• trixie kernel and cifs-utils mismatch

    From Andreas B. Mundt@1:229/2 to All on Mon May 19 13:40:01 2025
    XPost: linux.debian.bugs.dist
    From: andi@debian.org

    Hi,

    I would like to draw your attention to #1105747 [1] which affects
    enterprise installations where kerberized cifs mounts are used. I am
    not sure if anybody is aware of this issue.

    Right now, the trixie kernel and cifs-utils do not fit together
    (c.f. reported bug above). Some more background I could find:

    • cifs-utils added a new mount option upcall_target [2] recently
    • the corresponding linux patch landed in the 6.13 kernel tree [3]

    Perhaps the patch can be backported to trixie's kernel version.

    Thanks and best regards,

    Andi


    [1] https://bugs.debian.org/1105747
    [2] https://git.samba.org/?p=cifs-utils.git;a=commit;h=89b679228cc1be9739d54203d28289b03352c174
    [3] https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=db363b0a1d9e6b9dc556296f1b1007aeb496a8cf

    --- SoupGate-Win32 v1.05
    * Origin: you cannot sedate... all the things you hate (1:229/2)
  • From Salvatore Bonaccorso@1:229/2 to Andreas B. Mundt on Tue May 20 10:30:01 2025
    XPost: linux.debian.bugs.dist
    From: carnil@debian.org

    Hi Andi,

    On Mon, May 19, 2025 at 01:13:14PM +0200, Andreas B. Mundt wrote:
    Hi,

    I would like to draw your attention to #1105747 [1] which affects
    enterprise installations where kerberized cifs mounts are used. I am
    not sure if anybody is aware of this issue.

    Right now, the trixie kernel and cifs-utils do not fit together
    (c.f. reported bug above). Some more background I could find:

    • cifs-utils added a new mount option upcall_target [2] recently
    • the corresponding linux patch landed in the 6.13 kernel tree [3]

    Perhaps the patch can be backported to trixie's kernel version.

    I have asked stable maintainers upstream, and it got queued for the
    6.12.y stable series: https://git.kernel.org/pub/scm/linux/kernel/git/stable/stable-queue.git/commit/?id=7294cf6ff42482531f62f9b2a74d9c7ee00bbb59

    I'm not reassigning the bug to src:linux as I think the cifs-utils
    might need to handle such situation more gracefully? I.e. usually we
    cannot across packages in distribution guarantee such a lockstep
    update.

    cifs-uitls maintainers, might that be something which can be handled
    on both sides? At least for upstream 6.12.y should look good now per
    se.

    Regards,
    Salvatore

    --- SoupGate-Win32 v1.05
    * Origin: you cannot sedate... all the things you hate (1:229/2)