[Bug,215693] New: [xfstests generic/673] file on XFS lose its sgid bit after reflink, if there's only sgid bit

https://bugzilla.kernel.org/show_bug.cgi?id=215693

            Bug ID: 215693
           Summary: [xfstests generic/673] file on XFS lose its sgid bit
                    after reflink, if there's only sgid bit
           Product: File System
           Version: 2.5
    Kernel Version: xfs-5.18-merge-1
          Hardware: All
                OS: Linux
              Tree: Mainline
            Status: NEW
          Severity: normal
          Priority: P1
         Component: XFS
          Assignee: filesystem_xfs@kernel-bugs.kernel.org
          Reporter: zlang@redhat.com
        Regression: No

xfstests suddently generic/673 fails on latest xfs-5.18-merge-1:
FSTYP         -- xfs (debug)
PLATFORM      -- Linux/aarch64 hpe-xxx-xx-x-xxxx-xx 5.17.0-rc8+ #1 SMP Mon Mar
14 15:30:26 EDT 2022
MKFS_OPTIONS  -- -f -m crc=1,finobt=1,rmapbt=0,reflink=1,bigtime=1,inobtcount=1
/dev/vda3
MOUNT_OPTIONS -- -o context=system_u:object_r:root_t:s0 /dev/vda3
/mnt/xfstests/scratch

generic/673     - output mismatch (see
/var/lib/xfstests/results//generic/673.out.bad)
    --- tests/generic/673.out   2022-03-14 19:50:16.969436417 -0400
    +++ /var/lib/xfstests/results//generic/673.out.bad  2022-03-15
12:12:33.080620337 -0400
    @@ -3,7 +3,7 @@
     310f146ce52077fcd3308dcbe7632bb2  SCRATCH_MNT/a
     6666 -rwSrwSrw- SCRATCH_MNT/a
     3784de23efab7a2074c9ec66901e39e5  SCRATCH_MNT/a
    -2666 -rw-rwSrw- SCRATCH_MNT/a
    +666 -rw-rw-rw- SCRATCH_MNT/a

     Test 2 - qa_user, group-exec file
    ...
    (Run 'diff -u /var/lib/xfstests/tests/generic/673.out
/var/lib/xfstests/results//generic/673.out.bad'  to see the entire diff)
Ran: generic/673
Failures: generic/673
Failed 1 of 1 tests

The diff output is:

According to the generic/673.out, the failed lines are
QA output created by 673
Test 1 - qa_user, non-exec file
310f146ce52077fcd3308dcbe7632bb2  SCRATCH_MNT/a
6666 -rwSrwSrw- SCRATCH_MNT/a
3784de23efab7a2074c9ec66901e39e5  SCRATCH_MNT/a
2666 -rw-rwSrw- SCRATCH_MNT/a                    <------ fail

Test 2 - qa_user, group-exec file
310f146ce52077fcd3308dcbe7632bb2  SCRATCH_MNT/a
6676 -rwSrwsrw- SCRATCH_MNT/a
3784de23efab7a2074c9ec66901e39e5  SCRATCH_MNT/a
676 -rw-rwxrw- SCRATCH_MNT/a

Test 3 - qa_user, user-exec file
310f146ce52077fcd3308dcbe7632bb2  SCRATCH_MNT/a
6766 -rwsrwSrw- SCRATCH_MNT/a
3784de23efab7a2074c9ec66901e39e5  SCRATCH_MNT/a
2766 -rwxrwSrw- SCRATCH_MNT/a                    <------ fail
...
...

I've reported another about losing sgid or suid bits after chown:
https://bugzilla.kernel.org/show_bug.cgi?id=215687
I doubt they're same issue, but due to they're about different testing (chown
and reflink), and the file not always lose its sgid bit after reflink(only Test
1 and 3). So I hope to get review from xfs expert, to make sure the behavior
changes are all as your expected, then we can change the test case's expect
result.

https://bugzilla.kernel.org/show_bug.cgi?id=215693

The Linux kernel's regression tracker (Thorsten Leemhuis) (regressions@leemhuis.info) changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |regressions@leemhuis.info

--- Comment #1 from The Linux kernel's regression tracker (Thorsten Leemhuis) (regressions@leemhuis.info) ---
Is this issue still happening? was it also discussed any maybe solved already,
like the other bug you mentioned? or should I poke the developers to get things
moving?

https://bugzilla.kernel.org/show_bug.cgi?id=215693

--- Comment #2 from Zorro Lang (zlang@redhat.com) ---
(In reply to The Linux kernel's regression tracker (Thorsten Leemhuis) from
comment #1)
> Is this issue still happening? was it also discussed any maybe solved
> already, like the other bug you mentioned? or should I poke the developers
> to get things moving?

Please refer to:
https://bugzilla.kernel.org/show_bug.cgi?id=215687#c4