Merge branch 'net-hold-netdev-instance-lock-during-ndo-operations'

Stanislav Fomichev says:

====================
net: Hold netdev instance lock during ndo operations

As the gradual purging of rtnl continues, start grabbing netdev
instance lock in more places so we can get to the state where
most paths are working without rtnl. Start with requiring the
drivers that use shaper api (and later queue mgmt api) to work
with both rtnl and netdev instance lock. Eventually we might
attempt to drop rtnl. This mostly affects iavf, gve, bnxt and
netdev sim (as the drivers that implement shaper/queue mgmt)
so those drivers are converted in the process.

call_netdevice_notifiers locking is very inconsistent and might need
a separate follow up. Some notified events are covered by the
instance lock, some are not, which might complicate the driver
expectations.

Reviewed-by: Eric Dumazet <edumazet@google.com>
====================

Link: https://patch.msgid.link/20250305163732.2766420-1-sdf@fomichev.me
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

1 files changed, 81 insertions, 9 deletions

diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index 7ab86ec228b7..d206c9592b60 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -2486,19 +2486,29 @@ struct net_device {
 	bool			up;
 
 	/**
+	 * @request_ops_lock: request the core to run all @netdev_ops and
+	 * @ethtool_ops under the @lock.
+	 */
+	bool			request_ops_lock;
+
+	/**
 	 * @lock: netdev-scope lock, protects a small selection of fields.
 	 * Should always be taken using netdev_lock() / netdev_unlock() helpers.
 	 * Drivers are free to use it for other protection.
 	 *
 	 * Protects:
 	 *	@gro_flush_timeout, @napi_defer_hard_irqs, @napi_list,
-	 *	@net_shaper_hierarchy, @reg_state, @threaded
+	 *	@net_shaper_hierarchy, @reg_state, @threaded, @dev_addr
 	 *
 	 * Partially protects (writers must hold both @lock and rtnl_lock):
 	 *	@up
 	 *
 	 * Also protects some fields in struct napi_struct.
 	 *
+	 * For the drivers that implement shaper or queue API, the scope
+	 * of this lock is expanded to cover most ndo/queue/ethtool/sysfs
+	 * operations.
+	 *
 	 * Ordering: take after rtnl_lock.
 	 */
 	struct mutex		lock;
@@ -2620,16 +2630,35 @@ static inline void netdev_for_each_tx_queue(struct net_device *dev,
 		f(dev, &dev->_tx[i], arg);
 }
 
+static inline int netdev_lock_cmp_fn(const struct lockdep_map *a,
+				     const struct lockdep_map *b)
+{
+	/* Only lower devices currently grab the instance lock, so no
+	 * real ordering issues can occur. In the near future, only
+	 * hardware devices will grab instance lock which also does not
+	 * involve any ordering. Suppress lockdep ordering warnings
+	 * until (if) we start grabbing instance lock on pure SW
+	 * devices (bond/team/veth/etc).
+	 */
+	if (a == b)
+		return 0;
+	return -1;
+}
+
 #define netdev_lockdep_set_classes(dev)				\
 {								\
 	static struct lock_class_key qdisc_tx_busylock_key;	\
 	static struct lock_class_key qdisc_xmit_lock_key;	\
 	static struct lock_class_key dev_addr_list_lock_key;	\
+	static struct lock_class_key dev_instance_lock_key;	\
 	unsigned int i;						\
 								\
 	(dev)->qdisc_tx_busylock = &qdisc_tx_busylock_key;	\
 	lockdep_set_class(&(dev)->addr_list_lock,		\
 			  &dev_addr_list_lock_key);		\
+	lockdep_set_class(&(dev)->lock,				\
+			  &dev_instance_lock_key);		\
+	lock_set_cmp_fn(&dev->lock, netdev_lock_cmp_fn, NULL);	\
 	for (i = 0; i < (dev)->num_tx_queues; i++)		\
 		lockdep_set_class(&(dev)->_tx[i]._xmit_lock,	\
 				  &qdisc_xmit_lock_key);	\
@@ -2736,6 +2765,11 @@ static inline void netdev_lock(struct net_device *dev)
 	mutex_lock(&dev->lock);
 }
 
+static inline bool netdev_trylock(struct net_device *dev)
+{
+	return mutex_trylock(&dev->lock);
+}
+
 static inline void netdev_unlock(struct net_device *dev)
 {
 	mutex_unlock(&dev->lock);
@@ -2753,6 +2787,35 @@ static inline void netdev_assert_locked_or_invisible(struct net_device *dev)
 		netdev_assert_locked(dev);
 }
 
