netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [PATCH net-next v1 resend 0/2] Update bhash2 when socket's rcv saddr changes
@ 2022-06-01 20:14 Joanne Koong
  2022-06-01 20:14 ` [PATCH net-next v1 resend 1/2] net: " Joanne Koong
  2022-06-01 20:14 ` [PATCH net-next v1 resend 2/2] selftests/net: Add sk_bind_sendto_listen test Joanne Koong
  0 siblings, 2 replies; 6+ messages in thread
From: Joanne Koong @ 2022-06-01 20:14 UTC (permalink / raw)
  To: netdev, bpf; +Cc: edumazet, kafai, kuba, davem, pabeni, testing, Joanne Koong

From: Joanne Koong <joannelkoong@gmail.com>

As syzbot noted [1], there is an inconsistency in the bhash2 table in the case
where a socket's rcv saddr changes after it is binded. (For more details,
please see the commit message of the first patch)

This patchset fixes that and adds a test that triggers the case where the sk's
rcv saddr changes. The subsequent listen() call should succeed.

[1] https://lore.kernel.org/netdev/0000000000003f33bc05dfaf44fe@google.com/

Joanne Koong (2):
  net: Update bhash2 when socket's rcv saddr changes
  selftests/net: Add sk_bind_sendto_listen test

 include/net/inet_hashtables.h                 |  6 +-
 include/net/ipv6.h                            |  2 +-
 net/dccp/ipv4.c                               | 10 ++-
 net/dccp/ipv6.c                               |  4 +-
 net/ipv4/af_inet.c                            |  7 +-
 net/ipv4/inet_hashtables.c                    | 69 ++++++++++++++--
 net/ipv4/tcp_ipv4.c                           |  8 +-
 net/ipv6/inet6_hashtables.c                   |  4 +-
 net/ipv6/tcp_ipv6.c                           |  4 +-
 tools/testing/selftests/net/.gitignore        |  1 +
 tools/testing/selftests/net/Makefile          |  1 +
 .../selftests/net/sk_bind_sendto_listen.c     | 82 +++++++++++++++++++
 12 files changed, 180 insertions(+), 18 deletions(-)
 create mode 100644 tools/testing/selftests/net/sk_bind_sendto_listen.c

-- 
2.30.2


^ permalink raw reply	[flat|nested] 6+ messages in thread

* [PATCH net-next v1 resend 1/2] net: Update bhash2 when socket's rcv saddr changes
  2022-06-01 20:14 [PATCH net-next v1 resend 0/2] Update bhash2 when socket's rcv saddr changes Joanne Koong
@ 2022-06-01 20:14 ` Joanne Koong
  2022-06-02  1:09   ` kernel test robot
                     ` (2 more replies)
  2022-06-01 20:14 ` [PATCH net-next v1 resend 2/2] selftests/net: Add sk_bind_sendto_listen test Joanne Koong
  1 sibling, 3 replies; 6+ messages in thread
From: Joanne Koong @ 2022-06-01 20:14 UTC (permalink / raw)
  To: netdev, bpf
  Cc: edumazet, kafai, kuba, davem, pabeni, testing, Joanne Koong,
	syzbot+015d756bbd1f8b5c8f09, Eric Dumazet

From: Joanne Koong <joannelkoong@gmail.com>

Commit d5a42de8bdbe ("net: Add a second bind table hashed by port and
address") added a second bind table, bhash2, that hashes by a socket's port
and rcv address.

However, there are two cases where the socket's rcv saddr can change
after it has been binded:

1) The case where there is a bind() call on "::" (IPADDR_ANY) and then
a connect() call. The kernel will assign the socket an address when it
handles the connect()

2) In inet_sk_reselect_saddr(), which is called when rerouting fails
when rebuilding the sk header (invoked by inet_sk_rebuild_header)

In these two cases, we need to update the bhash2 table by removing the
entry for the old address, and adding a new entry reflecting the updated
address.

