Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Message-ID: <27614af23cd7ae4433b909194062c553a6ae16ac.camel@redhat.com>
Subject: Re: [PATCH] udp6: Fix race condition in udp6_sendmsg & connect
From:   Paolo Abeni <pabeni@redhat.com>
To:     Vladislav Efanov <VEfanov@ispras.ru>,
        Willem de Bruijn <willemdebruijn.kernel@gmail.com>
Cc:     "David S. Miller" <davem@davemloft.net>,
        David Ahern <dsahern@kernel.org>,
        Eric Dumazet <edumazet@google.com>,
        Jakub Kicinski <kuba@kernel.org>, netdev@vger.kernel.org,
        linux-kernel@vger.kernel.org, lvc-project@linuxtesting.org
Date:   Fri, 26 May 2023 17:33:33 +0200
In-Reply-To: <20230526150806.1457828-1-VEfanov@ispras.ru>
References: <20230526150806.1457828-1-VEfanov@ispras.ru>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
User-Agent: Evolution 3.46.4 (3.46.4-1.fc37) 
MIME-Version: 1.0
Precedence: bulk

On Fri, 2023-05-26 at 18:08 +0300, Vladislav Efanov wrote:
> Syzkaller got the following report:
> BUG: KASAN: use-after-free in sk_setup_caps+0x621/0x690 net/core/sock.c:2=
018
> Read of size 8 at addr ffff888027f82780 by task syz-executor276/3255
>=20
> The function sk_setup_caps (called by ip6_sk_dst_store_flow->
> ip6_dst_store) referenced already freed memory as this memory was
> freed by parallel task in udpv6_sendmsg->ip6_sk_dst_lookup_flow->
> sk_dst_check.
>=20
>           task1 (connect)              task2 (udp6_sendmsg)
>         sk_setup_caps->sk_dst_set |
>                                   |  sk_dst_check->
>                                   |      sk_dst_set
>                                   |      dst_release
>         sk_setup_caps references  |
>         to already freed dst_entry|
>=20
> The reason for this race condition is: udp6_sendmsg() calls
> ip6_sk_dst_lookup() without lock for sock structure and tries to
> allocate/add dst_entry structure to sock structure in parallel with
> "connect" task.
>=20
> Found by Linux Verification Center (linuxtesting.org) with syzkaller.
>=20
> Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
> Signed-off-by: Vladislav Efanov <VEfanov@ispras.ru>

Thank you for the detailed report!

> ---
>  net/ipv6/udp.c | 3 +++
>  1 file changed, 3 insertions(+)
>=20
> diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c
> index e5a337e6b970..a5ecd5d93b0a 100644
> --- a/net/ipv6/udp.c
> +++ b/net/ipv6/udp.c
> @@ -1563,12 +1563,15 @@ int udpv6_sendmsg(struct sock *sk, struct msghdr =
*msg, size_t len)
> =20
>  	fl6->flowlabel =3D ip6_make_flowinfo(ipc6.tclass, fl6->flowlabel);
> =20
> +	lock_sock(sk);

Acquiring the socket lock in this fast-path is going to kill the xmit
performances, I think we can't do that.

What about something like the following instead? Does that addresses
the UaF? (completely untested, not even built ;) If so, feel free to
take it over.

Thanks.

Paolo
---
diff --git a/net/core/sock.c b/net/core/sock.c
index 5440e67bcfe3..24f2761bdb1d 100644
--- a/net/core/sock.c
+++ b/net/core/sock.c
@@ -2381,7 +2381,6 @@ void sk_setup_caps(struct sock *sk, struct dst_entry =
*dst)
 {
 	u32 max_segs =3D 1;
=20
-	sk_dst_set(sk, dst);
 	sk->sk_route_caps =3D dst->dev->features;
 	if (sk_is_tcp(sk))
 		sk->sk_route_caps |=3D NETIF_F_GSO;
@@ -2400,6 +2399,7 @@ void sk_setup_caps(struct sock *sk, struct dst_entry =
*dst)
 		}
 	}
 	sk->sk_gso_max_segs =3D max_segs;
+	sk_dst_set(sk, dst);
 }
 EXPORT_SYMBOL_GPL(sk_setup_caps);
=20