Re: linux-next: duplicate patch in the nolibc tree

From: Shuah Khan
Date: Mon Aug 21 2023 - 11:53:02 EST


On 8/18/23 07:59, Shuah Khan wrote:
On 8/18/23 07:27, Christian Brauner wrote:
On Thu, Aug 17, 2023 at 01:41:45PM -0700, Paul E. McKenney wrote:
On Thu, Aug 17, 2023 at 09:39:09PM +0200, Willy Tarreau wrote:
On Thu, Aug 17, 2023 at 11:46:57AM -0700, Paul E. McKenney wrote:
On Thu, Aug 17, 2023 at 12:27:46PM -0600, Shuah Khan wrote:
On 8/17/23 10:30, Thomas Weißschuh wrote:
On 2023-08-17 13:38:11+1000, Stephen Rothwell wrote:
The following commit is also in the vfs-brauner tree as a different commit
(but the same patch):

    ba859b2e419c ("selftests/nolibc: drop test chmod_net")

This is commit

    49319832de90 ("selftests/nolibc: drop test chmod_net")

in the vfs-brauner tree.

I think we can drop the patch from the nolibc tree.
The patch is only really necessary in combination with
commit 18e66ae67673 ("proc: use generic setattr() for /proc/$PID/net")
which already is and should stay in the vfs tree.

Thomas,

Do the rest of the nolibc patches build without this if we were
to drop this patch? Dorpping requires rebase and please see below.

Willy, Paul,

How do we want to handle this so we can avoid rebasing to keep
the Commit IDs the same as one ones in Willy's nolibc branch?

The usual way would be for Willy to drop the patch, rebase, and republish
his branch.  You would then discard the current branch and pull the
new one.

I would recommend dropping this commit from vfs-brauner if it
doesn't cause problems.

It might be good for nolibc patches to be going through Willy's tree.

It would indeed be more logical as a general rule. However, here I don't
care as I don't see any issue caused by dropping it, I can adapt to what
is most convenient for most of us.

Let's maybe just wait a little bit for Christian to suggest what he
prefers then we can adapt.

Or does Christian have some situation where it is necessary to make
a coordinated vfs/nolibc change?

I don't think there's any need for coordination on this one.

It is always good when either option can be make to work.  ;-)

The patch in the vfs tree will make the test fail so it makes sense to
have both go in together. I would normally be happy to drop it but I'm
rather unenthusiastic in this particular case because I replied to this
almost 5 weeks ago on Thursday, July 13 and since then this has been in
-next.


I totally understand you being unenthusiastic. Considering summer
vacation schedules and all, emails get missed at times.

I sincerely request you to consider dropping as it is the simpler route
for all involved.


Christian,

Please let us know if my request failed to raise your enthusiasm level.
We will go to our plan b of having Willy drop the patch, resend the
pull request to me ....


thanks,
-- Shuah