netdev - Re: [Patch bpf-next] unix_bpf: fix a potential deadlock in unix_dgram_bpf

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <87h7ge3fya.fsf@cloudflare.com>
Date:   Wed, 28 Jul 2021 12:12:13 +0200
From:   Jakub Sitnicki <jakub@...udflare.com>
To:     Cong Wang <xiyou.wangcong@...il.com>
Cc:     netdev@...r.kernel.org, Cong Wang <cong.wang@...edance.com>,
        Eric Dumazet <eric.dumazet@...il.com>,
        John Fastabend <john.fastabend@...il.com>,
        Daniel Borkmann <daniel@...earbox.net>,
        Lorenz Bauer <lmb@...udflare.com>
Subject: Re: [Patch bpf-next] unix_bpf: fix a potential deadlock in
 unix_dgram_bpf_recvmsg()

On Fri, Jul 23, 2021 at 08:36 PM CEST, Cong Wang wrote:
> From: Cong Wang <cong.wang@...edance.com>
>
> As Eric noticed, __unix_dgram_recvmsg() may acquire u->iolock
> too, so we have to release it before calling this function.
>
> Fixes: 9825d866ce0d ("af_unix: Implement unix_dgram_bpf_recvmsg()")
> Reported-by: Eric Dumazet <eric.dumazet@...il.com>
> Cc: John Fastabend <john.fastabend@...il.com>
> Cc: Daniel Borkmann <daniel@...earbox.net>
> Cc: Jakub Sitnicki <jakub@...udflare.com>
> Cc: Lorenz Bauer <lmb@...udflare.com>
> Signed-off-by: Cong Wang <cong.wang@...edance.com>
> ---
>  net/unix/unix_bpf.c | 11 ++++++-----
>  1 file changed, 6 insertions(+), 5 deletions(-)
>
> diff --git a/net/unix/unix_bpf.c b/net/unix/unix_bpf.c
> index db0cda29fb2f..b07cb30e87b1 100644
> --- a/net/unix/unix_bpf.c
> +++ b/net/unix/unix_bpf.c
> @@ -53,8 +53,9 @@ static int unix_dgram_bpf_recvmsg(struct sock *sk, struct msghdr *msg,
>  	mutex_lock(&u->iolock);
>  	if (!skb_queue_empty(&sk->sk_receive_queue) &&
>  	    sk_psock_queue_empty(psock)) {
> -		ret = __unix_dgram_recvmsg(sk, msg, len, flags);
> -		goto out;
> +		mutex_unlock(&u->iolock);
> +		sk_psock_put(sk, psock);
> +		return __unix_dgram_recvmsg(sk, msg, len, flags);
>  	}
>  
>  msg_bytes_ready:
> @@ -68,13 +69,13 @@ static int unix_dgram_bpf_recvmsg(struct sock *sk, struct msghdr *msg,
>  		if (data) {
>  			if (!sk_psock_queue_empty(psock))
>  				goto msg_bytes_ready;
> -			ret = __unix_dgram_recvmsg(sk, msg, len, flags);
> -			goto out;
> +			mutex_unlock(&u->iolock);
> +			sk_psock_put(sk, psock);
> +			return __unix_dgram_recvmsg(sk, msg, len, flags);
>  		}
>  		copied = -EAGAIN;
>  	}
>  	ret = copied;
> -out:
>  	mutex_unlock(&u->iolock);
>  	sk_psock_put(sk, psock);
>  	return ret;

Nit: Can be just `return copied`. `ret` became useless.

Acked-by: Jakub Sitnicki <jakub@...udflare.com>