linux-kernel - Re: [PATCH v2] dt-bindings: power: Convert bq27xxx dt to yaml

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] [day] [month] [year] [list]

Date:   Mon, 11 May 2020 14:06:33 -0500
From:   Dan Murphy <dmurphy@...com>
To:     Sebastian Reichel <sebastian.reichel@...labora.com>
CC:     <linux-pm@...r.kernel.org>, <robh@...nel.org>,
        <linux-kernel@...r.kernel.org>, <devicetree@...r.kernel.org>,
        Pali Rohár <pali@...nel.org>,
        "Andrew F . Davis" <afd@...com>
Subject: Re: [PATCH v2] dt-bindings: power: Convert bq27xxx dt to yaml

Sebastian

On 5/11/20 10:30 AM, Sebastian Reichel wrote:
> Hi,
>
> On Mon, May 11, 2020 at 09:55:11AM -0500, Dan Murphy wrote:
>> On 5/11/20 9:57 AM, Sebastian Reichel wrote:
>>> On Mon, May 11, 2020 at 09:29:59AM -0500, Dan Murphy wrote:
>>>> On 5/11/20 9:32 AM, Sebastian Reichel wrote:
>>>>> On Mon, May 11, 2020 at 07:25:06AM -0500, Dan Murphy wrote:
>>>>>> On 5/10/20 11:17 AM, Sebastian Reichel wrote:
>>>>>>> This needs is missing the power-supplies property. The N900 DT
>>>>>>> contains a bq27200 referencing the charger, so it should fail the DT
>>>>>>> check without the property being listed here.
>>>>>> Hmm.  I ran the dt checker specifically on the binding and it did not fail.
>>>>>> Unless I need to build some other DTs as well.
>>>>>> Either I will have the power-supplies property
>>>>> I just tried it myself. The problem is the way you are specifying
>>>>> the compatible strings. This is the parsing result:
>>>>>
>>>>> enum: ['ti,bq27200 - BQ27200', 'ti,bq27210 - BQ27210', 'ti,bq27500 - deprecated,
>>>>>          use revision specific property below', ...
>>>>>
>>>>> You can see this in Documentation/devicetree/bindings/processed-schema.yaml, which
>>>>> is generated by running the check. The compatible comments need a # as separation
>>>>> character like this to generate proper bindings:
>>>>>
>>>>> properties:
>>>>>      compatible:
>>>>>        enum:
>>>>>          - ti,bq27200 # BQ27200
>>>>>          - ti,bq27210 # BQ27210
>>>>>          - ti,bq27500 # deprecated, use revision specific property below
>>>> Well honestly not sure why we need the comment either. These are pretty
>>>> self explanatory maybe we should just remove the additional comments
>>> Fine with me.
>> Ack
>>>> Any consideration on just removing the deprecated values?
>>> Let's keep them with their comment for now. Removing them should
>>> start with marking them as depracated in the binding and generating
>>> a runtime warning in the driver, so that people become aware of the
>>> problem. At least for ti,bq27500 we have mainline users At least for
>>> ti,bq27500 we have mainline users.
>> There are only 2 dts files that have this reference unless we are not sure
>> which device is actually in use.
> DT is considered ABI and one is supposed to be able to boot a new
> kernel with an old DT. It's not enough to just update the in-tree
> dts files. I suppose we can consider removing support for the old
> compatible values after having the warning being printed for some
> time and the mainline users being converted to the new binding.

Yes I know. I may have said that before.

After looking at the driver and how this is all stitched together I 
think I am just going to stick to the DT conversion as is.

I will make the basic changes for conversion but any changes to the 
compatibles should be done later.

Dan


> -- Sebastian