On Tue, Oct 01, 2024 at 08:52:28AM +0200, Greg Kroah-Hartman wrote:

On Tue, Oct 01, 2024 at 06:02:45AM +0200, Jason A. Donenfeld wrote:

...

Hi Sasha,

I've been getting emails from your bots...

I sent two pulls to Linus for 6.13-rc1:

https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?i... https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?i...

In these, I'm not sure there's actually much valid stable material. I didn't mark anything as Cc: stable@vger.kernel.org, I don't think.

As such, can you make sure none of those get backported?

Alternatively, if you do have reason to want to pick some of these, can you be clear with what and why, and actually carefully decide which ones and which dependencies are required as such in a non-automated way?

They say so directly in the commit, i.e.: Stable-dep-of: 6eda706a535c ("selftests: vDSO: fix the way vDSO functions are called for powerpc")

Wha? Sasha added that. It's not in the original commit.

Is that tag anywhere in Linus' tree?

in each one. So this seem to be needed to fix up the powerpc stuff.

I'll drop them all if you feel these should not be applied.

Yes, thanks.

Jason

On Tue, Oct 01, 2024 at 02:13:14PM +0200, Greg Kroah-Hartman wrote:

Ok, I'll try to rework the other dependant patches to see if we can get that fix in somehow without this change. But why not take this, what is it hurting?

I just don't see the need to backport *any* patches from my tree that don't have an explicit Cc: stable@ marker on them. I'm pretty careful about adding those, and when I forget, I send them manually onward to stable@. If there's some judgement that a certain patch needs to be backported that I didn't mark, that sounds like something to deliberately raise, rather than a heap of emails that this patch and that patch have been added willy-nilly.

The reason I care about this is that I generally care about stable and consistency of rationale and such, and so if you *do* want to backport some stuff, I am going to spend time checking and verifying and being careful. I don't want to do that work if it's just the consequence of a random script and not somebody's technical decision.