Reported-by: syzbot+015d756bbd1f8b5c8f09@syzkaller.appspotmail.com
Fixes: d5a42de8bdbe ("net: Add a second bind table hashed by port and address")
Signed-off-by: Joanne Koong <joannelkoong@gmail.com>
Reviewed-by: Eric Dumazet <edumzet@google.com>
---
 include/net/inet_hashtables.h |  6 ++-
 include/net/ipv6.h            |  2 +-
 net/dccp/ipv4.c               | 10 +++--
 net/dccp/ipv6.c               |  4 +-
 net/ipv4/af_inet.c            |  7 +++-
 net/ipv4/inet_hashtables.c    | 69 ++++++++++++++++++++++++++++++++---
 net/ipv4/tcp_ipv4.c           |  8 +++-
 net/ipv6/inet6_hashtables.c   |  4 +-
 net/ipv6/tcp_ipv6.c           |  4 +-
 9 files changed, 96 insertions(+), 18 deletions(-)

diff --git a/include/net/inet_hashtables.h b/include/net/inet_hashtables.h
index a0887b70967b..2c331ce6ca73 100644
--- a/include/net/inet_hashtables.h
+++ b/include/net/inet_hashtables.h
@@ -448,11 +448,13 @@ static inline void sk_rcv_saddr_set(struct sock *sk, __be32 addr)
 }
 
 int __inet_hash_connect(struct inet_timewait_death_row *death_row,
-			struct sock *sk, u64 port_offset,
+			struct sock *sk, u64 port_offset, bool prev_inaddr_any,
 			int (*check_established)(struct inet_timewait_death_row *,
 						 struct sock *, __u16,
 						 struct inet_timewait_sock **));
 
 int inet_hash_connect(struct inet_timewait_death_row *death_row,
-		      struct sock *sk);
+		      struct sock *sk, bool prev_inaddr_any);
+
+int inet_bhash2_update_saddr(struct sock *sk);
 #endif /* _INET_HASHTABLES_H */
diff --git a/include/net/ipv6.h b/include/net/ipv6.h
index 5b38bf1a586b..6a50aca56d50 100644
--- a/include/net/ipv6.h
+++ b/include/net/ipv6.h
@@ -1187,7 +1187,7 @@ int inet6_compat_ioctl(struct socket *sock, unsigned int cmd,
 		unsigned long arg);
 
 int inet6_hash_connect(struct inet_timewait_death_row *death_row,
-			      struct sock *sk);
+		       struct sock *sk, bool prev_inaddr_any);
 int inet6_sendmsg(struct socket *sock, struct msghdr *msg, size_t size);
 int inet6_recvmsg(struct socket *sock, struct msghdr *msg, size_t size,
 		  int flags);
diff --git a/net/dccp/ipv4.c b/net/dccp/ipv4.c
index da6e3b20cd75..37a8bc3ee49e 100644
--- a/net/dccp/ipv4.c
+++ b/net/dccp/ipv4.c
@@ -47,12 +47,13 @@ int dccp_v4_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len)
 	const struct sockaddr_in *usin = (struct sockaddr_in *)uaddr;
 	struct inet_sock *inet = inet_sk(sk);
 	struct dccp_sock *dp = dccp_sk(sk);
+	struct ip_options_rcu *inet_opt;
 	__be16 orig_sport, orig_dport;
+	bool prev_inaddr_any = false;
 	__be32 daddr, nexthop;
 	struct flowi4 *fl4;
 	struct rtable *rt;
 	int err;
-	struct ip_options_rcu *inet_opt;
 
 	dp->dccps_role = DCCP_ROLE_CLIENT;
 
@@ -89,8 +90,11 @@ int dccp_v4_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len)
 	if (inet_opt == NULL || !inet_opt->opt.srr)
 		daddr = fl4->daddr;
 
-	if (inet->inet_saddr == 0)
+	if (inet->inet_saddr == 0) {
 		inet->inet_saddr = fl4->saddr;
+		prev_inaddr_any = true;
+	}
+
 	sk_rcv_saddr_set(sk, inet->inet_saddr);
 	inet->inet_dport = usin->sin_port;
 	sk_daddr_set(sk, daddr);
@@ -105,7 +109,7 @@ int dccp_v4_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len)
 	 * complete initialization after this.
 	 */
 	dccp_set_state(sk, DCCP_REQUESTING);
-	err = inet_hash_connect(&dccp_death_row, sk);
+	err = inet_hash_connect(&dccp_death_row, sk, prev_inaddr_any);
 	if (err != 0)
 		goto failure;
 
diff --git a/net/dccp/ipv6.c b/net/dccp/ipv6.c
index fd44638ec16b..03013522acab 100644
--- a/net/dccp/ipv6.c
+++ b/net/dccp/ipv6.c
@@ -824,6 +824,7 @@ static int dccp_v6_connect(struct sock *sk, struct sockaddr *uaddr,
 	struct ipv6_pinfo *np = inet6_sk(sk);
 	struct dccp_sock *dp = dccp_sk(sk);
 	struct in6_addr *saddr = NULL, *final_p, final;
+	bool prev_inaddr_any = false;
 	struct ipv6_txoptions *opt;
 	struct flowi6 fl6;
 	struct dst_entry *dst;
@@ -936,6 +937,7 @@ static int dccp_v6_connect(struct sock *sk, struct sockaddr *uaddr,
 	if (saddr == NULL) {
 		saddr = &fl6.saddr;
 		sk->sk_v6_rcv_saddr = *saddr;
+		prev_inaddr_any = true;
 	}
 
