[PATCH net-next v4 0/7] net: maintain netif vs dev prefix semantics

[PATCH net-next v4 0/7] net: maintain netif vs dev prefix semantics

[Stanislav Fomichev]

Commit cc34acd577f1 ("docs: net: document new locking reality")
introduced netif_ vs dev_ function semantics: the former expects locked
netdev, the latter takes care of the locking. We don't strictly
follow this semantics on either side, but there are more dev_xxx handlers
now that don't fit. Rename them to netif_xxx where appropriate. We care only
about driver-visible APIs, don't touch stack-internal routines.

The rest seem to be ok:
  * dev_xdp_prog_count - mostly called by sw drivers (bonding), should not matter
  * dev_get_by_xxx - too many to reasonably cleanup, already have different flavors
  * dev_fetch_sw_netstats - don't need instance lock
  * dev_get_tstats64 - never called directly, only as an ndo callback
  * dev_pick_tx_zero - never called directly, only as an ndo callback
  * dev_add_pack / dev_remove_pack - called early enough (in module init) to not matter
  * dev_get_iflink - mostly called by sw drivers, should not matter
  * dev_fill_forward_path - ditto
  * dev_getbyhwaddr_rcu - ditto
  * dev_getbyhwaddr - ditto
  * dev_getfirstbyhwtype - ditto
  * dev_valid_name - ditto
  * __dev_forward_skb dev_forward_skb dev_queue_xmit_nit - established helpers, no netif vs dev distinction

v4:
- drop dev_get_stats patch (Jakub)
- switch to ops-locked assert in netif_set_mtu_ext (Jakub)
- fix more docs (Jakub)

v3:
- move MODULE_IMPORT_NS closer to other module metadata (Willem)
- add MODULE_IMPORT_NS to bridge (build bot)
- add ops lock assert where appropriate (Willem)
- netif_close_many - calls __dev_close_many which has an assert
- netif_set_threaded - already has an assert
- netif_get_flags - READ_ONCE, does not need any locks
- netif_set_mtu_ext - added ops lock assert, updated the doc
- netif_get_mac_address - uses dev_addr_sem
- netif_get_port_parent_id - needs only rtnl?
- netif_get_stats - uses other mechanisms (seqlock)

v2:
- move a bunch of symbols into NETDEV_INTERNAL (Willem)
- netif_close_many
- __netif_set_mtu
- netif_pre_changeaddr_notify
- netif_get_mac_address
- keep more info in the individual patches' commit message (Willem)

Stanislav Fomichev (7):
  net: s/dev_get_port_parent_id/netif_get_port_parent_id/
  net: s/dev_get_mac_address/netif_get_mac_address/
  net: s/dev_pre_changeaddr_notify/netif_pre_changeaddr_notify/
  net: s/__dev_set_mtu/__netif_set_mtu/
  net: s/dev_get_flags/netif_get_flags/
  net: s/dev_set_threaded/netif_set_threaded/
  net: s/dev_close_many/netif_close_many/

