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;
Sender: Guenter Roeck <groeck7@gmail.com>
Message-ID: <820cef00-4768-46ae-c5a5-ea7c0dff71c5@roeck-us.net>
Date:   Wed, 31 May 2023 09:46:21 -0700
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101
 Thunderbird/102.10.0
Subject: Re: [PATCH] hwmon: (oxp-sensors) Add tt_toggle attribute on supported
 boards
Content-Language: en-US
To:     Joaquin Aramendia <samsagax@gmail.com>
Cc:     derekjohn.clark@gmail.com, jdelvare@suse.com,
        linux-hwmon@vger.kernel.org, linux-kernel@vger.kernel.org
References: <20230527012206.133464-1-samsagax@gmail.com>
 <8ae3e2ad-27ff-4339-88d4-504c2f59e501@roeck-us.net>
 <CABgtM3h8DXs0swGQth=dcE3J_W8k8iejvfFgjVSm9nKbRmxHDQ@mail.gmail.com>
From:   Guenter Roeck <linux@roeck-us.net>
In-Reply-To: <CABgtM3h8DXs0swGQth=dcE3J_W8k8iejvfFgjVSm9nKbRmxHDQ@mail.gmail.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
Precedence: bulk

On 5/31/23 05:04, Joaquin Aramendia wrote:
>> This attribute is a no-go. It is not even remotely related to hardware
>> monitoring, and thus must not be attached to the hwmon device.
>>
>> I don't know exactly where it belongs, but it appears to be related
>> to the keyboard. Its natural place therefore seems to be a keyboard driver.
>> We could possibly also attach it to the platform device, but there would
>> have to be some precedence of other drivers doing the same. Question
>> in that case though would be if this is just the first of many attributes
>> to come. If so, we would need to find a different solution.
>>
>> Guenter
> 
> Sure! Should this driver with those changes go into a platform driver?

You are attaching the attribute to the hwmon device. Moving the driver
to another directory would just be an attempt to avoid review by a
hwmon maintainer. Just for that reason I would NACK such an attempt.

Guenter

> Seems a better fit to me. The case against keyboard driver is the
> switch changes behaviour of the key but both the behaviour with the
> switch on and off is device defined. Some use the key as part of an AT
> Translated Keyboard and others just operate on the EC itself to grab
> the fan and set a special TDP for "Silent mode".
> For now this is the first such attribute found by the community and
> some talks with the manufacturer but it doesn't mean there wouldn't be
> others. Specially with such new form factors adding some "control
> panels" on Windows to set some hardware behaviour via EC writes. My
> goal is to allow those same functions to be available to linux users
> in a way some other userspace tools can serve as front ends.
> 
> Would taking this same driver to the platform side be a solution to
> that going forward? It would be a combination of hwmon monitoring
> attributes and some other special functions with custom attributes.
> Seems a better fit to me.