linux-kernel - Re: [PATCH v4 2/7] dt-bindings: ti-lmu: Modify dt bindings for the LM3697

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <ade3121a-58ab-c0c3-6605-cf7d6a638da0@gmail.com>
Date:   Thu, 25 Oct 2018 20:27:18 +0200
From:   Jacek Anaszewski <jacek.anaszewski@...il.com>
To:     Dan Murphy <dmurphy@...com>, Rob Herring <robh@...nel.org>
Cc:     Pavel Machek <pavel@....cz>, devicetree@...r.kernel.org,
        linux-kernel@...r.kernel.org, linux-leds@...r.kernel.org,
        lee.jones@...aro.org, tony@...mide.com
Subject: Re: [PATCH v4 2/7] dt-bindings: ti-lmu: Modify dt bindings for the
 LM3697

On 10/25/2018 08:07 PM, Dan Murphy wrote:
> Rob
> 
> On 10/24/2018 09:54 AM, Rob Herring wrote:
>> On Wed, Oct 24, 2018 at 07:07:57AM -0500, Dan Murphy wrote:
>>> Pavel
>>>
>>> On 10/24/2018 04:04 AM, Pavel Machek wrote:
>>>> Hi!
>>>>
>>>>> The LM3697 is a single function LED driver. The single function LED
>>>>> driver needs to reside in the LED directory as a dedicated LED driver
>>>>> and not as a MFD device.  The device does have common brightness and ramp
>>>>
>>>> So it is single function LED driver. That does not mean it can not
>>>> share bindings with the rest. Where the bindings live is not imporant.
>>>>
>>>
>>> It can share bindings that are correctly done, not ones that are incomplete and incorrect.
>>>
>>> Where bindings live is important to new Linux kernel developers and product 
>>> developers looking for the proper documentation on the H/W bindings.
>>>
>>>>> reside in the Documentation/devicetree/bindings/leds directory and follow the
>>>>> current LED and general bindings guidelines.
>>>>
>>>> What you forgot to tell us in the changelog:
>>>
>>> I can add this to the changelog.
>>>
>>>>
>>>>> +Optional child properties:
>>>>> +	- runtime-ramp-up-msec: Current ramping from one brightness level to
>>>>> +				the a higher brightness level.
>>>>> +				Range from 2048 us - 117.44 s
>>>>
>>>> The other binding uses "ramp-up-msec". Tell us why you are changing this, or
>>>> better don't change things needlessly.
>>>>
>>>> We don't want to be using "runtime-ramp-up-msec" for one device and
>>>> "ramp-up-msec" for the other.
>>>
>>> This is another example of how the original bindings were incorrect and misleading.
>>>
>>> The LM3697 have 2 ramp implementations that can be used.
>>>
>>> Startup/Shutdown ramp and Runtime Ramp.  Same Ramp rates different registers and
>>> different end user experience.
>>>
>>> So having a single node call ramp-up-msec is misleading and it does not
>>> indicate what the H/W will do.
>>
>> The existing ones aren't documented (present in the example is not 
>> documented). This seems like something that should be common rather than 
>> TI specific. Though it also seems more like something the user would 
>> want to control (i.e. sysfs) rather than fixed in DT.
>>
> 
> Changing the runtime ramping or startup/shutdown ramping could also be done via sysfs.
> I am not dedicated to having it in the DT file I was following prior art.
> 
> Jacek
> 
> Do you have an opinion on this?

This is this problem with the Device Tree's scope of responsibility.
It is defined as a means for "describing the hardware", but often
this rule is abused by the properties that fall into "configuration"
category. E.g. default-state, retain-state-suspended from leds-gpio.txt
or linux-default-trigger from common LED bindings.

In some cases this is justified. The question is whether it is something
that necessarily needs to be configured on driver probing? If not, then
I'd go for sysfs interface.

-- 
Best regards,
Jacek Anaszewski