 	/* set the source address */
@@ -951,7 +953,7 @@ static int dccp_v6_connect(struct sock *sk, struct sockaddr *uaddr,
 	inet->inet_dport = usin->sin6_port;
 
 	dccp_set_state(sk, DCCP_REQUESTING);
-	err = inet6_hash_connect(&dccp_death_row, sk);
+	err = inet6_hash_connect(&dccp_death_row, sk, prev_inaddr_any);
 	if (err)
 		goto late_failure;
 
diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c
index 93da9f783bec..ad627a99ff9d 100644
--- a/net/ipv4/af_inet.c
+++ b/net/ipv4/af_inet.c
@@ -1221,10 +1221,11 @@ static int inet_sk_reselect_saddr(struct sock *sk)
 	struct inet_sock *inet = inet_sk(sk);
 	__be32 old_saddr = inet->inet_saddr;
 	__be32 daddr = inet->inet_daddr;
+	struct ip_options_rcu *inet_opt;
 	struct flowi4 *fl4;
 	struct rtable *rt;
 	__be32 new_saddr;
-	struct ip_options_rcu *inet_opt;
+	int err;
 
 	inet_opt = rcu_dereference_protected(inet->inet_opt,
 					     lockdep_sock_is_held(sk));
@@ -1253,6 +1254,10 @@ static int inet_sk_reselect_saddr(struct sock *sk)
 
 	inet->inet_saddr = inet->inet_rcv_saddr = new_saddr;
 
+	err = inet_bhash2_update_saddr(sk);
+	if (err)
+		return err;
+
 	/*
 	 * XXX The only one ugly spot where we need to
 	 * XXX really change the sockets identity after
diff --git a/net/ipv4/inet_hashtables.c b/net/ipv4/inet_hashtables.c
index e8de5e699b3f..0b43a30b6363 100644
--- a/net/ipv4/inet_hashtables.c
+++ b/net/ipv4/inet_hashtables.c
@@ -826,6 +826,54 @@ inet_bind2_bucket_find(struct inet_hashinfo *hinfo, struct net *net,
 	return bhash2;
 }
 
+/* the lock for the socket's corresponding bhash entry must be held */
+int __inet_bhash2_update_saddr(struct sock *sk, struct inet_hashinfo *hinfo,
+			       struct net *net, int port, int l3mdev)
+{
+	struct inet_bind2_hashbucket *head2;
+	struct inet_bind2_bucket *tb2;
+
+	tb2 = inet_bind2_bucket_find(hinfo, net, port, l3mdev, sk,
+				     &head2);
+	if (!tb2) {
+		tb2 = inet_bind2_bucket_create(hinfo->bind2_bucket_cachep,
+					       net, head2, port, l3mdev, sk);
+		if (!tb2)
+			return -ENOMEM;
+	}
+
+	/* Remove the socket's old entry from bhash2 */
+	__sk_del_bind2_node(sk);
+
+	sk_add_bind2_node(sk, &tb2->owners);
+	inet_csk(sk)->icsk_bind2_hash = tb2;
+
+	return 0;
+}
+
+/* This should be called if/when a socket's rcv saddr changes after it has
+ * been binded.
+ */
+int inet_bhash2_update_saddr(struct sock *sk)
+{
+	struct inet_hashinfo *hinfo = sk->sk_prot->h.hashinfo;
+	int l3mdev = inet_sk_bound_l3mdev(sk);
+	struct inet_bind_hashbucket *head;
+	int port = inet_sk(sk)->inet_num;
+	struct net *net = sock_net(sk);
+	int err;
+
+	head = &hinfo->bhash[inet_bhashfn(net, port, hinfo->bhash_size)];
+
+	spin_lock_bh(&head->lock);
+
+	err = __inet_bhash2_update_saddr(sk, hinfo, net, port, l3mdev);
+
+	spin_unlock_bh(&head->lock);
+
+	return err;
+}
+
 /* RFC 6056 3.3.4.  Algorithm 4: Double-Hash Port Selection Algorithm
  * Note that we use 32bit integers (vs RFC 'short integers')
  * because 2^16 is not a multiple of num_ephemeral and this
@@ -840,7 +888,7 @@ inet_bind2_bucket_find(struct inet_hashinfo *hinfo, struct net *net,
 static u32 *table_perturb;
 
