Merge branch 'rtnetlink-per-netns-rtnl'

Kuniyuki Iwashima says:

====================
rtnetlink: Per-netns RTNL.

rtnl_lock() is a "Big Kernel Lock" in the networking slow path and
serialised all rtnetlink requests until 4.13.

Since RTNL_FLAG_DOIT_UNLOCKED and RTNL_FLAG_DUMP_UNLOCKED have been
introduced in 4.14 and 6.9, respectively, rtnetlink message handlers
are ready to be converted to RTNL-less/free.

15 out of 44 dumpit()s have been converted to RCU so far, and the
progress is pretty good.  We can now dump various major network
resources without RTNL.

12 out of 87 doit()s have been converted, but most of the converted
doit()s are also on the reader side of RTNL; their message types are
RTM_GET*.

So, most of RTM_(NEW|DEL|SET)* operations are still serialised by RTNL.

For example, one of our services creates 2K netns and a small number
of network interfaces in each netns that require too many writer-side
rtnetlink requests, and setting up a single host takes 10+ minutes.

RTNL is still a huge pain for network configuration paths, and we need
more granular locking, given converting all doit()s would be unfeasible.

Actually, most RTNL users do not need to freeze multiple netns, and such
users can be protected by per-netns RTNL mutex.  The exceptions would be
RTM_NEWLINK, RTM_DELLINK, and RTM_SETLINK.  (See [0] and [1])

This series is the first step of the per-netns RTNL conversion that
gradually replaces rtnl_lock() with rtnl_net_lock(net) under
CONFIG_DEBUG_NET_SMALL_RTNL.

[0]: https://netdev.bots.linux.dev/netconf/2024/index.html
[1]: https://lpc.events/event/18/contributions/1959/
====================

Link: https://patch.msgid.link/20241004221031.77743-1-kuniyu@amazon.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>

7 files changed, 289 insertions, 7 deletions

diff --git a/include/linux/rtnetlink.h b/include/linux/rtnetlink.h
index a7da7dfc06a2..8468a4ce8510 100644
--- a/include/linux/rtnetlink.h
+++ b/include/linux/rtnetlink.h
@@ -7,7 +7,6 @@
 #include <linux/netdevice.h>
 #include <linux/wait.h>
 #include <linux/refcount.h>
-#include <linux/cleanup.h>
 #include <uapi/linux/rtnetlink.h>
 
 extern int rtnetlink_send(struct sk_buff *skb, struct net *net, u32 pid, u32 group, int echo);
@@ -47,13 +46,15 @@ extern int rtnl_is_locked(void);
 extern int rtnl_lock_killable(void);
 extern bool refcount_dec_and_rtnl_lock(refcount_t *r);
 
-DEFINE_LOCK_GUARD_0(rtnl, rtnl_lock(), rtnl_unlock())
-
 extern wait_queue_head_t netdev_unregistering_wq;
 extern atomic_t dev_unreg_count;
 extern struct rw_semaphore pernet_ops_rwsem;
 extern struct rw_semaphore net_rwsem;
 
+#define ASSERT_RTNL() \
+	WARN_ONCE(!rtnl_is_locked(), \
+		  "RTNL: assertion failed at %s (%d)\n", __FILE__,  __LINE__)
+
 #ifdef CONFIG_PROVE_LOCKING
 extern bool lockdep_rtnl_is_held(void);
 #else
@@ -95,6 +96,64 @@ static inline bool lockdep_rtnl_is_held(void)
 #define rcu_replace_pointer_rtnl(rp, p)			\
 	rcu_replace_pointer(rp, p, lockdep_rtnl_is_held())
 
