Re: [PATCH v2 2/3] clk: qcom: dispcc-eliza: Add Eliza display clock controller support

From: Krzysztof Kozlowski

Date: Wed Mar 18 2026 - 11:42:45 EST


On 18/03/2026 16:05, Dmitry Baryshkov wrote:
>>>
>>> Then please follow the internal company guidelines as outlined in the
>>> legal&marketing documents.
>>
>> That's not your task to instruct people what internal stuff should they
>> follow or not.
>
> Well... For me it's not different from your comments telling other
> submitters to follow "internal guidelines" when submitting patches. Or
> not to follow them.

I am telling them to use go/upstream, only because they do not follow
public guidelines, so we created a streamlined version for them.

It is only convenience and all of the internal guidelines are based
purely on upstream rules. They do not introduce anything else, no new
rules, no weird copyright rewriting rules.

I understand however how my comments could be misleading and someone
might think I recommend following some corporate legal stuff. Thank you
for bringing it up.

I will indeed change my comments and say they need to read Linux kernel
docs.


>
> I don't want to argue about the corporate guidelines. If you think they
> are incorrect, please change them.

No. You brought them. I did not want to discuss them, I did not say they
are right or wrong. You brought them as review so you must defend them.

Otherwise do not bring arguments you are not willing to discuss and
reply "please change them [corpo rules]".

>
>>
>> Especially not implied by previous comment "Then".
>>
>>>
>>> JFYI, several other Qualcomm maintainers also enforce use of copyright
>>> headers for Qualcomm-provided patches. Konrad is not unique here.
>>
>> I already objected to one of them, so I know.
>>
>> You do understand that this is completely broken review process? As
>> every contributor, I can object to that comment with arguments (and I
>> did in the past), however you as reviewer do not bring any
>> counter-arguments for that all. You just refer "follow legal internal
>> stuff". No, this does not work for that.
>>
>> If you bring review comment you must be able to justify it, when it is
>> being discussed. You cannot refer "but legal team said".
>
> If you want to put it that way, sure. As a Qualcomm employee you have a
> set of internal rules you have to adhere to. One of them is this

In the context of discussing code and contributions, it's none of your
business where do I work, what is my employment status and what rules my
employment or lack of that implies.

> copyright string. I'd rather not have legal department pre-review all
> our contributions. Been there (in another company), thanks, but no.

Again, nothing related here. If these are your feelings, fine, but do
not impose some rules on the community because of that. Whatever you
have with legal department, please keep for yourself.

>
> In my opinion, the maintainers and reviewers should ensure correctness
> of the patch. Correct legal header is one of those. Consider someone

And the patch is perfectly correct. I carried over original copyrights
which is the correct thing to do.

Please bring any document from upstream Linux kernel suggesting it is
not correct.

> submitting patches which has copyright strings such as "(c) qwalkomm" or
> "(c) lunix foundacion". They would be questioned for correctness.

Patch hash perfectly correct copyright statements, carried over from
code I was using.

The rule is rather that we do not touch copyright statements.

> Likewise when somebody from Qualcomm submits a patch with "(c) QuIC",
> they were asked to change it to the current form. It doesn't concern

That's not the case here.

> non-Qualcomm employees, because they cannot change the copyright of the
> material.
>
>>
>> Otherwise look for comments for your contributions where you are going
>> to receive review "please remove all boilerplate because my legal team
>> told me that and I am not going to provide actual arguments why".
>
> In this case there is one. "Because I assume that you have a requirement
> to do so from your company". If I were reviewing patches for e.g.

You have no clue what are my requirements from whatever there is my
company and it is none of your business to even suggest or decide that I
have.

Don't make assumptions about other people's employment, regardless
whether the actual guess is correct or not, because my contract and my
employment is only my business.

> Mediatek driver, if I knew the guideline for the patches and if I saw
> any of the guidelines to be breached, I'd have reacted in exactly the
> same way.

Then don't. I have no clue what rules internally Mediatek has thus I
absolutely should never say that one has to follow some internal
copyright rules. Neither should you.

Of course you can comment about public community rules and guidelines
about such things, but that is not the case here or in your example from
Mediatek.

Best regards,
Krzysztof