 .../networking/net_cachelines/net_device.rst  |  2 +-
 drivers/infiniband/sw/rxe/rxe_verbs.c         |  2 +-
 drivers/net/bonding/bond_main.c               |  5 +-
 .../net/ethernet/atheros/atl1c/atl1c_main.c   |  2 +-
 drivers/net/ethernet/mediatek/mtk_eth_soc.c   |  3 +-
 .../net/ethernet/mellanox/mlx5/core/en_tc.c   |  2 +-
 drivers/net/ethernet/mellanox/mlxsw/pci.c     |  2 +-
 drivers/net/ethernet/renesas/ravb_main.c      |  2 +-
 drivers/net/ipvlan/ipvlan_main.c              |  7 +-
 drivers/net/tap.c                             |  5 +-
 drivers/net/tun.c                             |  3 +-
 drivers/net/wireguard/device.c                |  2 +-
 drivers/net/wireless/ath/ath10k/snoc.c        |  2 +-
 fs/smb/server/smb2pdu.c                       |  2 +-
 include/linux/netdevice.h                     | 17 ++--
 net/8021q/vlan.c                              |  5 +-
 net/bridge/br.c                               |  7 +-
 net/bridge/br_if.c                            |  3 +-
 net/bridge/br_netlink.c                       |  2 +-
 net/bridge/br_switchdev.c                     |  2 +-
 net/core/dev.c                                | 95 ++++++++++---------
 net/core/dev_addr_lists.c                     |  2 +-
 net/core/dev_api.c                            | 12 +++
 net/core/dev_ioctl.c                          |  5 +-
 net/core/net-sysfs.c                          |  6 +-
 net/core/rtnetlink.c                          |  6 +-
 net/dsa/dsa.c                                 |  3 +-
 net/dsa/user.c                                |  2 +-
 net/ipv4/fib_frontend.c                       |  2 +-
 net/ipv4/fib_semantics.c                      |  2 +-
 net/ipv4/ipmr.c                               |  2 +-
 net/ipv4/nexthop.c                            |  2 +-
 net/ipv6/addrconf.c                           |  2 +-
 net/mpls/af_mpls.c                            |  6 +-
 net/wireless/wext-core.c                      |  2 +-
 35 files changed, 128 insertions(+), 98 deletions(-)

-- 
2.50.0

[PATCH net-next v4 1/7] net: s/dev_get_port_parent_id/netif_get_port_parent_id/

[Stanislav Fomichev]

Commit cc34acd577f1 ("docs: net: document new locking reality")
introduced netif_ vs dev_ function semantics: the former expects locked
netdev, the latter takes care of the locking. We don't strictly
follow this semantics on either side, but there are more dev_xxx handlers
now that don't fit. Rename them to netif_xxx where appropriate.

Signed-off-by: Stanislav Fomichev <sdf@fomichev.me>
---
 .../net/ethernet/mellanox/mlx5/core/en_tc.c   |  2 +-
 include/linux/netdevice.h                     |  4 +--
 net/bridge/br_switchdev.c                     |  2 +-
 net/core/dev.c                                | 25 ++++++++++---------
 net/core/net-sysfs.c                          |  2 +-
 net/core/rtnetlink.c                          |  2 +-
 net/ipv4/ipmr.c                               |  2 +-
 7 files changed, 20 insertions(+), 19 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index fef418e1ed1a..32c07a8b03d1 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -5446,7 +5446,7 @@ int mlx5e_tc_esw_init(struct mlx5_rep_uplink_priv *uplink_priv)
 		goto err_action_counter;
 	}
 
-	err = dev_get_port_parent_id(priv->netdev, &ppid, false);
+	err = netif_get_port_parent_id(priv->netdev, &ppid, false);
 	if (!err) {
 		memcpy(&key, &ppid.id, sizeof(key));
 		mlx5_esw_offloads_devcom_init(esw, key);
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index a80d21a14612..a8a275fb05c3 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -4222,8 +4222,8 @@ int dev_set_mac_address(struct net_device *dev, struct sockaddr_storage *ss,
 int dev_set_mac_address_user(struct net_device *dev, struct sockaddr_storage *ss,
 			     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);
+int netif_get_port_parent_id(struct net_device *dev,
+			     struct netdev_phys_item_id *ppid, bool recurse);
 bool netdev_port_same_parent_id(struct net_device *a, struct net_device *b);
 
 struct sk_buff *validate_xmit_skb_list(struct sk_buff *skb, struct net_device *dev, bool *again);
diff --git a/net/bridge/br_switchdev.c b/net/bridge/br_switchdev.c
index 95d7355a0407..1385ff604bbd 100644
--- a/net/bridge/br_switchdev.c
+++ b/net/bridge/br_switchdev.c
@@ -834,7 +834,7 @@ int br_switchdev_port_offload(struct net_bridge_port *p,
 	struct netdev_phys_item_id ppid;
 	int err;
 