+#ifdef CONFIG_DEBUG_NET_SMALL_RTNL
+void __rtnl_net_lock(struct net *net);
+void __rtnl_net_unlock(struct net *net);
+void rtnl_net_lock(struct net *net);
+void rtnl_net_unlock(struct net *net);
+int rtnl_net_lock_cmp_fn(const struct lockdep_map *a, const struct lockdep_map *b);
+
+bool rtnl_net_is_locked(struct net *net);
+
+#define ASSERT_RTNL_NET(net)						\
+	WARN_ONCE(!rtnl_net_is_locked(net),				\
+		  "RTNL_NET: assertion failed at %s (%d)\n",		\
+		  __FILE__,  __LINE__)
+
+bool lockdep_rtnl_net_is_held(struct net *net);
+
+#define rcu_dereference_rtnl_net(net, p)				\
+	rcu_dereference_check(p, lockdep_rtnl_net_is_held(net))
+#define rtnl_net_dereference(net, p)					\
+	rcu_dereference_protected(p, lockdep_rtnl_net_is_held(net))
+#define rcu_replace_pointer_rtnl_net(net, rp, p)			\
+	rcu_replace_pointer(rp, p, lockdep_rtnl_net_is_held(net))
+#else
+static inline void __rtnl_net_lock(struct net *net) {}
+static inline void __rtnl_net_unlock(struct net *net) {}
+
+static inline void rtnl_net_lock(struct net *net)
+{
+	rtnl_lock();
+}
+
+static inline void rtnl_net_unlock(struct net *net)
+{
+	rtnl_unlock();
+}
+
+static inline void ASSERT_RTNL_NET(struct net *net)
+{
+	ASSERT_RTNL();
+}
+
+static inline void *rcu_dereference_rtnl_net(struct net *net, void *p)
+{
+	return rcu_dereference_rtnl(p);
+}
+
+static inline void *rtnl_net_dereference(struct net *net, void *p)
+{
+	return rtnl_dereference(p);
+}
+
+static inline void *rcu_replace_pointer_rtnl_net(struct net *net,
+						 void *rp, void *p)
+{
+	return rcu_replace_pointer_rtnl(rp, p);
+}
+#endif
+
 static inline struct netdev_queue *dev_ingress_queue(struct net_device *dev)
 {
 	return rtnl_dereference(dev->ingress_queue);
@@ -122,10 +181,6 @@ void rtnetlink_init(void);
 void __rtnl_unlock(void);
 void rtnl_kfree_skbs(struct sk_buff *head, struct sk_buff *tail);
 
-#define ASSERT_RTNL() \
-	WARN_ONCE(!rtnl_is_locked(), \
-		  "RTNL: assertion failed at %s (%d)\n", __FILE__,  __LINE__)
-
 extern int ndo_dflt_fdb_dump(struct sk_buff *skb,
 			     struct netlink_callback *cb,
 			     struct net_device *dev,
diff --git a/include/net/net_namespace.h b/include/net/net_namespace.h
index e67b483cc8bb..873c0f9fdac6 100644
--- a/include/net/net_namespace.h
+++ b/include/net/net_namespace.h
@@ -188,6 +188,10 @@ struct net {
 #if IS_ENABLED(CONFIG_SMC)
 	struct netns_smc	smc;
 #endif
+#ifdef CONFIG_DEBUG_NET_SMALL_RTNL
+	/* Move to a better place when the config guard is removed. */
+	struct mutex		rtnl_mutex;
+#endif
 } __randomize_layout;
 
 #include <linux/seq_file_net.h>
diff --git a/net/Kconfig.debug b/net/Kconfig.debug
index 5e3fffe707dd..277fab8c4d77 100644
--- a/net/Kconfig.debug
+++ b/net/Kconfig.debug
@@ -24,3 +24,18 @@ config DEBUG_NET
 	help
 	  Enable extra sanity checks in networking.
 	  This is mostly used by fuzzers, but is safe to select.
+
+config DEBUG_NET_SMALL_RTNL
+	bool "Add extra per-netns mutex inside RTNL"
+	depends on DEBUG_KERNEL && NET && LOCK_DEBUGGING_SUPPORT
+	select PROVE_LOCKING
+	default n
+	help
+	  rtnl_lock() is being replaced with rtnl_net_lock() that
+	  acquires the global RTNL and a small per-netns RTNL mutex.
+
+	  During the conversion, rtnl_net_lock() just adds an extra
+	  mutex in every RTNL scope and slows down the operations.
+
+	  Once the conversion completes, rtnl_lock() will be removed
+	  and rtnetlink will gain per-netns scalability.
diff --git a/net/core/Makefile b/net/core/Makefile
index c3ebbaf9c81e..5a72a87ee0f1 100644
--- a/net/core/Makefile
+++ b/net/core/Makefile
@@ -45,3 +45,4 @@ obj-$(CONFIG_BPF_SYSCALL) += bpf_sk_storage.o
 obj-$(CONFIG_OF)	+= of_net.o
 obj-$(CONFIG_NET_TEST) += net_test.o
 obj-$(CONFIG_NET_DEVMEM) += devmem.o
+obj-$(CONFIG_DEBUG_NET_SMALL_RTNL) += rtnl_net_debug.o
diff --git a/net/core/net_namespace.c b/net/core/net_namespace.c
index e39479f1c9a4..105e3cd26763 100644
--- a/net/core/net_namespace.c
+++ b/net/core/net_namespace.c
@@ -334,6 +334,12 @@ static __net_init void preinit_net(struct net *net, struct user_namespace *user_
 	idr_init(&net->netns_ids);
 	spin_lock_init(&net->nsid_lock);
 	mutex_init(&net->ipv4.ra_mutex);
+
+#ifdef CONFIG_DEBUG_NET_SMALL_RTNL
+	mutex_init(&net->rtnl_mutex);
+	lock_set_cmp_fn(&net->rtnl_mutex, rtnl_net_lock_cmp_fn, NULL);
+#endif
+
 	preinit_net_sysctl(net);
 }
 
diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c
index 682d8d3127db..6d68247aea70 100644
--- a/net/core/rtnetlink.c
+++ b/net/core/rtnetlink.c
@@ -179,6 +179,76 @@ bool lockdep_rtnl_is_held(void)
 EXPORT_SYMBOL(lockdep_rtnl_is_held);
 #endif /* #ifdef CONFIG_PROVE_LOCKING */
 
+#ifdef CONFIG_DEBUG_NET_SMALL_RTNL
+void __rtnl_net_lock(struct net *net)
+{
+	ASSERT_RTNL();
+
+	mutex_lock(&net->rtnl_mutex);
+}
+EXPORT_SYMBOL(__rtnl_net_lock);
+
+void __rtnl_net_unlock(struct net *net)
+{
+	ASSERT_RTNL();
+
+	mutex_unlock(&net->rtnl_mutex);
+}
+EXPORT_SYMBOL(__rtnl_net_unlock);
+
+void rtnl_net_lock(struct net *net)
+{
+	rtnl_lock();
+	__rtnl_net_lock(net);
+}
+EXPORT_SYMBOL(rtnl_net_lock);
+
+void rtnl_net_unlock(struct net *net)
+{
+	__rtnl_net_unlock(net);
+	rtnl_unlock();
+}
+EXPORT_SYMBOL(rtnl_net_unlock);
+
+static int rtnl_net_cmp_locks(const struct net *net_a, const struct net *net_b)
+{
+	if (net_eq(net_a, net_b))
+		return 0;
+
+	/* always init_net first */
+	if (net_eq(net_a, &init_net))
+		return -1;
+
+	if (net_eq(net_b, &init_net))
+		return 1;
+
+	/* otherwise lock in ascending order */
+	return net_a < net_b ? -1 : 1;
+}
+
+int rtnl_net_lock_cmp_fn(const struct lockdep_map *a, const struct lockdep_map *b)
+{
+	const struct net *net_a, *net_b;
+
+	net_a = container_of(a, struct net, rtnl_mutex.dep_map);
+	net_b = container_of(b, struct net, rtnl_mutex.dep_map);
+
+	return rtnl_net_cmp_locks(net_a, net_b);
+}
+
+bool rtnl_net_is_locked(struct net *net)
+{
+	return rtnl_is_locked() && mutex_is_locked(&net->rtnl_mutex);
+}
+EXPORT_SYMBOL(rtnl_net_is_locked);
+
+bool lockdep_rtnl_net_is_held(struct net *net)
+{
+	return lockdep_rtnl_is_held() && lockdep_is_held(&net->rtnl_mutex);
+}
+EXPORT_SYMBOL(lockdep_rtnl_net_is_held);
+#endif
+
 static struct rtnl_link __rcu *__rcu *rtnl_msg_handlers[RTNL_FAMILY_MAX + 1];
 
