Re: [PATCH net 4/7] net: dsa: mt7530: set both CPU port interfaces to PHY_INTERFACE_MODE_NA

[Arınç ÜNAL]

On 28/03/2023 14:20, Vladimir Oltean wrote:
> On Tue, Mar 28, 2023 at 12:57:57AM +0300, Arınç ÜNAL wrote:
> > I don't appreciate your consistent use of the word "abuse" on my patches.
>
> Consistent would mean that, when given the same kind of input, I respond
> with the same kind of output. I'm thinking you'd want a reviewer to do that?

Of course.

> Last time I said: "It's best not to abuse the net.git tree with non-bugfix patches."
> https://patchwork.kernel.org/project/netdevbpf/patch/20230307220328.11186-1-arinc.unal@xxxxxxxxxx/
>
> If anything, Jakub was/is slightly inconsistent by accepting those previous
> non-bugfix patches to the net.git tree, and then agreeing with me. He probably
> did that thinking it wasn't a hill worth dying on, which I can agree with.
> But I'm afraid that this didn't help you realize that yes, maybe you really
> are abusing the process by submitting exclusively non-bugfix commits to the
> net tree. There's a fine balance between trying to be nice and trying not to
> transmit the wrong message.
>
> It would be good if you could clarify your objection regarding my consistent
> use of the word "abuse" on your patches.
>
> There is a document at Documentation/process/stable-kernel-rules.rst
> which I remember having shared with you before, where there are some
> indications as to what constitutes a legitimate candidate for "stable"
> and what does not.

I forgot this existed, sorry about that. I had this thought left in my 
mind that "any changes that are not new features must go to the net 
tree", which clearly is not the case. I see what you mean now. None of 
my patches on the series satisfy all of the rules specified on the document.

I just think your response could've been less harsh considering I didn't 
intentionally do this. Anyway, it's all resolved now so let's not drag 
this further.

> > I'm by no means a senior C programmer. I'm doing my best to correct the
> > driver.
> >
> > Thank you for explaining the process of phylink with DSA, I will adjust my
> > patches accordingly.
> >
> > I suggest you don't take my patches seriously for a while, until I know
> > better.
>
> Whether you're a junior or a senior C programmer is entirely irrelevant
> here. I have no choice but to take your patches seriously unless otherwise
> specified, in the commit message, cover letter, or by marking them as
> RFC/RFT (but even then, their intention must be very clearly specified,
> so that I know what to comment on, or test).
>
> I don't think you really want what you're asking for, which is for
> people to not take your patches seriously. I recommend forming a smaller
> community of people which does preliminary patch review and discusses
> issues around the hardware you're working on, prior to upstream submission.
> That would, at least, be more productive.

Yes, of course. I'm actually planning something similar that involves 
OpenWrt, thanks.

Arınç