-	err = dev_get_port_parent_id(dev, &ppid, false);
+	err = netif_get_port_parent_id(dev, &ppid, false);
 	if (err)
 		return err;
 
diff --git a/net/core/dev.c b/net/core/dev.c
index e365b099484e..3567cd09920e 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -9815,16 +9815,17 @@ int dev_get_phys_port_name(struct net_device *dev,
 }
 
 /**
- *	dev_get_port_parent_id - Get the device's port parent identifier
- *	@dev: network device
- *	@ppid: pointer to a storage for the port's parent identifier
- *	@recurse: allow/disallow recursion to lower devices
+ * netif_get_port_parent_id() - Get the device's port parent identifier
+ * @dev: network device
+ * @ppid: pointer to a storage for the port's parent identifier
+ * @recurse: allow/disallow recursion to lower devices
+ *
+ * Get the devices's port parent identifier.
  *
- *	Get the devices's port parent identifier
+ * Return: 0 on success, -errno on failure.
  */
-int dev_get_port_parent_id(struct net_device *dev,
-			   struct netdev_phys_item_id *ppid,
-			   bool recurse)
+int netif_get_port_parent_id(struct net_device *dev,
+			     struct netdev_phys_item_id *ppid, bool recurse)
 {
 	const struct net_device_ops *ops = dev->netdev_ops;
 	struct netdev_phys_item_id first = { };
@@ -9843,7 +9844,7 @@ int dev_get_port_parent_id(struct net_device *dev,
 		return err;
 
 	netdev_for_each_lower_dev(dev, lower_dev, iter) {
-		err = dev_get_port_parent_id(lower_dev, ppid, true);
+		err = netif_get_port_parent_id(lower_dev, ppid, true);
 		if (err)
 			break;
 		if (!first.id_len)
@@ -9854,7 +9855,7 @@ int dev_get_port_parent_id(struct net_device *dev,
 
 	return err;
 }
-EXPORT_SYMBOL(dev_get_port_parent_id);
+EXPORT_SYMBOL(netif_get_port_parent_id);
 
 /**
  *	netdev_port_same_parent_id - Indicate if two network devices have
@@ -9867,8 +9868,8 @@ bool netdev_port_same_parent_id(struct net_device *a, struct net_device *b)
 	struct netdev_phys_item_id a_id = { };
 	struct netdev_phys_item_id b_id = { };
 
-	if (dev_get_port_parent_id(a, &a_id, true) ||
-	    dev_get_port_parent_id(b, &b_id, true))
+	if (netif_get_port_parent_id(a, &a_id, true) ||
+	    netif_get_port_parent_id(b, &b_id, true))
 		return false;
 
 	return netdev_phys_item_id_same(&a_id, &b_id);
diff --git a/net/core/net-sysfs.c b/net/core/net-sysfs.c
index 8f897e2c8b4f..f7a6cc7aea79 100644
--- a/net/core/net-sysfs.c
+++ b/net/core/net-sysfs.c
@@ -687,7 +687,7 @@ static ssize_t phys_switch_id_show(struct device *dev,
 	if (ret)
 		return ret;
 
-	ret = dev_get_port_parent_id(netdev, &ppid, false);
+	ret = netif_get_port_parent_id(netdev, &ppid, false);
 	if (!ret)
 		ret = sysfs_emit(buf, "%*phN\n", ppid.id_len, ppid.id);
 
diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c
index a9555bfc372f..108995b6eced 100644
--- a/net/core/rtnetlink.c
+++ b/net/core/rtnetlink.c
@@ -1448,7 +1448,7 @@ static int rtnl_phys_switch_id_fill(struct sk_buff *skb, struct net_device *dev)
 	struct netdev_phys_item_id ppid = { };
 	int err;
 
