linux-kernel - "immediate fixes" for user-reported regressions (was: Re: [GIT PULL] pwm: Two fixes and a maintainer update)

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <1f9ecc30-1624-4d2b-aa86-a1f0dff6b97a@leemhuis.info>
Date: Wed, 28 Jan 2026 17:12:16 +0100
From: Thorsten Leemhuis <regressions@...mhuis.info>
To: Linus Torvalds <torvalds@...ux-foundation.org>
Cc: Uwe Kleine-König <u.kleine-koenig@...libre.com>,
 linux-pwm@...r.kernel.org, linux-kernel@...r.kernel.org,
 Michal Wilczynski <m.wilczynski@...sung.com>,
 Maxime Ripard <mripard@...hat.com>,
 Richard Genoud <richard.genoud@...tlin.com>
Subject: "immediate fixes" for user-reported regressions (was: Re: [GIT PULL]
 pwm: Two fixes and a maintainer update)

On 1/22/26 19:07, Linus Torvalds wrote:

> But a user complaining should basically result in an immediate fix -
> possibly a "revert and rethink".

So how can I/we make "immediate fixes" happen more often without
contributing to maintainer burnout?

Because I see quite a few user-reported regressions that take quite a
while to get resolved. I usually tried prodding maintainers, but well,
that led to major complaints reg. maintainer burnout and in one case
contributed to a maintainer stepping down. And many of those complaints
even happened with simple cases, like me prodding when (a) fixes/reverts
with a Tested-by: from the reporter were waiting on the list for review
and merging for one or two weeks or (b) were merged in the right
subsystem tree already for a similar timeframe and just not sent along.

And I can even understand that, as maintainers have a lot on their plate
already and sometimes need a break, too.

We could obviously start bypassing the regular channels occasionally
when no "immediate fix" comes forward through them if that's what you
want. I could, for example, point you to fixes for recent regressions or
even collect and send them your way when they fall into one of these
categories:
- Reverts that I or someone else submitted that got a Tested-by: from
the reporter.
- Unreviewed fixes that got a Tested-by: from the reporter and no "this
is all wrong" complaints from the developers within a two or three days.
- Fixes cherry-picked from the subsystem trees not sent along.
- Fixes maintainers ask me to pick up to send along.

But well, bypassing the regular channels obviously has it's downsides,
as things are more likely to get wrong that way. But I guess if you
really want "immediate fixes" for user-reported regressions, we afaics
likely need something like that. Unless you have a better idea.

Ciao, Thorsten