 int __inet_hash_connect(struct inet_timewait_death_row *death_row,
-		struct sock *sk, u64 port_offset,
+		struct sock *sk, u64 port_offset, bool prev_inaddr_any,
 		int (*check_established)(struct inet_timewait_death_row *,
 			struct sock *, __u16, struct inet_timewait_sock **))
 {
@@ -858,11 +906,24 @@ int __inet_hash_connect(struct inet_timewait_death_row *death_row,
 	int l3mdev;
 	u32 index;
 
+	l3mdev = inet_sk_bound_l3mdev(sk);
+
 	if (port) {
 		head = &hinfo->bhash[inet_bhashfn(net, port,
 						  hinfo->bhash_size)];
 		tb = inet_csk(sk)->icsk_bind_hash;
+
 		spin_lock_bh(&head->lock);
+
+		if (prev_inaddr_any) {
+			ret = __inet_bhash2_update_saddr(sk, hinfo, net, port,
+							 l3mdev);
+			if (ret) {
+				spin_unlock_bh(&head->lock);
+				return ret;
+			}
+		}
+
 		if (sk_head(&tb->owners) == sk && !sk->sk_bind_node.next) {
 			inet_ehash_nolisten(sk, NULL, NULL);
 			spin_unlock_bh(&head->lock);
@@ -875,8 +936,6 @@ int __inet_hash_connect(struct inet_timewait_death_row *death_row,
 		return ret;
 	}
 
-	l3mdev = inet_sk_bound_l3mdev(sk);
-
 	inet_get_local_port_range(net, &low, &high);
 	high++; /* [32768, 60999] -> [32768, 61000[ */
 	remaining = high - low;
@@ -987,13 +1046,13 @@ int __inet_hash_connect(struct inet_timewait_death_row *death_row,
  * Bind a port for a connect operation and hash it.
  */
 int inet_hash_connect(struct inet_timewait_death_row *death_row,
-		      struct sock *sk)
+		      struct sock *sk, bool prev_inaddr_any)
 {
 	u64 port_offset = 0;
 
 	if (!inet_sk(sk)->inet_num)
 		port_offset = inet_sk_port_offset(sk);
-	return __inet_hash_connect(death_row, sk, port_offset,
+	return __inet_hash_connect(death_row, sk, port_offset, prev_inaddr_any,
 				   __inet_check_established);
 }
 EXPORT_SYMBOL_GPL(inet_hash_connect);
diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index dac2650f3863..adf8d750933d 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -203,6 +203,7 @@ int tcp_v4_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len)
 	struct inet_sock *inet = inet_sk(sk);
 	struct tcp_sock *tp = tcp_sk(sk);
 	__be16 orig_sport, orig_dport;
+	bool prev_inaddr_any = false;
 	__be32 daddr, nexthop;
 	struct flowi4 *fl4;
 	struct rtable *rt;
@@ -246,8 +247,11 @@ int tcp_v4_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len)
 	if (!inet_opt || !inet_opt->opt.srr)
 		daddr = fl4->daddr;
 
-	if (!inet->inet_saddr)
+	if (!inet->inet_saddr) {
 		inet->inet_saddr = fl4->saddr;
+		prev_inaddr_any = true;
+	}
+
 	sk_rcv_saddr_set(sk, inet->inet_saddr);
 
 	if (tp->rx_opt.ts_recent_stamp && inet->inet_daddr != daddr) {
@@ -273,7 +277,7 @@ int tcp_v4_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len)
 	 * complete initialization after this.
 	 */
 	tcp_set_state(sk, TCP_SYN_SENT);
-	err = inet_hash_connect(tcp_death_row, sk);
+	err = inet_hash_connect(tcp_death_row, sk, prev_inaddr_any);
 	if (err)
 		goto failure;
 
diff --git a/net/ipv6/inet6_hashtables.c b/net/ipv6/inet6_hashtables.c
index 7d53d62783b1..c87c5933f3be 100644
--- a/net/ipv6/inet6_hashtables.c
+++ b/net/ipv6/inet6_hashtables.c
@@ -317,13 +317,13 @@ static u64 inet6_sk_port_offset(const struct sock *sk)
 }
 
