Re: [PATCH v2 1/4] gpio: aspeed-sgpio: propagate IRQ setup error in probe

From: shao.mingyin

Date: Mon Nov 16 2026 - 04:22:36 EST


Hi Krzysztof,

Thank you for the direct reply. You are right about the patch, and you
are right about the wider pattern. I was wrong on both, and I am
responsible for them.

I cannot show how the parent interrupt controller could defer probing,
and stable was not justified for this change. I should have established
that before sending it. I am withdrawing it (v2 1/4); there will be no
v3.

As I wrote in my reply to Gregory Price:

https://lore.kernel.org/all/20261115051526117H0NSID837XMXH6qUeh_JS@xxxxxxxxxx/

I sent too many patches, across too many subsystems, faster than I could
properly judge them. I started from the belief that finding defects with
an LLM and sending the patches was a contribution to the community, and
I did not weigh the review time every submission takes from maintainers,
including the ones that are rejected. The mm patches were the clearest
example: they were noise, and the cost to you and the other maintainers
was real. I used an LLM to help find and draft these submissions; I was
the one who decided to send them, so the responsibility is mine.

I will not send another patch until I can state a reachable trigger and
a user-visible effect, and I will hold myself to that bar. I do want to
contribute to the Linux community, and I understand that this means
reducing the review burden I create, not adding to it. I am sorry for
the time this has cost you, and I will not waste maintainers' time
again.

Thanks,
Shao