[PATCH net-next 0/4] net: prot_inuse and sock_inuse cleanups

[PATCH net-next 0/4] net: prot_inuse and sock_inuse cleanups

[Eric Dumazet]

From: Eric Dumazet <edumazet@google.com>

Small series cleaning and optimizing sock_prot_inuse_add()
and sock_inuse_add().

Eric Dumazet (4):
  net: inline sock_prot_inuse_add()
  net: make sock_inuse_add() available
  net: merge net->core.prot_inuse and net->core.sock_inuse
  net: drop nopreempt requirement on sock_prot_inuse_add()

 include/net/netns/core.h |  1 -
 include/net/sock.h       | 27 +++++++++++++++++++++++----
 net/core/sock.c          | 33 +--------------------------------
 net/ieee802154/socket.c  |  4 ++--
 net/ipv4/raw.c           |  2 +-
 net/ipv6/ipv6_sockglue.c |  8 ++++----
 net/mptcp/subflow.c      |  4 +---
 net/netlink/af_netlink.c |  4 ----
 net/packet/af_packet.c   |  4 ----
 net/sctp/socket.c        |  5 -----
 net/smc/af_smc.c         |  2 +-
 net/unix/af_unix.c       |  4 ----
 net/xdp/xsk.c            |  4 ----
 13 files changed, 33 insertions(+), 69 deletions(-)

-- 
2.34.0.rc1.387.gb447b232ab-goog

[PATCH net-next 1/4] net: inline sock_prot_inuse_add()

[Eric Dumazet]

From: Eric Dumazet <edumazet@google.com>

sock_prot_inuse_add() is very small, we can inline it.

Signed-off-by: Eric Dumazet <edumazet@google.com>
---
 include/net/sock.h | 14 +++++++++++---
 net/core/sock.c    | 11 -----------
 2 files changed, 11 insertions(+), 14 deletions(-)

diff --git a/include/net/sock.h b/include/net/sock.h
index b32906e1ab55527b5418f203d3de05853863f166..3f08e9d55f0ceed4ec4593012e6c856b400fc33f 100644
--- a/include/net/sock.h
+++ b/include/net/sock.h
@@ -1419,13 +1419,21 @@ proto_memory_pressure(struct proto *prot)
 
 
 #ifdef CONFIG_PROC_FS
+#define PROTO_INUSE_NR	64	/* should be enough for the first time */
+struct prot_inuse {
+	int val[PROTO_INUSE_NR];
+};
 /* Called with local bh disabled */
-void sock_prot_inuse_add(struct net *net, struct proto *prot, int inc);
+static inline void sock_prot_inuse_add(const struct net *net,
+				       const struct proto *prot, int val)
+{
+	__this_cpu_add(net->core.prot_inuse->val[prot->inuse_idx], val);
+}
 int sock_prot_inuse_get(struct net *net, struct proto *proto);
 int sock_inuse_get(struct net *net);
 #else
-static inline void sock_prot_inuse_add(struct net *net, struct proto *prot,
-		int inc)
+static inline void sock_prot_inuse_add(const struct net *net,
+				       const struct proto *prot, int val)
 {
 }
 #endif
diff --git a/net/core/sock.c b/net/core/sock.c
index 8f2b2f2c0e7b1decdb4a5c8d86327ed7caa62c99..fac46efd31fd44b4105c6004b4491aa11e1ed67c 100644
--- a/net/core/sock.c
+++ b/net/core/sock.c
@@ -3532,19 +3532,8 @@ void sk_get_meminfo(const struct sock *sk, u32 *mem)
 }
 
 #ifdef CONFIG_PROC_FS
-#define PROTO_INUSE_NR	64	/* should be enough for the first time */
-struct prot_inuse {
-	int val[PROTO_INUSE_NR];
-};
-
 static DECLARE_BITMAP(proto_inuse_idx, PROTO_INUSE_NR);
 
