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;
Date:   Mon, 15 Aug 2022 13:17:54 +0000
To:     Dmitry Torokhov <dmitry.torokhov@gmail.com>
From:   Caleb Connolly <caleb@connolly.tech>
Cc:     Andy Gross <agross@kernel.org>,
        Bjorn Andersson <bjorn.andersson@linaro.org>,
        Rob Herring <robh+dt@kernel.org>,
        Krzysztof Kozlowski <krzysztof.kozlowski+dt@linaro.org>,
        linux-arm-msm@vger.kernel.org, devicetree@vger.kernel.org,
        linux-kernel@vger.kernel.org, linux-input@vger.kernel.org,
        phone-devel@vger.kernel.org
Reply-To: Caleb Connolly <caleb@connolly.tech>
Subject: Re: [PATCH 1/3] input: add event codes for user programmable switch events
Message-ID: <70be22f4-7851-200b-9741-4032033f7bfc@connolly.tech>
In-Reply-To: <YvQyOhgSGjQLHHvv@google.com>
References: <20220516142158.1612109-1-caleb@connolly.tech> <YvQyOhgSGjQLHHvv@google.com>
Feedback-ID: 10753939:user:proton
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
Precedence: bulk


On 10/08/2022 23:33, Dmitry Torokhov wrote:
> Hi Caleb,
>
> On Mon, May 16, 2022 at 02:22:57PM +0000, Caleb Connolly wrote:
>> Add SW_PROG{1,2,3,4} for device switches which are handled by userspace.
>>
>> This can be used for devices with "generic" switches which are intended
>> to be user-programmable, for example OnePlus phones contain a tri-state
>> key which can be used for switching between mute/vibrate/ring, or
>> programmed by the user to perform any arbitrary actions.
>>
>> These are analogous to the keys KEY_PROG{1,2,3,4} found on some
>> keyboards.

Hi Dmitry,
>
> This has been proposed a few times but this goes against the spirit of
> the input subsystem where each key or switch has a defined [default]
> purpose and userspace is normally expected to act upon input events
> without paying attention to what device they actually are coming from.
> IOW if you want to deal with particular GPIO signals you are better off
> using GPIO subsystem.
>
> In this context adding KEY_PROG1-4 was actually a mistake and we should
> try not extend it further.
>
> As fas as modeling the sliders with multiple settings goes I'd look into
> ABS events, probably ABS_MISC, with a set of values.

Ah I wasn't aware of ABS_MISC, this looks like a good way to go. Thanks.

>
> Thanks.
>
> --
> Dmitry

--
Kind Regards,
Caleb