 int inet6_hash_connect(struct inet_timewait_death_row *death_row,
-		       struct sock *sk)
+		       struct sock *sk, bool prev_inaddr_any)
 {
 	u64 port_offset = 0;
 
 	if (!inet_sk(sk)->inet_num)
 		port_offset = inet6_sk_port_offset(sk);
-	return __inet_hash_connect(death_row, sk, port_offset,
+	return __inet_hash_connect(death_row, sk, port_offset, prev_inaddr_any,
 				   __inet6_check_established);
 }
 EXPORT_SYMBOL_GPL(inet6_hash_connect);
diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c
index f37dd4aa91c6..81e3312c2a97 100644
--- a/net/ipv6/tcp_ipv6.c
+++ b/net/ipv6/tcp_ipv6.c
@@ -152,6 +152,7 @@ static int tcp_v6_connect(struct sock *sk, struct sockaddr *uaddr,
 	struct ipv6_pinfo *np = tcp_inet6_sk(sk);
 	struct tcp_sock *tp = tcp_sk(sk);
 	struct in6_addr *saddr = NULL, *final_p, final;
+	bool prev_inaddr_any = false;
 	struct ipv6_txoptions *opt;
 	struct flowi6 fl6;
 	struct dst_entry *dst;
@@ -289,6 +290,7 @@ static int tcp_v6_connect(struct sock *sk, struct sockaddr *uaddr,
 	if (!saddr) {
 		saddr = &fl6.saddr;
 		sk->sk_v6_rcv_saddr = *saddr;
+		prev_inaddr_any = true;
 	}
 
 	/* set the source address */
@@ -309,7 +311,7 @@ static int tcp_v6_connect(struct sock *sk, struct sockaddr *uaddr,
 
 	tcp_set_state(sk, TCP_SYN_SENT);
 	tcp_death_row = sock_net(sk)->ipv4.tcp_death_row;
-	err = inet6_hash_connect(tcp_death_row, sk);
+	err = inet6_hash_connect(tcp_death_row, sk, prev_inaddr_any);
 	if (err)
 		goto late_failure;
 
-- 
2.30.2


^ permalink raw reply related	[flat|nested] 6+ messages in thread

* [PATCH net-next v1 resend 2/2] selftests/net: Add sk_bind_sendto_listen test
  2022-06-01 20:14 [PATCH net-next v1 resend 0/2] Update bhash2 when socket's rcv saddr changes Joanne Koong
  2022-06-01 20:14 ` [PATCH net-next v1 resend 1/2] net: " Joanne Koong
@ 2022-06-01 20:14 ` Joanne Koong
  1 sibling, 0 replies; 6+ messages in thread
From: Joanne Koong @ 2022-06-01 20:14 UTC (permalink / raw)
  To: netdev, bpf; +Cc: edumazet, kafai, kuba, davem, pabeni, testing, Joanne Koong

From: Joanne Koong <joannelkoong@gmail.com>

This patch adds a new test called sk_bind_sendto_listen.

This test exercises the path where a socket's rcv saddr changes after it
has been added to the binding tables, and then a listen() on the socket
is invoked. The listen() should succeed.

This test is copied over from one of syzbot's tests:
https://syzkaller.appspot.com/x/repro.c?x=1673a38df00000

Signed-off-by: Joanne Koong <joannelkoong@gmail.com>
---
 tools/testing/selftests/net/.gitignore        |  1 +
 tools/testing/selftests/net/Makefile          |  1 +
 .../selftests/net/sk_bind_sendto_listen.c     | 82 +++++++++++++++++++
 3 files changed, 84 insertions(+)
 create mode 100644 tools/testing/selftests/net/sk_bind_sendto_listen.c

diff --git a/tools/testing/selftests/net/.gitignore b/tools/testing/selftests/net/.gitignore
index b984f8c8d523..69f1a2aafde4 100644
--- a/tools/testing/selftests/net/.gitignore
+++ b/tools/testing/selftests/net/.gitignore
@@ -38,3 +38,4 @@ ioam6_parser
 toeplitz
 cmsg_sender
 bind_bhash_test
+sk_bind_sendto_listen
diff --git a/tools/testing/selftests/net/Makefile b/tools/testing/selftests/net/Makefile
index 464df13831f2..45f4f57bf1f4 100644
--- a/tools/testing/selftests/net/Makefile
+++ b/tools/testing/selftests/net/Makefile
@@ -60,6 +60,7 @@ TEST_GEN_FILES += cmsg_sender
 TEST_GEN_FILES += stress_reuseport_listen
 TEST_PROGS += test_vxlan_vnifiltering.sh
 TEST_GEN_FILES += bind_bhash_test
+TEST_GEN_FILES += sk_bind_sendto_listen
 