-void sock_prot_inuse_add(struct net *net, struct proto *prot, int val)
-{
-	__this_cpu_add(net->core.prot_inuse->val[prot->inuse_idx], val);
-}
-EXPORT_SYMBOL_GPL(sock_prot_inuse_add);
-
 int sock_prot_inuse_get(struct net *net, struct proto *prot)
 {
 	int cpu, idx = prot->inuse_idx;
-- 
2.34.0.rc1.387.gb447b232ab-goog

[PATCH net-next 2/4] net: make sock_inuse_add() available

[Eric Dumazet]

From: Eric Dumazet <edumazet@google.com>

MPTCP hard codes it, let us instead provide this helper.

Signed-off-by: Eric Dumazet <edumazet@google.com>
---
 include/net/sock.h  | 10 ++++++++++
 net/core/sock.c     | 10 ----------
 net/mptcp/subflow.c |  4 +---
 3 files changed, 11 insertions(+), 13 deletions(-)

diff --git a/include/net/sock.h b/include/net/sock.h
index 3f08e9d55f0ceed4ec4593012e6c856b400fc33f..cdc7ebc049b41b00aa7c851a6f1df6e58bae8430 100644
--- a/include/net/sock.h
+++ b/include/net/sock.h
@@ -1429,6 +1429,12 @@ static inline void sock_prot_inuse_add(const struct net *net,
 {
 	__this_cpu_add(net->core.prot_inuse->val[prot->inuse_idx], val);
 }
+
+static inline void sock_inuse_add(const struct net *net, int val)
+{
+	this_cpu_add(*net->core.sock_inuse, val);
+}
+
 int sock_prot_inuse_get(struct net *net, struct proto *proto);
 int sock_inuse_get(struct net *net);
 #else
@@ -1436,6 +1442,10 @@ static inline void sock_prot_inuse_add(const struct net *net,
 				       const struct proto *prot, int val)
 {
 }
+
+static inline void sock_inuse_add(const struct net *net, int val)
+{
+}
 #endif
 
 
diff --git a/net/core/sock.c b/net/core/sock.c
index fac46efd31fd44b4105c6004b4491aa11e1ed67c..214c2e816c63dba9146557a622516e73c1da142e 100644
--- a/net/core/sock.c
+++ b/net/core/sock.c
@@ -144,8 +144,6 @@
 static DEFINE_MUTEX(proto_list_mutex);
 static LIST_HEAD(proto_list);
 
-static void sock_inuse_add(struct net *net, int val);
-
 /**
  * sk_ns_capable - General socket capability test
  * @sk: Socket to use a capability on or through
@@ -3546,11 +3544,6 @@ int sock_prot_inuse_get(struct net *net, struct proto *prot)
 }
 EXPORT_SYMBOL_GPL(sock_prot_inuse_get);
 
-static void sock_inuse_add(struct net *net, int val)
-{
-	this_cpu_add(*net->core.sock_inuse, val);
-}
-
 int sock_inuse_get(struct net *net)
 {
 	int cpu, res = 0;
@@ -3629,9 +3622,6 @@ static inline void release_proto_idx(struct proto *prot)
 {
 }
 
-static void sock_inuse_add(struct net *net, int val)
-{
-}
 #endif
 
 static void tw_prot_cleanup(struct timewait_sock_ops *twsk_prot)
diff --git a/net/mptcp/subflow.c b/net/mptcp/subflow.c
index 6172f380dfb763b43c6d996b4896215cad9c7d7b..49787a1d7b3467acdfe284fd1494ac4c4a6eaf5c 100644
--- a/net/mptcp/subflow.c
+++ b/net/mptcp/subflow.c
@@ -1534,9 +1534,7 @@ int mptcp_subflow_create_socket(struct sock *sk, struct socket **new_sock)
 	 */
 	sf->sk->sk_net_refcnt = 1;
 	get_net(net);
-#ifdef CONFIG_PROC_FS
-	this_cpu_add(*net->core.sock_inuse, 1);
-#endif
+	sock_inuse_add(net, 1);
 	err = tcp_set_ulp(sf->sk, "mptcp");
 	release_sock(sf->sk);
 
