get_maintainer, b4, and CC: stable

get_maintainer, b4, and CC: stable

[Nick Desaulniers]

Hi Joe,
A recent mistake I made when using b4 was that I added a fixes tag,
but forgot to Cc: stable in the commit message.

Speaking with Konstantine, it seems that b4 may just be importing the
recommendations from get_maintainer.pl.

I suspect that either b4 or get_maintainer could see the Fixes tag and
then suggest to Cc stable for me.

Should get_maintainer.pl make such recommendations?
-- 
Thanks,
~Nick Desaulniers

Re: get_maintainer, b4, and CC: stable

[Mark Brown]

[-- Attachment #1: Type: text/plain, Size: 316 bytes --]

On Wed, Aug 09, 2023 at 09:50:00AM -0700, Nick Desaulniers wrote:

> I suspect that either b4 or get_maintainer could see the Fixes tag and
> then suggest to Cc stable for me.

> Should get_maintainer.pl make such recommendations?

People use the Fixes tag all the time for bugs that never made it into a
release...

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

Re: get_maintainer, b4, and CC: stable

[Joe Perches]

On Wed, 2023-08-09 at 18:21 +0100, Mark Brown wrote:
> On Wed, Aug 09, 2023 at 09:50:00AM -0700, Nick Desaulniers wrote:
> 
> > I suspect that either b4 or get_maintainer could see the Fixes tag and
> > then suggest to Cc stable for me.
> 
> > Should get_maintainer.pl make such recommendations?
> 
> People use the Fixes tag all the time for bugs that never made it into a
> release...


As Nick suggests, likely not.  Many fixes do
not belong in stable and are only commits
that update the current rc.

get_maintainer checks the "fixes:" tag and
cc's the signers of the commit being fixed.

Re: get_maintainer, b4, and CC: stable

[Theodore Ts'o]

On Wed, Aug 09, 2023 at 06:21:51PM +0100, Mark Brown wrote:
> On Wed, Aug 09, 2023 at 09:50:00AM -0700, Nick Desaulniers wrote:
> 
> > I suspect that either b4 or get_maintainer could see the Fixes tag and
> > then suggest to Cc stable for me.
> 
> > Should get_maintainer.pl make such recommendations?
> 
> People use the Fixes tag all the time for bugs that never made it into a
> release...

I agree that it probably shouldn't.

Sometimes the bug was introduced by a commit that didn't have a Cc:
stable@kernel.org, but it gets automatically pulled into a LTS kernel
due to dependency reasons, or otherwise gets auto-selected into an LTS
kernel.  So I try to add Fixes tags even for bugs that never make it
into the stable kernel --- but that doesn't mean that it should
automatically get a cc stable tag.

(Of course, it might be that the AUTOSEL process will automatically
pull in such commit, and then pull in something probably should not
been pulled into a stable tree, but this is why XFS has stable
backports maintainers --- because they don't trust the LTS automation.
For ext4, we probably see one of those sorts of the auto-backports
caused a regression maybe once a year?  But that's a different
debate.)

	    	    	       	 	     - Ted