linux-input.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Dmitry Torokhov <dmitry.torokhov@gmail.com>
To: Artur Rojek <contact@artur-rojek.eu>
Cc: Chris Morgan <macroalpha82@gmail.com>,
	linux-input@vger.kernel.org, devicetree@vger.kernel.org,
	conor+dt@kernel.org, krzysztof.kozlowski+dt@linaro.org,
	robh+dt@kernel.org, Chris Morgan <macromorgan@hotmail.com>,
	Paul Cercueil <paul@crapouillou.net>
Subject: Re: [RFC] dt-bindings: input: Clarify that abs_min must be less than abs_max
Date: Tue, 19 Dec 2023 12:32:44 -0800	[thread overview]
Message-ID: <ZYH97KVDO4lFsbmi@google.com> (raw)
In-Reply-To: <03a9a56362b0559234d4a21a4de3e32e@artur-rojek.eu>

On Fri, Dec 15, 2023 at 12:19:51PM +0100, Artur Rojek wrote:
> On 2023-12-15 03:40, Chris Morgan wrote:
> > From: Chris Morgan <macromorgan@hotmail.com>
> > 
> > uinput refuses to work with abs devices where the min value is greater
> > than the max value. uinput_validate_absinfo() returns -EINVAL if this
> > is the case and prevents using uinput on such a device. Since uinput
> > has worked this way since at least kernel 2.6 (or prior) I presume that
> > this is the correct way of doing things, and that this documentation
> > needs to be clarified that min must always be less than max.
> > 
> > uinput is used in my use case to bind together adc-joystick devices
> > with gpio-keys devices to create a single unified gamepad for
> > userspace.
> > 
> > Note that there are several boards that will need to be corrected,
> > all but a few of them I maintain. Submitting as an RFC for now to get
> > comments from the input team and the original author in case there is
> > something I am missing.
> > 
> > Fixes: 7956b0d4694f ("dt-bindings: input: Add docs for ADC driven
> > joystick")
> > 
> > Signed-off-by: Chris Morgan <macromorgan@hotmail.com>
> > ---
> >  Documentation/devicetree/bindings/input/adc-joystick.yaml | 5 +++--
> >  1 file changed, 3 insertions(+), 2 deletions(-)
> > 
> > diff --git a/Documentation/devicetree/bindings/input/adc-joystick.yaml
> > b/Documentation/devicetree/bindings/input/adc-joystick.yaml
> > index 6c244d66f8ce..8f5cdd5ef190 100644
> > --- a/Documentation/devicetree/bindings/input/adc-joystick.yaml
> > +++ b/Documentation/devicetree/bindings/input/adc-joystick.yaml
> > @@ -73,8 +73,9 @@ patternProperties:
> >          description: >
> >            Minimum and maximum values produced by the axis.
> >            For an ABS_X axis this will be the left-most and right-most
> > -          inclination of the joystick. If min > max, it is left to
> > userspace to
> > -          treat the axis as inverted.
> > +          inclination of the joystick. The axis must always be
> > expressed as
> > +          min < max, if the axis is inverted it is left to userspace to
> > handle
> > +          the inversion.
> 
> Hi Chris,
> 
> Device Tree is supposed to depict the actual state of the hardware.
> I worded the adc-joytick's adc-range property specifically, so that it
> covers a case of GCW Zero hardware [1], which has a joystick,  where the
> ABS_X axis reports increasing values for the left-wards inclination of
> the joystick, and decreasing values for the right-wards inclination. You
> are saying that there are even more boards that need to be corrected -
> those are all situations, where DT depicts the actual behavior of the
> hardware.
> What you are trying to do is change hardware description, because of how
> a driver in an OS works. You should instead fix behavior of said driver,
> even if nobody stumbled upon that issue since 2.6 :) We fixed libSDL [2]
> for the same reason.

We have several places in the kernel (such as mousedev and joydev) where
we expect that max is greater or equal to min if they are specified. I
am sure that at least some userspace components also have this
assumption. In general, we expect min to be a minimum value reported and
max being maximum value reported, and orientation expressed via
different properties (see [1]).

Since we codified min > max as inversion for adc-joystick devices in the
bindings, I think we need to handle this *in that driver* and leave the
rest alone.

[1] Documentation/devicetree/bindings/input/touchscreen/touchscreen.yaml

Thanks.

-- 
Dmitry

  parent reply	other threads:[~2023-12-19 20:32 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-12-15  2:40 [RFC] dt-bindings: input: Clarify that abs_min must be less than abs_max Chris Morgan
2023-12-15 11:19 ` Artur Rojek
2023-12-18 17:11   ` Chris Morgan
2023-12-19 20:32   ` Dmitry Torokhov [this message]
2023-12-19 20:34     ` Dmitry Torokhov
2023-12-19 23:48       ` Peter Hutterer
2023-12-20  0:23     ` Paul Cercueil
2023-12-20  1:22       ` Dmitry Torokhov

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=ZYH97KVDO4lFsbmi@google.com \
    --to=dmitry.torokhov@gmail.com \
    --cc=conor+dt@kernel.org \
    --cc=contact@artur-rojek.eu \
    --cc=devicetree@vger.kernel.org \
    --cc=krzysztof.kozlowski+dt@linaro.org \
    --cc=linux-input@vger.kernel.org \
    --cc=macroalpha82@gmail.com \
    --cc=macromorgan@hotmail.com \
    --cc=paul@crapouillou.net \
    --cc=robh+dt@kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).