 TEST_FILES := settings
 
diff --git a/tools/testing/selftests/net/sk_bind_sendto_listen.c b/tools/testing/selftests/net/sk_bind_sendto_listen.c
new file mode 100644
index 000000000000..3e09c5631997
--- /dev/null
+++ b/tools/testing/selftests/net/sk_bind_sendto_listen.c
@@ -0,0 +1,82 @@
+// SPDX-License-Identifier: GPL-2.0
+
+#define _GNU_SOURCE
+
+#include <arpa/inet.h>
+#include <error.h>
+#include <errno.h>
+#include <unistd.h>
+
+int main(void)
+{
+	int fd1, fd2, one = 1;
+	struct sockaddr_in6 bind_addr = {
+		.sin6_family = AF_INET6,
+		.sin6_port = htons(20000),
+		.sin6_flowinfo = htonl(0),
+		.sin6_addr = {},
+		.sin6_scope_id = 0,
+	};
+
+	inet_pton(AF_INET6, "::", &bind_addr.sin6_addr);
+
+	fd1 = socket(AF_INET6, SOCK_STREAM, IPPROTO_IP);
+	if (fd1 < 0) {
+		error(1, errno, "socket fd1");
+		return -1;
+	}
+
+	if (setsockopt(fd1, SOL_SOCKET, SO_REUSEADDR, &one, sizeof(one))) {
+		error(1, errno, "setsockopt(SO_REUSEADDR) fd1");
+		goto out_err1;
+	}
+
+	if (bind(fd1, (struct sockaddr *)&bind_addr, sizeof(bind_addr))) {
+		error(1, errno, "bind fd1");
+		goto out_err1;
+	}
+
+	if (sendto(fd1, NULL, 0, MSG_FASTOPEN, (struct sockaddr *)&bind_addr,
+		   sizeof(bind_addr))) {
+		error(1, errno, "sendto fd1");
+		goto out_err1;
+	}
+
+	fd2 = socket(AF_INET6, SOCK_STREAM, IPPROTO_IP);
+	if (fd2 < 0) {
+		error(1, errno, "socket fd2");
+		goto out_err1;
+	}
+
+	if (setsockopt(fd2, SOL_SOCKET, SO_REUSEADDR, &one, sizeof(one))) {
+		error(1, errno, "setsockopt(SO_REUSEADDR) fd2");
+		goto out_err2;
+	}
+
+	if (bind(fd2, (struct sockaddr *)&bind_addr, sizeof(bind_addr))) {
+		error(1, errno, "bind fd2");
+		goto out_err2;
+	}
+
+	if (sendto(fd2, NULL, 0, MSG_FASTOPEN, (struct sockaddr *)&bind_addr,
+		   sizeof(bind_addr)) != -1) {
+		error(1, errno, "sendto fd2");
+		goto out_err2;
+	}
+
+	if (listen(fd2, 0)) {
+		error(1, errno, "listen");
+		goto out_err2;
+	}
+
+	close(fd2);
+	close(fd1);
+	return 0;
+
+out_err2:
+	close(fd2);
+
+out_err1:
+	close(fd1);
+	return -1;
+}
-- 
2.30.2


^ permalink raw reply related	[flat|nested] 6+ messages in thread

* Re: [PATCH net-next v1 resend 1/2] net: Update bhash2 when socket's rcv saddr changes
  2022-06-01 20:14 ` [PATCH net-next v1 resend 1/2] net: " Joanne Koong
@ 2022-06-02  1:09   ` kernel test robot
  2022-06-02  1:13   ` Jakub Kicinski
  2022-06-02  6:50   ` kernel test robot
  2 siblings, 0 replies; 6+ messages in thread
From: kernel test robot @ 2022-06-02  1:09 UTC (permalink / raw)
  To: Joanne Koong, netdev, bpf
  Cc: llvm, kbuild-all, edumazet, kafai, kuba, davem, pabeni, testing,
	Joanne Koong, syzbot+015d756bbd1f8b5c8f09, Eric Dumazet

Hi Joanne,

Thank you for the patch! Perhaps something to improve:

[auto build test WARNING on net-next/master]

url:    https://github.com/intel-lab-lkp/linux/commits/Joanne-Koong/Update-bhash2-when-socket-s-rcv-saddr-changes/20220602-050108
base:   https://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next.git 7e062cda7d90543ac8c7700fc7c5527d0c0f22ad
config: hexagon-defconfig (https://download.01.org/0day-ci/archive/20220602/202206020958.MfddzQxe-lkp@intel.com/config)
compiler: clang version 15.0.0 (https://github.com/llvm/llvm-project c825abd6b0198fb088d9752f556a70705bc99dfd)
reproduce (this is a W=1 build):
        wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
        chmod +x ~/bin/make.cross
        # https://github.com/intel-lab-lkp/linux/commit/d4e9d3ab2c5210670fbe995cc8b13310a5aa6310
        git remote add linux-review https://github.com/intel-lab-lkp/linux
        git fetch --no-tags linux-review Joanne-Koong/Update-bhash2-when-socket-s-rcv-saddr-changes/20220602-050108
        git checkout d4e9d3ab2c5210670fbe995cc8b13310a5aa6310
        # save the config file
        mkdir build_dir && cp config build_dir/.config
        COMPILER_INSTALL_PATH=$HOME/0day COMPILER=clang make.cross W=1 O=build_dir ARCH=hexagon SHELL=/bin/bash net/ipv4/

If you fix the issue, kindly add following tag where applicable
Reported-by: kernel test robot <lkp@intel.com>

All warnings (new ones prefixed by >>):

>> net/ipv4/inet_hashtables.c:830:5: warning: no previous prototype for function '__inet_bhash2_update_saddr' [-Wmissing-prototypes]
   int __inet_bhash2_update_saddr(struct sock *sk, struct inet_hashinfo *hinfo,
       ^
   net/ipv4/inet_hashtables.c:830:1: note: declare 'static' if the function is not intended to be used outside of this translation unit
   int __inet_bhash2_update_saddr(struct sock *sk, struct inet_hashinfo *hinfo,
   ^
   static 
   1 warning generated.


vim +/__inet_bhash2_update_saddr +830 net/ipv4/inet_hashtables.c

   828	
   829	/* the lock for the socket's corresponding bhash entry must be held */
 > 830	int __inet_bhash2_update_saddr(struct sock *sk, struct inet_hashinfo *hinfo,
   831				       struct net *net, int port, int l3mdev)
   832	{
   833		struct inet_bind2_hashbucket *head2;
   834		struct inet_bind2_bucket *tb2;
   835	
   836		tb2 = inet_bind2_bucket_find(hinfo, net, port, l3mdev, sk,
   837					     &head2);
   838		if (!tb2) {
   839			tb2 = inet_bind2_bucket_create(hinfo->bind2_bucket_cachep,
   840						       net, head2, port, l3mdev, sk);
   841			if (!tb2)
   842				return -ENOMEM;
   843		}
   844	
   845		/* Remove the socket's old entry from bhash2 */
   846		__sk_del_bind2_node(sk);
   847	
   848		sk_add_bind2_node(sk, &tb2->owners);
   849		inet_csk(sk)->icsk_bind2_hash = tb2;
   850	
   851		return 0;
   852	}
   853	

-- 
0-DAY CI Kernel Test Service
https://01.org/lkp

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: [PATCH net-next v1 resend 1/2] net: Update bhash2 when socket's rcv saddr changes
  2022-06-01 20:14 ` [PATCH net-next v1 resend 1/2] net: " Joanne Koong
  2022-06-02  1:09   ` kernel test robot
@ 2022-06-02  1:13   ` Jakub Kicinski
  2022-06-02  6:50   ` kernel test robot
  2 siblings, 0 replies; 6+ messages in thread
From: Jakub Kicinski @ 2022-06-02  1:13 UTC (permalink / raw)
  To: Joanne Koong
  Cc: netdev, bpf, edumazet, kafai, davem, pabeni, testing,
	Joanne Koong, syzbot+015d756bbd1f8b5c8f09, Eric Dumazet