 static inline int rtm_msgindex(int msgtype)
diff --git a/net/core/rtnl_net_debug.c b/net/core/rtnl_net_debug.c
new file mode 100644
index 000000000000..e90a32242e22
--- /dev/null
+++ b/net/core/rtnl_net_debug.c
@@ -0,0 +1,131 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/* Copyright Amazon.com Inc. or its affiliates. */
+
+#include <linux/init.h>
+#include <linux/netdevice.h>
+#include <linux/notifier.h>
+#include <linux/rtnetlink.h>
+#include <net/net_namespace.h>
+#include <net/netns/generic.h>
+
+static int rtnl_net_debug_event(struct notifier_block *nb,
+				unsigned long event, void *ptr)
+{
+	struct net_device *dev = netdev_notifier_info_to_dev(ptr);
+	struct net *net = dev_net(dev);
+	enum netdev_cmd cmd = event;
+
+	/* Keep enum and don't add default to trigger -Werror=switch */
+	switch (cmd) {
+	case NETDEV_UP:
+	case NETDEV_DOWN:
+	case NETDEV_REBOOT:
+	case NETDEV_CHANGE:
+	case NETDEV_REGISTER:
+	case NETDEV_UNREGISTER:
+	case NETDEV_CHANGEMTU:
+	case NETDEV_CHANGEADDR:
+	case NETDEV_PRE_CHANGEADDR:
+	case NETDEV_GOING_DOWN:
+	case NETDEV_CHANGENAME:
+	case NETDEV_FEAT_CHANGE:
+	case NETDEV_BONDING_FAILOVER:
+	case NETDEV_PRE_UP:
+	case NETDEV_PRE_TYPE_CHANGE:
+	case NETDEV_POST_TYPE_CHANGE:
+	case NETDEV_POST_INIT:
+	case NETDEV_PRE_UNINIT:
+	case NETDEV_RELEASE:
+	case NETDEV_NOTIFY_PEERS:
+	case NETDEV_JOIN:
+	case NETDEV_CHANGEUPPER:
+	case NETDEV_RESEND_IGMP:
+	case NETDEV_PRECHANGEMTU:
+	case NETDEV_CHANGEINFODATA:
+	case NETDEV_BONDING_INFO:
+	case NETDEV_PRECHANGEUPPER:
+	case NETDEV_CHANGELOWERSTATE:
+	case NETDEV_UDP_TUNNEL_PUSH_INFO:
+	case NETDEV_UDP_TUNNEL_DROP_INFO:
+	case NETDEV_CHANGE_TX_QUEUE_LEN:
+	case NETDEV_CVLAN_FILTER_PUSH_INFO:
+	case NETDEV_CVLAN_FILTER_DROP_INFO:
+	case NETDEV_SVLAN_FILTER_PUSH_INFO:
+	case NETDEV_SVLAN_FILTER_DROP_INFO:
+	case NETDEV_OFFLOAD_XSTATS_ENABLE:
+	case NETDEV_OFFLOAD_XSTATS_DISABLE:
+	case NETDEV_OFFLOAD_XSTATS_REPORT_USED:
+	case NETDEV_OFFLOAD_XSTATS_REPORT_DELTA:
+	case NETDEV_XDP_FEAT_CHANGE:
+		ASSERT_RTNL();
+		break;
+
+	/* Once an event fully supports RTNL_NET, move it here
+	 * and remove "if (0)" below.
+	 *
+	 * case NETDEV_XXX:
+	 *	ASSERT_RTNL_NET(net);
+	 *	break;
+	 */
+	}
+
+	/* Just to avoid unused-variable error for dev and net. */
+	if (0)
+		ASSERT_RTNL_NET(net);
+
+	return NOTIFY_DONE;
+}
+
+static int rtnl_net_debug_net_id;
+
+static int __net_init rtnl_net_debug_net_init(struct net *net)
+{
+	struct notifier_block *nb;
+
+	nb = net_generic(net, rtnl_net_debug_net_id);
+	nb->notifier_call = rtnl_net_debug_event;
+
+	return register_netdevice_notifier_net(net, nb);
+}
+
+static void __net_exit rtnl_net_debug_net_exit(struct net *net)
+{
+	struct notifier_block *nb;
+
+	nb = net_generic(net, rtnl_net_debug_net_id);
+	unregister_netdevice_notifier_net(net, nb);
+}
+
+static struct pernet_operations rtnl_net_debug_net_ops __net_initdata = {
+	.init = rtnl_net_debug_net_init,
+	.exit = rtnl_net_debug_net_exit,
+	.id = &rtnl_net_debug_net_id,
+	.size = sizeof(struct notifier_block),
+};
+
+static struct notifier_block rtnl_net_debug_block = {
+	.notifier_call = rtnl_net_debug_event,
+};
+
+static int __init rtnl_net_debug_init(void)
+{
+	int ret;
+
+	ret = register_pernet_device(&rtnl_net_debug_net_ops);
+	if (ret)
+		return ret;
+
+	ret = register_netdevice_notifier(&rtnl_net_debug_block);
+	if (ret)
+		unregister_pernet_subsys(&rtnl_net_debug_net_ops);
+
+	return ret;
+}
+
+static void __exit rtnl_net_debug_exit(void)
+{
+	unregister_netdevice_notifier(&rtnl_net_debug_block);
+	unregister_pernet_device(&rtnl_net_debug_net_ops);
+}
+
+subsys_initcall(rtnl_net_debug_init);