[PATCH net] af_packet: do not use READ_ONCE() in packet_bind()

[PATCH net] af_packet: do not use READ_ONCE() in packet_bind()

[Eric Dumazet]

A recent patch added READ_ONCE() in packet_bind() and packet_bind_spkt()

This is better handled by reading pkt_sk(sk)->num later
in packet_do_bind() while appropriate lock is held.

READ_ONCE() in writers are often an evidence of something being wrong.

Fixes: 822b5a1c17df ("af_packet: Fix data-races of pkt_sk(sk)->num.")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Kuniyuki Iwashima <kuniyu@amazon.com>
Cc: Willem de Bruijn <willemb@google.com>
---
 net/packet/af_packet.c | 8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)

diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c
index a1f9a0e9f3c8a72e5a95f96473b7b6c63f893935..a2dbeb264f260e5b8923ece9aac99fe19ddfeb62 100644
--- a/net/packet/af_packet.c
+++ b/net/packet/af_packet.c
@@ -3201,6 +3201,9 @@ static int packet_do_bind(struct sock *sk, const char *name, int ifindex,
 
 	lock_sock(sk);
 	spin_lock(&po->bind_lock);
+	if (!proto)
+		proto = po->num;
+
 	rcu_read_lock();
 
 	if (po->fanout) {
@@ -3299,7 +3302,7 @@ static int packet_bind_spkt(struct socket *sock, struct sockaddr *uaddr,
 	memcpy(name, uaddr->sa_data, sizeof(uaddr->sa_data_min));
 	name[sizeof(uaddr->sa_data_min)] = 0;
 
-	return packet_do_bind(sk, name, 0, READ_ONCE(pkt_sk(sk)->num));
+	return packet_do_bind(sk, name, 0, 0);
 }
 
 static int packet_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len)
@@ -3316,8 +3319,7 @@ static int packet_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len
 	if (sll->sll_family != AF_PACKET)
 		return -EINVAL;
 
-	return packet_do_bind(sk, NULL, sll->sll_ifindex,
-			      sll->sll_protocol ? : READ_ONCE(pkt_sk(sk)->num));
+	return packet_do_bind(sk, NULL, sll->sll_ifindex, sll->sll_protocol);
 }
 
 static struct proto packet_proto = {
-- 
2.41.0.rc0.172.g3f132b7071-goog

Re: [PATCH net] af_packet: do not use READ_ONCE() in packet_bind()

[Willem de Bruijn]

On Fri, May 26, 2023 at 11:43 AM Eric Dumazet <edumazet@google.com> wrote:
>
> A recent patch added READ_ONCE() in packet_bind() and packet_bind_spkt()
>
> This is better handled by reading pkt_sk(sk)->num later
> in packet_do_bind() while appropriate lock is held.
>
> READ_ONCE() in writers are often an evidence of something being wrong.
>
> Fixes: 822b5a1c17df ("af_packet: Fix data-races of pkt_sk(sk)->num.")
> Signed-off-by: Eric Dumazet <edumazet@google.com>
> Cc: Kuniyuki Iwashima <kuniyu@amazon.com>
> Cc: Willem de Bruijn <willemb@google.com>

Reviewed-by: Willem de Bruijn <willemb@google.com>

Thanks for catching this Eric.

Re: [PATCH net] af_packet: do not use READ_ONCE() in packet_bind()

[Jiri Pirko]

Fri, May 26, 2023 at 05:43:42PM CEST, edumazet@google.com wrote:
>A recent patch added READ_ONCE() in packet_bind() and packet_bind_spkt()
>
>This is better handled by reading pkt_sk(sk)->num later
>in packet_do_bind() while appropriate lock is held.

Nit: easier to understand the desctiption if you use imperative mood,
telling the codebase what to do.


>
>READ_ONCE() in writers are often an evidence of something being wrong.
>
>Fixes: 822b5a1c17df ("af_packet: Fix data-races of pkt_sk(sk)->num.")
>Signed-off-by: Eric Dumazet <edumazet@google.com>

Reviewed-by: Jiri Pirko <jiri@nvidia.com>

[PATCH net] af_packet: do not use READ_ONCE() in packet_bind()

[Kuniyuki Iwashima]

From: Eric Dumazet <edumazet@google.com>
Date: Fri, 26 May 2023 15:43:42 +0000
> A recent patch added READ_ONCE() in packet_bind() and packet_bind_spkt()
> 
> This is better handled by reading pkt_sk(sk)->num later
> in packet_do_bind() while appropriate lock is held.
> 
> READ_ONCE() in writers are often an evidence of something being wrong.
> 
> Fixes: 822b5a1c17df ("af_packet: Fix data-races of pkt_sk(sk)->num.")
> Signed-off-by: Eric Dumazet <edumazet@google.com>
> Cc: Kuniyuki Iwashima <kuniyu@amazon.com>
> Cc: Willem de Bruijn <willemb@google.com>

Reviewed-by: Kuniyuki Iwashima <kuniyu@amazon.com>

Thanks!


> ---
>  net/packet/af_packet.c | 8 +++++---
>  1 file changed, 5 insertions(+), 3 deletions(-)
> 
> diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c
> index a1f9a0e9f3c8a72e5a95f96473b7b6c63f893935..a2dbeb264f260e5b8923ece9aac99fe19ddfeb62 100644
> --- a/net/packet/af_packet.c
> +++ b/net/packet/af_packet.c
> @@ -3201,6 +3201,9 @@ static int packet_do_bind(struct sock *sk, const char *name, int ifindex,
>  
>  	lock_sock(sk);
>  	spin_lock(&po->bind_lock);
> +	if (!proto)
> +		proto = po->num;
> +
>  	rcu_read_lock();
>  
>  	if (po->fanout) {
> @@ -3299,7 +3302,7 @@ static int packet_bind_spkt(struct socket *sock, struct sockaddr *uaddr,
>  	memcpy(name, uaddr->sa_data, sizeof(uaddr->sa_data_min));
>  	name[sizeof(uaddr->sa_data_min)] = 0;
>  
> -	return packet_do_bind(sk, name, 0, READ_ONCE(pkt_sk(sk)->num));
> +	return packet_do_bind(sk, name, 0, 0);
>  }
>  
>  static int packet_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len)
> @@ -3316,8 +3319,7 @@ static int packet_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len
>  	if (sll->sll_family != AF_PACKET)
>  		return -EINVAL;
>  
> -	return packet_do_bind(sk, NULL, sll->sll_ifindex,
> -			      sll->sll_protocol ? : READ_ONCE(pkt_sk(sk)->num));
> +	return packet_do_bind(sk, NULL, sll->sll_ifindex, sll->sll_protocol);
>  }
>  
>  static struct proto packet_proto = {
> -- 
> 2.41.0.rc0.172.g3f132b7071-goog

Re: [PATCH net] af_packet: do not use READ_ONCE() in packet_bind()

[patchwork-bot+netdevbpf]

Hello:

This patch was applied to netdev/net.git (main)
by Jakub Kicinski <kuba@kernel.org>:

On Fri, 26 May 2023 15:43:42 +0000 you wrote:
> A recent patch added READ_ONCE() in packet_bind() and packet_bind_spkt()
> 
> This is better handled by reading pkt_sk(sk)->num later
> in packet_do_bind() while appropriate lock is held.
> 
> READ_ONCE() in writers are often an evidence of something being wrong.
> 
> [...]

Here is the summary with links:
  - [net] af_packet: do not use READ_ONCE() in packet_bind()
    https://git.kernel.org/netdev/net/c/6ffc57ea0042

You are awesome, thank you!
-- 
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html