Re: [PATCH] net: macb: rate limit netdev error info print in the data path

["Théo Lebrun" <theo.lebrun@bootlin.com>]

Hello Zijin,

Please use interleaved posting for replies, not top posting.

https://docs.kernel.org/process/submitting-patches.html#use-trimmed-interleaved-replies-in-email-discussions

On Tue Sep 29, 2026 at 5:56 AM CEST, Zijin wrote:
> Thank you very much for your kindly reply.
>
> For all your concerns in this mail thread and [0][1], I try to give
> my answers here and hope it does not miss anything:
>
> 1. Yes, I missed the prefix [PATCH net], I will add it in my next
> commit; I would like to send it again as the 1st version
> since I missed this needed part [PATCH net];

If patch stays the same it is OK, if it changes (for example because you
add trailers) then you must iterate onto v2.

> 2. For the To/Cc list, I will use `git send-email --cc-cmd=scripts/get_maintainer.pl` as you indicated;
>
> 3. This whole patch is  NOT LLM generated, it's really a bug hit in practice, the "PHYT0036:00" 
> device name is the real name that showed in our console output, I can't just imagine or made this name and it's really of out of my imagination.
>  I think it may be related to a vendor with some of the PCI addr but I think it may not be polite to say it directly. 
> Actually hallucinating a kernel log even a kernel patch is out of my knowledge, if you know to how and 
> wish to tell me, I think it should be really interesting!

No I don't encourage you to hallucinate! It's what LLMs do sometimes
when you ask one to generate a commit message for a bug fix. They have
seen tons of commits with logs so they reproduce.

[...]

> 5. For the severity tag and should this carry a Fixes tag and name the
> intended tree,  please give out a clear indication from your
> experiences for what I described above. I really have not too much
> knowledge about that.

Yes. You put those two trailers:

   Fixes: 89e5785fc8a6 ("[PATCH] Atmel MACB ethernet driver")
   Cc: stable@vger.kernel.org

See documentation link below (or kernel git log for many examples).

In your case you want the oldest commit that introduced one of those
logs. The macb_start_xmit() one for example got introduced with the
first driver commit, so that's what you target.

https://docs.kernel.org/process/submitting-patches.html#describe-your-changes

Thanks,

--
Théo Lebrun, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com