net: change sock.sk_ino and sock_i_ino() to u64

inode->i_ino is being converted to a u64. sock.sk_ino (which caches the
inode number) must also be widened to avoid truncation on 32-bit
architectures where unsigned long is only 32 bits.

Change sk_ino from unsigned long to u64, and update the return type
of sock_i_ino() to match. Fix all format strings that print the
result of sock_i_ino() (%lu -> %llu), and widen the intermediate
variables and function parameters in the diag modules that were
using int to hold the inode number.

Note that the UAPI socket diag structures (inet_diag_msg.idiag_inode,
unix_diag_msg.udiag_ino, etc.) are all __u32 and cannot be changed
without breaking the ABI. The assignments to those fields will
silently truncate, which is the existing behavior.

Acked-by: Marc Kleine-Budde <mkl@pengutronix.de> # for net/can
Signed-off-by: Jeff Layton <jlayton@kernel.org>
Link: https://patch.msgid.link/20260304-iino-u64-v3-3-2257ad83d372@kernel.org
Signed-off-by: Christian Brauner <brauner@kernel.org>

1 files changed, 2 insertions, 2 deletions

diff --git a/include/net/sock.h b/include/net/sock.h
index 66b56288c1d3..b08cb0347e5e 100644
--- a/include/net/sock.h
+++ b/include/net/sock.h
@@ -537,7 +537,7 @@ struct sock {
 	rwlock_t		sk_callback_lock;
 	u32			sk_ack_backlog;
 	u32			sk_max_ack_backlog;
-	unsigned long		sk_ino;
+	u64			sk_ino;
 	spinlock_t		sk_peer_lock;
 	int			sk_bind_phc;
 	struct pid		*sk_peer_pid;
@@ -2140,7 +2140,7 @@ static inline void sock_graft(struct sock *sk, struct socket *parent)
 	write_unlock_bh(&sk->sk_callback_lock);
 }
 
-static inline unsigned long sock_i_ino(const struct sock *sk)
+static inline u64 sock_i_ino(const struct sock *sk)
 {
 	/* Paired with WRITE_ONCE() in sock_graft() and sock_orphan() */
 	return READ_ONCE(sk->sk_ino);