On Tue, Nov 05, 2024 at 05:22:57AM +0100, Thorsten Glaser wrote:
> 
> So, why does 5.10.208-xfstests warn but 5.10.226-1 doesn’t?
> 
> (It *does* warn about r1 128-byte-inode ones.)

No idea...

% cd /usr/projects/linux/ext4-5.10
% git checkout v5.10.226
% install-kconfig
% kbuild
% kvm-xfstests shell
    ...
root@kvm-xfstests:~# mkfs.ext4 -Fq -r 0 /dev/vdc
/dev/vdc contains a ext2 file system
        created on Tue Nov  5 08:00:41 2024
root@kvm-xfstests:~# mount /dev/vdc /vdc
[   27.176032] ext2 filesystem being mounted at /vdc supports timestamps until 
2038 (0x7fffffff)
root@kvm-xfstests:~# uname -a
Linux kvm-xfstests 5.10.226-xfstests #9 SMP Tue Nov 5 07:59:55 EST 2024 x86_64 
GNU/Linux

So it works using the upstream Linux kernel.

Maybe double check, and then file a bug with the Debian kernel if you
can still reproduce?  Assuming that the Debian kernel maintainers care
about fixing the Debian old-stable kernel...

I could talk about all of the massive number of security fixes that
aren't backported to 5.10 and so are unsecure, but I assume you know
that.  Friends don't let friends use 5.10; there are reasons why
enterprise Linux customers pay $$$$$$ to Red Hat and SuSE.  It's
*expensive* to safely use ancient kernels.  :-)

For the record, at the moment as the ext4 maintainer I really only pay
attention to 6.1, 6.6, and 6.12 LTS kernels.  And there are a whole
bunch of xfstests which fail with the older LTS kernels.  Compare:

TESTRUNID: ltm-20241020233728-0001
KERNEL:    kernel 6.1.113-xfstests #1 SMP PREEMPT_DYNAMIC Sun Oct 20 23:44:12 
EDT 2024 x86_64
CMDLINE:   -c ext4/all,xfs/all -g auto --repo stable-rc.git --watch linux-6.1.y

ext4/4k: 576 tests, 3 failures, 55 skipped, 5166 seconds
  Failures: generic/365 generic/645 generic/743
xfs/4k: 1157 tests, 30 failures, 183 skipped, 20079 seconds
  Failures: generic/083 generic/091 generic/127 generic/263 generic/269 
    generic/270 generic/363 generic/365 generic/447 generic/475 
    generic/476 generic/616 generic/645 generic/743 generic/750 
    generic/754 xfs/104 xfs/168 xfs/305 xfs/348 xfs/442 xfs/538 xfs/597 
    xfs/598 xfs/629 xfs/630 xfs/631 xfs/632 xfs/806
...

with:

TESTRUNID: ltm-20241020234544-0011
KERNEL:    kernel 6.12.0-rc5-xfstests-00954-gcb7d894c0227 #1 SMP 
PREEMPT_DYNAMIC Sun Nov  3 18:09:05 EST 2024 x86_64
CMDLINE:   -c ext4/all,xfs/all,btrfs/all,f2fs/all -g auto --repo 
https://kernel.googlesource.com/pub/scm/linux/kernel/git/next/linux-next 
--watch fs-next

ext4/4k: 576 tests, 53 skipped, 5205 seconds
xfs/v4: 1156 tests, 3 failures, 476 skipped, 9951 seconds
  Failures: xfs/348 xfs/803 xfs/804
...

(And 5.10 and 5.15 are even worse, but I don't bother with them any
more.  Volunteers to manage the older ancient ext4 for stable kernels
would be greatly appreciated.  I don't have the time.)

                                                - Ted

Reply via email to