Re: [PATCH 0/5] rtw_sdio_if1_init cleanup and small logic tweak

[Konstantin Ryabitsev <mricon@kernel.org>]

On Wed, Apr 01, 2026 at 11:46:59AM +0200, Luka Gejak wrote:
> You're right, I used a tool to help format the response because 
> English is not my native language and I wanted the review to be clear.
> I see now that it made the response look like a bot report and I 
> apologize for that. I'll stick to writing reviews manually going 
> forward. However the technical issues I pointed out (like the inverted
> _SUCCESS/_FAIL logic in the staging headers and the uninitialized 
> pnetdev pointer) are real regressions I found while auditing the code 
> on my local tree. I'll make sure future feedback is direct and clearly
> identified if I use any tooling.

If you want to try something out, there is a new feature in b4 that allows
integrating agent reviews in your response in a format that would be better
received by kernel developers:

https://b4.docs.kernel.org/en/latest/reviewer/getting-started.html

HOWEVER, it does not remove the need to verify what the agent wrote -- LLMs
are *routinely* wrong.

Review and include only those comments that you have verified and are sure
about -- never anything that "looks valid," because you are likely to just
waste everyone's time. If you aren't absolutely sure that the agent's findings
are correct, do not include them, no matter the temptation.

Best regards,
-- 
KR