+static inline bool netdev_need_ops_lock(struct net_device *dev)
+{
+	bool ret = dev->request_ops_lock || !!dev->queue_mgmt_ops;
+
+#if IS_ENABLED(CONFIG_NET_SHAPER)
+	ret |= !!dev->netdev_ops->net_shaper_ops;
+#endif
+
+	return ret;
+}
+
+static inline void netdev_lock_ops(struct net_device *dev)
+{
+	if (netdev_need_ops_lock(dev))
+		netdev_lock(dev);
+}
+
+static inline void netdev_unlock_ops(struct net_device *dev)
+{
+	if (netdev_need_ops_lock(dev))
+		netdev_unlock(dev);
+}
+
+static inline void netdev_ops_assert_locked(struct net_device *dev)
+{
+	if (netdev_need_ops_lock(dev))
+		lockdep_assert_held(&dev->lock);
+}
+
 void netif_napi_set_irq_locked(struct napi_struct *napi, int irq);
 
 static inline void netif_napi_set_irq(struct napi_struct *napi, int irq)
@@ -3327,9 +3390,14 @@ struct net_device *dev_get_by_name_rcu(struct net *net, const char *name);
 struct net_device *__dev_get_by_name(struct net *net, const char *name);
 bool netdev_name_in_use(struct net *net, const char *name);
 int dev_alloc_name(struct net_device *dev, const char *name);
+int netif_open(struct net_device *dev, struct netlink_ext_ack *extack);
 int dev_open(struct net_device *dev, struct netlink_ext_ack *extack);
+void netif_close(struct net_device *dev);
 void dev_close(struct net_device *dev);
 void dev_close_many(struct list_head *head, bool unlink);
+int dev_setup_tc(struct net_device *dev, enum tc_setup_type type,
+		 void *type_data);
+void netif_disable_lro(struct net_device *dev);
 void dev_disable_lro(struct net_device *dev);
 int dev_loopback_xmit(struct net *net, struct sock *sk, struct sk_buff *newskb);
 u16 dev_pick_tx_zero(struct net_device *dev, struct sk_buff *skb,
@@ -4177,6 +4245,8 @@ int put_user_ifreq(struct ifreq *ifr, void __user *arg);
 int dev_ioctl(struct net *net, unsigned int cmd, struct ifreq *ifr,
 		void __user *data, bool *need_copyout);
 int dev_ifconf(struct net *net, struct ifconf __user *ifc);
+int dev_eth_ioctl(struct net_device *dev,
+		  struct ifreq *ifr, unsigned int cmd);
 int generic_hwtstamp_get_lower(struct net_device *dev,
 			       struct kernel_hwtstamp_config *kernel_cfg);
 int generic_hwtstamp_set_lower(struct net_device *dev,
@@ -4186,27 +4256,27 @@ int dev_ethtool(struct net *net, struct ifreq *ifr, void __user *userdata);
 unsigned int dev_get_flags(const struct net_device *);
 int __dev_change_flags(struct net_device *dev, unsigned int flags,
 		       struct netlink_ext_ack *extack);
+int netif_change_flags(struct net_device *dev, unsigned int flags,
+		       struct netlink_ext_ack *extack);
 int dev_change_flags(struct net_device *dev, unsigned int flags,
 		     struct netlink_ext_ack *extack);
+int netif_set_alias(struct net_device *dev, const char *alias, size_t len);
 int dev_set_alias(struct net_device *, const char *, size_t);
 int dev_get_alias(const struct net_device *, char *, size_t);
-int __dev_change_net_namespace(struct net_device *dev, struct net *net,
+int netif_change_net_namespace(struct net_device *dev, struct net *net,
 			       const char *pat, int new_ifindex,
 			       struct netlink_ext_ack *extack);
-static inline
 int dev_change_net_namespace(struct net_device *dev, struct net *net,
-			     const char *pat)
-{
-	return __dev_change_net_namespace(dev, net, pat, 0, NULL);
-}
+			     const char *pat);
 int __dev_set_mtu(struct net_device *, int);
+int netif_set_mtu(struct net_device *dev, int new_mtu);
 int dev_set_mtu(struct net_device *, int);
 int dev_pre_changeaddr_notify(struct net_device *dev, const char *addr,
 			      struct netlink_ext_ack *extack);
+int netif_set_mac_address(struct net_device *dev, struct sockaddr *sa,
+			  struct netlink_ext_ack *extack);
 int dev_set_mac_address(struct net_device *dev, struct sockaddr *sa,
 			struct netlink_ext_ack *extack);
-int dev_set_mac_address_user(struct net_device *dev, struct sockaddr *sa,
-			     struct netlink_ext_ack *extack);
 int dev_get_mac_address(struct sockaddr *sa, struct net *net, char *dev_name);
 int dev_get_port_parent_id(struct net_device *dev,
 			   struct netdev_phys_item_id *ppid, bool recurse);
@@ -4218,6 +4288,7 @@ struct sk_buff *dev_hard_start_xmit(struct sk_buff *skb, struct net_device *dev,
 
 int bpf_xdp_link_attach(const union bpf_attr *attr, struct bpf_prog *prog);
 u8 dev_xdp_prog_count(struct net_device *dev);
+int netif_xdp_propagate(struct net_device *dev, struct netdev_bpf *bpf);
 int dev_xdp_propagate(struct net_device *dev, struct netdev_bpf *bpf);
 u8 dev_xdp_sb_prog_count(struct net_device *dev);
 u32 dev_xdp_prog_id(struct net_device *dev, enum bpf_xdp_mode mode);
@@ -4960,6 +5031,7 @@ static inline void __dev_mc_unsync(struct net_device *dev,
 /* Functions used for secondary unicast and multicast support */
 void dev_set_rx_mode(struct net_device *dev);
 int dev_set_promiscuity(struct net_device *dev, int inc);
+int netif_set_allmulti(struct net_device *dev, int inc, bool notify);
 int dev_set_allmulti(struct net_device *dev, int inc);
 void netdev_state_change(struct net_device *dev);
 void __netdev_notify_peers(struct net_device *dev);