New subject: [PATCH net] net: clear sk->sk_ino in sk_set_socket(sk, NULL)

17 Sep 2025


      On Wed, Sep 17, 2025 at 10:03 AM Andrei Vagin avagin@google.com wrote:
...
is
On Wed, Sep 17, 2025 at 8:59 AM Eric Dumazet edumazet@google.com wrote:
...
On Wed, Sep 17, 2025 at 8:39 AM Andrei Vagin avagin@google.com wrote:
...
On Wed, Sep 17, 2025 at 6:53 AM Eric Dumazet edumazet@google.com wrote:
...
Andrei Vagin reported that blamed commit broke CRIU.
Indeed, while we want to keep sk_uid unchanged when a socket
is cloned, we want to clear sk->sk_ino.
Otherwise, sock_diag might report multiple sockets sharing
the same inode number.
Move the clearing part from sock_orphan() to sk_set_socket(sk, NULL),
called both from sock_orphan() and sk_clone_lock().
Fixes: 5d6b58c932ec ("net: lockless sock_i_ino()")
Closes: https://lore.kernel.org/netdev/aMhX-VnXkYDpKd9V@google.com/
Closes: https://github.com/checkpoint-restore/criu/issues/2744
Reported-by: Andrei Vagin avagin@google.com
Signed-off-by: Eric Dumazet edumazet@google.com
Acked-by: Andrei Vagin avagin@google.com
I think we need to add `Cc: stable@vger.kernel.org`.
I never do this. Note that the prior patch had no such CC.
The original patch has been ported to the v6.16 kernels. According to the
kernel documentation
(https://www.kernel.org/doc/html/v6.5/process/stable-kernel-rules.html),
adding Cc: stable@vger.kernel.org is required for automatic porting into
stable trees. Without this tag, someone will likely need to manually request
that this patch be ported. This is my understanding of how the stable
branch process works, sorry if I missed something.
Andrei, I think I know pretty well what I am doing. You do not have to
explain to me anything.
Thank you.

Re: [PATCH net] net: clear sk->sk_ino in sk_set_socket(sk, NULL)