Re: [REGRESSION] Linux kernel 6.12.75 fails to compile with -Werror=implicit-function-declaration

[Brett A C Sheffield <bacs@librecast.net>]

On 2026-03-04 20:00, Peter Schneider wrote:
> I already found and reported this in the RC cycle [1], and Sasha dropped it in -rc2 [2], and now in the release, it 
> obviously has, somewhat mysteriously, reappeared [3], affecting all of today's 6.x stable branch releases.

Greg, Sasha et al.

Can we make a small adjustment to the stable kernel testing process please,
whereby we release a kernel that we have actually tested, instead of adding and
dropping patches at the last moment and releasing a kernel that no one has
tested?

We are only a small pool of testers. If we find a bug, can we fix it, release a
new RC and test again please?  We can have an RC3. Even an RC4.  Perhaps if we
bogoselect fewer patches in the first place we might have less work to do. It's
better to miss a backport for a bug no one has reported than to pull stuff in
without proper review.

The current stable process is introducing bugs. Bugs that never existed in
mainline.

The 3 kernels released today were tested by no one before release. The seven
kernels yesterday were similarly tested by no one before release.  We weren't
given the opportunity.

There aren't enough of us to do this right, but we can do it less wrong with a
bit of caution.

Cheers,


Brett