On Wed, 1 Jun 2022 13:14:33 -0700 Joanne Koong wrote:
> From: Joanne Koong <joannelkoong@gmail.com>
> 
> Commit d5a42de8bdbe ("net: Add a second bind table hashed by port and
> address") added a second bind table, bhash2, that hashes by a socket's port
> and rcv address.
> 
> However, there are two cases where the socket's rcv saddr can change
> after it has been binded:
> 
> 1) The case where there is a bind() call on "::" (IPADDR_ANY) and then
> a connect() call. The kernel will assign the socket an address when it
> handles the connect()
> 
> 2) In inet_sk_reselect_saddr(), which is called when rerouting fails
> when rebuilding the sk header (invoked by inet_sk_rebuild_header)
> 
> In these two cases, we need to update the bhash2 table by removing the
> entry for the old address, and adding a new entry reflecting the updated
> address.
> 
> Reported-by: syzbot+015d756bbd1f8b5c8f09@syzkaller.appspotmail.com
> Fixes: d5a42de8bdbe ("net: Add a second bind table hashed by port and address")
> Signed-off-by: Joanne Koong <joannelkoong@gmail.com>
> Reviewed-by: Eric Dumazet <edumzet@google.com>

This one needs to be static:

net/ipv4/inet_hashtables.c:830:5: warning: no previous prototype for ‘__inet_bhash2_update_saddr’ [-Wmissing-prototypes]
  830 | int __inet_bhash2_update_saddr(struct sock *sk, struct inet_hashinfo *hinfo,

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: [PATCH net-next v1 resend 1/2] net: Update bhash2 when socket's rcv saddr changes
  2022-06-01 20:14 ` [PATCH net-next v1 resend 1/2] net: " Joanne Koong
  2022-06-02  1:09   ` kernel test robot
  2022-06-02  1:13   ` Jakub Kicinski
@ 2022-06-02  6:50   ` kernel test robot
  2 siblings, 0 replies; 6+ messages in thread
From: kernel test robot @ 2022-06-02  6:50 UTC (permalink / raw)
  To: Joanne Koong, netdev, bpf
  Cc: kbuild-all, edumazet, kafai, kuba, davem, pabeni, testing,
	Joanne Koong, syzbot+015d756bbd1f8b5c8f09, Eric Dumazet

Hi Joanne,

Thank you for the patch! Perhaps something to improve:

[auto build test WARNING on net-next/master]

url:    https://github.com/intel-lab-lkp/linux/commits/Joanne-Koong/Update-bhash2-when-socket-s-rcv-saddr-changes/20220602-050108
base:   https://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next.git 7e062cda7d90543ac8c7700fc7c5527d0c0f22ad
config: i386-randconfig-s001 (https://download.01.org/0day-ci/archive/20220602/202206021442.Rb7gLaDQ-lkp@intel.com/config)
compiler: gcc-11 (Debian 11.3.0-1) 11.3.0
reproduce:
        # apt-get install sparse
        # sparse version: v0.6.4-14-g5a0004b5-dirty
        # https://github.com/intel-lab-lkp/linux/commit/d4e9d3ab2c5210670fbe995cc8b13310a5aa6310
        git remote add linux-review https://github.com/intel-lab-lkp/linux
        git fetch --no-tags linux-review Joanne-Koong/Update-bhash2-when-socket-s-rcv-saddr-changes/20220602-050108
        git checkout d4e9d3ab2c5210670fbe995cc8b13310a5aa6310
        # save the config file
        mkdir build_dir && cp config build_dir/.config
        make W=1 C=1 CF='-fdiagnostic-prefix -D__CHECK_ENDIAN__' O=build_dir ARCH=i386 SHELL=/bin/bash

If you fix the issue, kindly add following tag where applicable
Reported-by: kernel test robot <lkp@intel.com>


sparse warnings: (new ones prefixed by >>)
>> net/ipv4/inet_hashtables.c:830:5: sparse: sparse: symbol '__inet_bhash2_update_saddr' was not declared. Should it be static?

Please review and possibly fold the followup patch.

-- 
0-DAY CI Kernel Test Service
https://01.org/lkp

^ permalink raw reply	[flat|nested] 6+ messages in thread

end of thread, other threads:[~2022-06-02  6:51 UTC | newest]

Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2022-06-01 20:14 [PATCH net-next v1 resend 0/2] Update bhash2 when socket's rcv saddr changes Joanne Koong
2022-06-01 20:14 ` [PATCH net-next v1 resend 1/2] net: " Joanne Koong
2022-06-02  1:09   ` kernel test robot
2022-06-02  1:13   ` Jakub Kicinski
2022-06-02  6:50   ` kernel test robot
2022-06-01 20:14 ` [PATCH net-next v1 resend 2/2] selftests/net: Add sk_bind_sendto_listen test Joanne Koong

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).