-- 
2.34.0.rc1.387.gb447b232ab-goog

[PATCH net-next 3/4] net: merge net->core.prot_inuse and net->core.sock_inuse

[Eric Dumazet]

From: Eric Dumazet <edumazet@google.com>

net->core.sock_inuse is a per cpu variable (int),
while net->core.prot_inuse is another per cpu variable
of 64 integers.

per cpu allocator tend to place them in very different places.

Grouping them together makes sense, since it makes
updates potentially faster, if hitting the same
cache line.

Signed-off-by: Eric Dumazet <edumazet@google.com>
---
 include/net/netns/core.h |  1 -
 include/net/sock.h       |  3 ++-
 net/core/sock.c          | 12 +-----------
 3 files changed, 3 insertions(+), 13 deletions(-)

diff --git a/include/net/netns/core.h b/include/net/netns/core.h
index 36c2d998a43c015c4b917428ec33f5839cde89b1..552bc25b19335e2ee07effa2550bbe5944eb33aa 100644
--- a/include/net/netns/core.h
+++ b/include/net/netns/core.h
@@ -12,7 +12,6 @@ struct netns_core {
 	int	sysctl_somaxconn;
 
 #ifdef CONFIG_PROC_FS
-	int __percpu *sock_inuse;
 	struct prot_inuse __percpu *prot_inuse;
 #endif
 };
diff --git a/include/net/sock.h b/include/net/sock.h
index cdc7ebc049b41b00aa7c851a6f1df6e58bae8430..fefffeb1cc3d5a11615afbc34e5cd7521bd6f502 100644
--- a/include/net/sock.h
+++ b/include/net/sock.h
@@ -1421,6 +1421,7 @@ proto_memory_pressure(struct proto *prot)
 #ifdef CONFIG_PROC_FS
 #define PROTO_INUSE_NR	64	/* should be enough for the first time */
 struct prot_inuse {
+	int all;
 	int val[PROTO_INUSE_NR];
 };
 /* Called with local bh disabled */
@@ -1432,7 +1433,7 @@ static inline void sock_prot_inuse_add(const struct net *net,
 
 static inline void sock_inuse_add(const struct net *net, int val)
 {
-	this_cpu_add(*net->core.sock_inuse, val);
+	this_cpu_add(net->core.prot_inuse->all, val);
 }
 
 int sock_prot_inuse_get(struct net *net, struct proto *proto);
diff --git a/net/core/sock.c b/net/core/sock.c
index 214c2e816c63dba9146557a622516e73c1da142e..2f58e4d3e76296280aece28314b8695d0d40cf02 100644
--- a/net/core/sock.c
+++ b/net/core/sock.c
@@ -3549,7 +3549,7 @@ int sock_inuse_get(struct net *net)
 	int cpu, res = 0;
 
 	for_each_possible_cpu(cpu)
-		res += *per_cpu_ptr(net->core.sock_inuse, cpu);
+		res += per_cpu_ptr(net->core.prot_inuse, cpu)->all;
 
 	return res;
 }
@@ -3561,22 +3561,12 @@ static int __net_init sock_inuse_init_net(struct net *net)
 	net->core.prot_inuse = alloc_percpu(struct prot_inuse);
 	if (net->core.prot_inuse == NULL)
 		return -ENOMEM;
-
-	net->core.sock_inuse = alloc_percpu(int);
-	if (net->core.sock_inuse == NULL)
-		goto out;
-
 	return 0;
-
-out:
-	free_percpu(net->core.prot_inuse);
-	return -ENOMEM;
 }
 
 static void __net_exit sock_inuse_exit_net(struct net *net)
 {
 	free_percpu(net->core.prot_inuse);
-	free_percpu(net->core.sock_inuse);
 }
 
 static struct pernet_operations net_inuse_ops = {
-- 
2.34.0.rc1.387.gb447b232ab-goog

[PATCH net-next 4/4] net: drop nopreempt requirement on sock_prot_inuse_add()

[Eric Dumazet]

From: Eric Dumazet <edumazet@google.com>

This is distracting really, let's make this simpler,
because many callers had to take care of this
by themselves, even if on x86 this adds more
code than really needed.

