, here maybe:
>
> https://lkml.org/lkml/2014/1/21/164

They're also mentioned in:

https://lkml.org/lkml/2014/2/25/581

> However I don't see a clear conclusion that we should implement that
> scheme. Not that I am strongly against it, but I'd like to see a
> practical purpose for it.

Again no clear conclusion on this. I wrote this as it was one possible
way out of the index-based GPIO stuff.

Hopefully others will chime in and we can decide whether this is what
we want or not.

Cheers
ChenYu