Re: [PATCH 4/4] af_unix: convert to lock_cmp_fn

[Kuniyuki Iwashima]

From: Kent Overstreet <kent.overstreet@xxxxxxxxx>
Date: Fri, 26 Jan 2024 21:08:31 -0500
> Kill
>  - unix_state_lock_nested
>  - _nested usage for net->unx.table.locks[].
> 
> replace both with lock_set_cmp_fn_ptr_order(&u->lock).
> 
> The lock ordering in sk_diag_dump_icons() looks suspicious; this may
> turn up a real issue.

Yes, you cannot use lock_cmp_fn() for unix_state_lock_nested().

The lock order in sk_diag_dump_icons() is

  listening socket -> child socket in the listener's queue

, and the inverse order never happens.  ptr comparison does not make
sense in this case, and lockdep will complain about false positive.


> 
> Cc: netdev@xxxxxxxxxxxxxxx
> Signed-off-by: Kent Overstreet <kent.overstreet@xxxxxxxxx>
> ---
>  include/net/af_unix.h |  3 ---
>  net/unix/af_unix.c    | 20 ++++++++------------
>  net/unix/diag.c       |  2 +-
>  3 files changed, 9 insertions(+), 16 deletions(-)
> 
> diff --git a/include/net/af_unix.h b/include/net/af_unix.h
> index 49c4640027d8..4eff0a089640 100644
> --- a/include/net/af_unix.h
> +++ b/include/net/af_unix.h
> @@ -48,9 +48,6 @@ struct scm_stat {
>  
>  #define unix_state_lock(s)	spin_lock(&unix_sk(s)->lock)
>  #define unix_state_unlock(s)	spin_unlock(&unix_sk(s)->lock)
> -#define unix_state_lock_nested(s) \
> -				spin_lock_nested(&unix_sk(s)->lock, \
> -				SINGLE_DEPTH_NESTING)
>  
>  /* The AF_UNIX socket */
>  struct unix_sock {
> diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c
> index d013de3c5490..1a0d273799c1 100644
> --- a/net/unix/af_unix.c
> +++ b/net/unix/af_unix.c
> @@ -170,7 +170,7 @@ static void unix_table_double_lock(struct net *net,
>  		swap(hash1, hash2);
>  
>  	spin_lock(&net->unx.table.locks[hash1]);
> -	spin_lock_nested(&net->unx.table.locks[hash2], SINGLE_DEPTH_NESTING);
> +	spin_lock(&net->unx.table.locks[hash2]);
>  }
>  
>  static void unix_table_double_unlock(struct net *net,
> @@ -997,6 +997,7 @@ static struct sock *unix_create1(struct net *net, struct socket *sock, int kern,
>  	u->path.dentry = NULL;
>  	u->path.mnt = NULL;
>  	spin_lock_init(&u->lock);
> +	lock_set_cmp_fn_ptr_order(&u->lock);
>  	atomic_long_set(&u->inflight, 0);
>  	INIT_LIST_HEAD(&u->link);
>  	mutex_init(&u->iolock); /* single task reading lock */
> @@ -1340,17 +1341,11 @@ static int unix_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len)
>  
>  static void unix_state_double_lock(struct sock *sk1, struct sock *sk2)
>  {
> -	if (unlikely(sk1 == sk2) || !sk2) {
> -		unix_state_lock(sk1);
> -		return;
> -	}
> -	if (sk1 < sk2) {
> +	if (sk1 > sk2)
> +		swap(sk1, sk2);
> +	if (sk1 && sk1 != sk2)
>  		unix_state_lock(sk1);
> -		unix_state_lock_nested(sk2);
> -	} else {
> -		unix_state_lock(sk2);
> -		unix_state_lock_nested(sk1);
> -	}
> +	unix_state_lock(sk2);
>  }
>  
>  static void unix_state_double_unlock(struct sock *sk1, struct sock *sk2)
> @@ -1591,7 +1586,7 @@ static int unix_stream_connect(struct socket *sock, struct sockaddr *uaddr,
>  		goto out_unlock;
>  	}
>  
> -	unix_state_lock_nested(sk);
> +	unix_state_lock(sk);
>  
>  	if (sk->sk_state != st) {
>  		unix_state_unlock(sk);
> @@ -3575,6 +3570,7 @@ static int __net_init unix_net_init(struct net *net)
>  
>  	for (i = 0; i < UNIX_HASH_SIZE; i++) {
>  		spin_lock_init(&net->unx.table.locks[i]);
> +		lock_set_cmp_fn_ptr_order(&net->unx.table.locks[i]);
>  		INIT_HLIST_HEAD(&net->unx.table.buckets[i]);
>  	}
>  
> diff --git a/net/unix/diag.c b/net/unix/diag.c
> index bec09a3a1d44..8ab5e2217e4c 100644
> --- a/net/unix/diag.c
> +++ b/net/unix/diag.c
> @@ -84,7 +84,7 @@ static int sk_diag_dump_icons(struct sock *sk, struct sk_buff *nlskb)
>  			 * queue lock. With the other's queue locked it's
>  			 * OK to lock the state.
>  			 */
> -			unix_state_lock_nested(req);
> +			unix_state_lock(req);
>  			peer = unix_sk(req)->peer;
>  			buf[i++] = (peer ? sock_i_ino(peer) : 0);
>  			unix_state_unlock(req);
> -- 
> 2.43.0