Signed-off-by: Eric Dumazet <edumazet@google.com>
---
 include/net/sock.h       | 4 ++--
 net/ieee802154/socket.c  | 4 ++--
 net/ipv4/raw.c           | 2 +-
 net/ipv6/ipv6_sockglue.c | 8 ++++----
 net/netlink/af_netlink.c | 4 ----
 net/packet/af_packet.c   | 4 ----
 net/sctp/socket.c        | 5 -----
 net/smc/af_smc.c         | 2 +-
 net/unix/af_unix.c       | 4 ----
 net/xdp/xsk.c            | 4 ----
 10 files changed, 10 insertions(+), 31 deletions(-)

diff --git a/include/net/sock.h b/include/net/sock.h
index fefffeb1cc3d5a11615afbc34e5cd7521bd6f502..7e8330599f633c5029e15e9c5e859d183ae87127 100644
--- a/include/net/sock.h
+++ b/include/net/sock.h
@@ -1424,11 +1424,11 @@ struct prot_inuse {
 	int all;
 	int val[PROTO_INUSE_NR];
 };
-/* Called with local bh disabled */
+
 static inline void sock_prot_inuse_add(const struct net *net,
 				       const struct proto *prot, int val)
 {
-	__this_cpu_add(net->core.prot_inuse->val[prot->inuse_idx], val);
+	this_cpu_add(net->core.prot_inuse->val[prot->inuse_idx], val);
 }
 
 static inline void sock_inuse_add(const struct net *net, int val)
diff --git a/net/ieee802154/socket.c b/net/ieee802154/socket.c
index 7bb9ef35c57076252c700617dad4cc6cf275987a..3b2366a88c3ccbae787669c58a2827048aa4552c 100644
--- a/net/ieee802154/socket.c
+++ b/net/ieee802154/socket.c
@@ -174,8 +174,8 @@ static int raw_hash(struct sock *sk)
 {
 	write_lock_bh(&raw_lock);
 	sk_add_node(sk, &raw_head);
-	sock_prot_inuse_add(sock_net(sk), sk->sk_prot, 1);
 	write_unlock_bh(&raw_lock);
+	sock_prot_inuse_add(sock_net(sk), sk->sk_prot, 1);
 
 	return 0;
 }
@@ -453,8 +453,8 @@ static int dgram_hash(struct sock *sk)
 {
 	write_lock_bh(&dgram_lock);
 	sk_add_node(sk, &dgram_head);
-	sock_prot_inuse_add(sock_net(sk), sk->sk_prot, 1);
 	write_unlock_bh(&dgram_lock);
+	sock_prot_inuse_add(sock_net(sk), sk->sk_prot, 1);
 
 	return 0;
 }
diff --git a/net/ipv4/raw.c b/net/ipv4/raw.c
index bb446e60cf58057b448f094b4d6f48d6e91d113c..3910bb08a986f37984ad1a5eebca14c706deb315 100644
--- a/net/ipv4/raw.c
+++ b/net/ipv4/raw.c
@@ -99,8 +99,8 @@ int raw_hash_sk(struct sock *sk)
 
 	write_lock_bh(&h->lock);
 	sk_add_node(sk, head);
+	write_unlock_bh(&h->lock);
 	sock_prot_inuse_add(sock_net(sk), sk->sk_prot, 1);
-	write_unlock_bh(&h->lock);
 
