From: Jakub Kicinski <kuba@kernel.org>
To: Linus Torvalds <torvalds@linux-foundation.org>
Cc: davem@davemloft.net, netdev@vger.kernel.org,
linux-kernel@vger.kernel.org, pabeni@redhat.com
Subject: Re: [GIT PULL] Networking for v7.3-rc5
Date: Thu, 24 Sep 2026 13:27:43 -0700 [thread overview]
Message-ID: <20260924132743.50090cf5@kernel.org> (raw)
In-Reply-To: <CAHk-=wiSnTE9vBZ=5_v+3EEkdazCCbBM5YABRzRHUAeRdyd4Xw@mail.gmail.com>
On Thu, 24 Sep 2026 12:04:05 -0700 Linus Torvalds wrote:
> I'm looking at those ovpn fixes, for example, and they fix old stuff
> and they aren't marked for stable. The ones I looked at seemed to go
> back to ovpn being basically merged initially.
>
> So they clearly aren't regression fixes. Why are they in an -rc at all then?
>
> Are they actually important enough that they shouldn't just go into linux-next?
>
> Your merge message notes don't even mention them, which to me says
> "not important".
>
> "Not a regression, not critical" should mean "next release, not rc material".
>
> Maybe it *is* important, but then it should be mentioned.
Yes, ovpn stuff was definitely pushing it :/ More of a decision
on whether I want to ship this to you today for rc5 or risk making
rc6 larger.
> I've pulled this, but I really am trying to start pushing back on
> pointless "fixes" after the merge window.
>
> For post-merge-window patches, it is NOT ENOUGH to just be "a fix".
>
> It needs to be a regression, or it needs to be critical. It needs to
> be something that would merit the change being marked for stable.
>
> Regressions are serious business. They break existing setups. They
> make it painful for people to upgrade. They are critical, in other
> words.
>
> But stuff that has been broken since the first release of the code?
>
> Sure, it *can* be a very important security fix. Or - more likely -
> it's just a codepath that doesn't matter, never did, and was found by
> some random LLM and that's the only reason it now shows up.
Not gonna argue, we'll continue to adjust the filters.
At the same time I'm very impressed by the overall code quality these
days. Regressions are very rare for us. Even mindless slop that people
send is an order of magnitude better than code we used to merge 2 years
ago. But more importantly with multiple frontier models reviewing each
merged patch the math on the risk of regressions changes quite a bit.
AFAIU the rc releases where about stabilizing the code merged during
the merge window. If the see no bug fixes for that code because LLMs
and CIs have beating all the bugs out of it before the merge window..
what are the rc releases about? 🤔️
next prev parent reply other threads:[~2026-09-24 20:27 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-09-24 18:32 Jakub Kicinski
2026-09-24 19:04 ` Linus Torvalds
2026-09-24 20:27 ` Jakub Kicinski [this message]
2026-09-24 21:02 ` Linus Torvalds
2026-09-24 19:16 ` pr-tracker-bot
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20260924132743.50090cf5@kernel.org \
--to=kuba@kernel.org \
--cc=davem@davemloft.net \
--cc=linux-kernel@vger.kernel.org \
--cc=netdev@vger.kernel.org \
--cc=pabeni@redhat.com \
--cc=torvalds@linux-foundation.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®