Message-ID: <20f8d441-914d-48b9-85d4-c1891f44d20f@gmail.com>
Date: Wed, 22 Oct 2025 18:01:42 +0100
From: Pavel Begunkov <asml.silence@...il.com>
To: Jens Axboe <axboe@...nel.dk>, David Wei <dw@...idwei.uk>,
 io-uring@...r.kernel.org, netdev@...r.kernel.org
Cc: Jakub Kicinski <kuba@...nel.org>, Mina Almasry <almasrymina@...gle.com>
Subject: Re: [PATCH 1/1] io_uring zcrx: add MAINTAINERS entry

On 10/22/25 15:34, Jens Axboe wrote:
> On 10/22/25 8:25 AM, Pavel Begunkov wrote:
>> On 10/22/25 14:17, Jens Axboe wrote:
>>> On 10/22/25 5:38 AM, Pavel Begunkov wrote:
>>>> On 10/21/25 21:29, David Wei wrote:
>>>>> Same as [1] but also with netdev@ as an additional mailing list.
>>>>> io_uring zero copy receive is of particular interest to netdev
>>>>> participants too, given its tight integration to netdev core.
>>>>
>>>> David, I can guess why you sent it, but it doesn't address the bigger
>>>> problem on the networking side. Specifically, why patches were blocked
>>>> due to a rule that had not been voiced before and remained blocked even
>>>> after pointing this out? And why accusations against me with the same
>>>> circumstances, which I equate to defamation, were left as is without
>>>> any retraction? To avoid miscommunication, those are questions to Jakub
>>>> and specifically about the v3 of the large buffer patchset without
>>>> starting a discussion here on later revisions.
>>>>
>>>> Without that cleared, considering that compliance with the new rule
>>>> was tried and lead to no results, this behaviour can only be accounted
>>>> to malice, and it's hard to see what cooperation is there to be had as
>>>> there is no indication Jakub is going to stop maliciously blocking
>>>> my work.
>>>
>>> The netdev side has been pretty explicit on wanting a MAINTAINERS entry
>>
>> Can you point out where that was requested dated before the series in
>> question? Because as far as I know, only CC'ing was mentioned and
>> only as a question, for which I proposed a fairly standard way of
>> dealing with it by introducing API and agreeing on any changes to that,
>> and got no reply. Even then, I was CC'ing netdev for changes that might
>> be interesting to netdev, that includes the blocked series.
> 
> Not interested in digging out those other discussions, but Mina had a
> patch back in August, and there was the previous discussion on the big

If August, I'm pretty sure you're referring to one of the
replies / follow ups after the mentioned series.

> patchset. At least I very much understood it as netdev wanting to be
> CC'ed, and the straight forward way to always have that is to make it
> explicit in MAINTAINERS.
> 
>>> so that they see changes. I don't think it's unreasonable to have that,
>>> and it doesn't mean that they need to ack things that are specific to
>>> zcrx. Nobody looks at all the various random lists, giving them easier
>>> insight is a good thing imho. I think we all agree on that.
>>>
>>> Absent that change, it's also not unreasonable for that side to drag
>>> their feet a bit on further changes. Could the communication have been
>>> better on that side? Certainly yes. But it's hard to blame them too much
>>> on that front, as any response would have predictably yielded an
>>> accusatory reply back.
>>
>> Not really, solely depends on the reply.
> 
> Well, statistically based on recent and earlier replies in those
> threads, if I was on that side, I'd say that would be a fair assumption.
> 
>>> And honestly, nobody wants to deal with that, if
>>
>> Understandable, but you're making it sound like I started by
>> throwing accusations and not the other way around. But it's
>> true that I never wanted to deal with it.
> 
> Honestly I don't even know where this all started, but it hasn't been
> going swimmingly the last few months would be my assessment.
> 
> My proposal is to put all of this behind us and move forward in a
> productive manner. There's absolutely nothing to be gained from
> continuing down the existing path of arguing about who did what and why,
> and frankly I have zero inclination to participate in that. It should be
> in everybody's best interest to move forward, productively. And if that
> starts with a simple MAINTAINERS entry, that seems like a good place to
> start. So _please_, can we do that?

I'm convinced it's not going to help with the work being
blocked or the aforementioned issues, but ok, let's have it
your way.

-- 
Pavel Begunkov

Powered by blists - more mailing lists