 	return 0;
 }
diff --git a/net/ipv6/ipv6_sockglue.c b/net/ipv6/ipv6_sockglue.c
index 41efca817db4228f265235a471449a3790075ce7..04683ad6d3df83bbe8ff6c10885b3bf16483688c 100644
--- a/net/ipv6/ipv6_sockglue.c
+++ b/net/ipv6/ipv6_sockglue.c
@@ -471,10 +471,10 @@ static int do_ipv6_setsockopt(struct sock *sk, int level, int optname,
 
 			if (sk->sk_protocol == IPPROTO_TCP) {
 				struct inet_connection_sock *icsk = inet_csk(sk);
-				local_bh_disable();
+
 				sock_prot_inuse_add(net, sk->sk_prot, -1);
 				sock_prot_inuse_add(net, &tcp_prot, 1);
-				local_bh_enable();
+
 				sk->sk_prot = &tcp_prot;
 				icsk->icsk_af_ops = &ipv4_specific;
 				sk->sk_socket->ops = &inet_stream_ops;
@@ -485,10 +485,10 @@ static int do_ipv6_setsockopt(struct sock *sk, int level, int optname,
 
 				if (sk->sk_protocol == IPPROTO_UDPLITE)
 					prot = &udplite_prot;
-				local_bh_disable();
+
 				sock_prot_inuse_add(net, sk->sk_prot, -1);
 				sock_prot_inuse_add(net, prot, 1);
-				local_bh_enable();
+
 				sk->sk_prot = prot;
 				sk->sk_socket->ops = &inet_dgram_ops;
 				sk->sk_family = PF_INET;
diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c
index 4c575324a98528bec4188acf27eecc2f98ae5e0a..1a19d179e913a3cf973aebab54af71b665345e0f 100644
--- a/net/netlink/af_netlink.c
+++ b/net/netlink/af_netlink.c
@@ -707,9 +707,7 @@ static int netlink_create(struct net *net, struct socket *sock, int protocol,
 	if (err < 0)
 		goto out_module;
 
-	local_bh_disable();
 	sock_prot_inuse_add(net, &netlink_proto, 1);
-	local_bh_enable();
 
 	nlk = nlk_sk(sock->sk);
 	nlk->module = module;
@@ -809,9 +807,7 @@ static int netlink_release(struct socket *sock)
 		netlink_table_ungrab();
 	}
 
-	local_bh_disable();
 	sock_prot_inuse_add(sock_net(sk), &netlink_proto, -1);
-	local_bh_enable();
 	call_rcu(&nlk->rcu, deferred_put_nlk_sk);
 	return 0;
 }
diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c
index 46943a18a10d5413db57955dbd24302af7ef1d97..a1ffdb48cc474dcf91bddfd1ab96386a89c20375 100644
--- a/net/packet/af_packet.c
+++ b/net/packet/af_packet.c
@@ -3102,9 +3102,7 @@ static int packet_release(struct socket *sock)
 	sk_del_node_init_rcu(sk);
 	mutex_unlock(&net->packet.sklist_lock);
 
-	preempt_disable();
 	sock_prot_inuse_add(net, sk->sk_prot, -1);
-	preempt_enable();
 
 	spin_lock(&po->bind_lock);
 	unregister_prot_hook(sk, false);
@@ -3368,9 +3366,7 @@ static int packet_create(struct net *net, struct socket *sock, int protocol,
 	sk_add_node_tail_rcu(sk, &net->packet.sklist);
 	mutex_unlock(&net->packet.sklist_lock);
 
-	preempt_disable();
 	sock_prot_inuse_add(net, &packet_proto, 1);
-	preempt_enable();
 
 	return 0;
 out2:
diff --git a/net/sctp/socket.c b/net/sctp/socket.c
index 33391254fa82b2277551387bfcdf95259bea7e0d..055a6d3ec6e2e90e31a21c4303e912168c8afe5b 100644
--- a/net/sctp/socket.c
+++ b/net/sctp/socket.c
@@ -5068,12 +5068,9 @@ static int sctp_init_sock(struct sock *sk)
 
 	SCTP_DBG_OBJCNT_INC(sock);
 
-	local_bh_disable();
 	sk_sockets_allocated_inc(sk);
 	sock_prot_inuse_add(net, sk->sk_prot, 1);
 
-	local_bh_enable();
-
 	return 0;
 }
 
@@ -5099,10 +5096,8 @@ static void sctp_destroy_sock(struct sock *sk)
 		list_del(&sp->auto_asconf_list);
 	}
 	sctp_endpoint_free(sp->ep);
-	local_bh_disable();
 	sk_sockets_allocated_dec(sk);
 	sock_prot_inuse_add(sock_net(sk), sk->sk_prot, -1);
-	local_bh_enable();
 }
 
 /* Triggered when there are no references on the socket anymore */
diff --git a/net/smc/af_smc.c b/net/smc/af_smc.c
index 59284da9116d787b95704be7fda6d62e351e36aa..ff5cd0c30741c0b562e66ce49a60bbc8d260f3b5 100644
--- a/net/smc/af_smc.c
+++ b/net/smc/af_smc.c
@@ -89,8 +89,8 @@ int smc_hash_sk(struct sock *sk)
 
 	write_lock_bh(&h->lock);
 	sk_add_node(sk, head);
+	write_unlock_bh(&h->lock);
 	sock_prot_inuse_add(sock_net(sk), sk->sk_prot, 1);
-	write_unlock_bh(&h->lock);
 
 	return 0;
 }
diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c
index 78e08e82c08c423ff2112fcef9e27995a6d39984..54e5553a150ec58d07b61a15f4823bc0622a0b6e 100644
--- a/net/unix/af_unix.c
+++ b/net/unix/af_unix.c
@@ -522,9 +522,7 @@ static void unix_sock_destructor(struct sock *sk)
 		unix_release_addr(u->addr);
 
 	atomic_long_dec(&unix_nr_socks);
-	local_bh_disable();
 	sock_prot_inuse_add(sock_net(sk), sk->sk_prot, -1);
-	local_bh_enable();
 #ifdef UNIX_REFCNT_DEBUG
 	pr_debug("UNIX %p is destroyed, %ld are still alive.\n", sk,
 		atomic_long_read(&unix_nr_socks));
@@ -889,9 +887,7 @@ static struct sock *unix_create1(struct net *net, struct socket *sock, int kern,
 	memset(&u->scm_stat, 0, sizeof(struct scm_stat));
 	unix_insert_socket(unix_sockets_unbound(sk), sk);
 
-	local_bh_disable();
 	sock_prot_inuse_add(sock_net(sk), sk->sk_prot, 1);
-	local_bh_enable();
 
 	return sk;
 
diff --git a/net/xdp/xsk.c b/net/xdp/xsk.c
index f16074eb53c72a7040421d23028d5762e0c5658d..28ef3f4465ae9e63bc5326311f923260b88f0fe8 100644
--- a/net/xdp/xsk.c
+++ b/net/xdp/xsk.c
@@ -794,9 +794,7 @@ static int xsk_release(struct socket *sock)
 	sk_del_node_init_rcu(sk);
 	mutex_unlock(&net->xdp.lock);
 
-	local_bh_disable();
 	sock_prot_inuse_add(net, sk->sk_prot, -1);
-	local_bh_enable();
 
 	xsk_delete_from_maps(xs);
 	mutex_lock(&xs->mutex);
@@ -1396,9 +1394,7 @@ static int xsk_create(struct net *net, struct socket *sock, int protocol,
 	sk_add_node_rcu(sk, &net->xdp.list);
 	mutex_unlock(&net->xdp.lock);
 
-	local_bh_disable();
 	sock_prot_inuse_add(net, &xsk_proto, 1);
-	local_bh_enable();
 
 	return 0;
 }
-- 
2.34.0.rc1.387.gb447b232ab-goog