-	err = dev_get_port_parent_id(dev, &ppid, false);
+	err = netif_get_port_parent_id(dev, &ppid, false);
 	if (err) {
 		if (err == -EOPNOTSUPP)
 			return 0;
diff --git a/net/ipv4/ipmr.c b/net/ipv4/ipmr.c
index 3a2044e6033d..e86a8a862c41 100644
--- a/net/ipv4/ipmr.c
+++ b/net/ipv4/ipmr.c
@@ -901,7 +901,7 @@ static int vif_add(struct net *net, struct mr_table *mrt,
 			vifc->vifc_flags | (!mrtsock ? VIFF_STATIC : 0),
 			(VIFF_TUNNEL | VIFF_REGISTER));
 
-	err = dev_get_port_parent_id(dev, &ppid, true);
+	err = netif_get_port_parent_id(dev, &ppid, true);
 	if (err == 0) {
 		memcpy(v->dev_parent_id.id, ppid.id, ppid.id_len);
 		v->dev_parent_id.id_len = ppid.id_len;
-- 
2.50.0

[PATCH net-next v4 2/7] net: s/dev_get_mac_address/netif_get_mac_address/

[Stanislav Fomichev]

Commit cc34acd577f1 ("docs: net: document new locking reality")
introduced netif_ vs dev_ function semantics: the former expects locked
netdev, the latter takes care of the locking. We don't strictly
follow this semantics on either side, but there are more dev_xxx handlers
now that don't fit. Rename them to netif_xxx where appropriate.

netif_get_mac_address is used only by tun/tap, so move it into
NETDEV_INTERNAL namespace.

Signed-off-by: Stanislav Fomichev <sdf@fomichev.me>
---
 drivers/net/tap.c         | 5 +++--
 drivers/net/tun.c         | 3 ++-
 include/linux/netdevice.h | 2 +-
 net/core/dev.c            | 4 ++--
 net/core/dev_ioctl.c      | 3 ++-
 net/core/net-sysfs.c      | 2 +-
 6 files changed, 11 insertions(+), 8 deletions(-)

diff --git a/drivers/net/tap.c b/drivers/net/tap.c
index d82eb7276a8b..1197f245e873 100644
--- a/drivers/net/tap.c
+++ b/drivers/net/tap.c
@@ -1000,8 +1000,8 @@ static long tap_ioctl(struct file *file, unsigned int cmd,
 			return -ENOLINK;
 		}
 		ret = 0;
-		dev_get_mac_address((struct sockaddr *)&ss, dev_net(tap->dev),
-				    tap->dev->name);
+		netif_get_mac_address((struct sockaddr *)&ss, dev_net(tap->dev),
+				      tap->dev->name);
 		if (copy_to_user(&ifr->ifr_name, tap->dev->name, IFNAMSIZ) ||
 		    copy_to_user(&ifr->ifr_hwaddr, &ss, sizeof(ifr->ifr_hwaddr)))
 			ret = -EFAULT;
@@ -1282,3 +1282,4 @@ MODULE_DESCRIPTION("Common library for drivers implementing the TAP interface");
 MODULE_AUTHOR("Arnd Bergmann <arnd@arndb.de>");
 MODULE_AUTHOR("Sainath Grandhi <sainath.grandhi@intel.com>");
 MODULE_LICENSE("GPL");
+MODULE_IMPORT_NS("NETDEV_INTERNAL");
diff --git a/drivers/net/tun.c b/drivers/net/tun.c
index 49bcd12a4ac8..4568fe4c7e58 100644
--- a/drivers/net/tun.c
+++ b/drivers/net/tun.c
@@ -3225,7 +3225,7 @@ static long __tun_chr_ioctl(struct file *file, unsigned int cmd,
 
 	case SIOCGIFHWADDR:
 		/* Get hw address */
-		dev_get_mac_address(&ifr.ifr_hwaddr, net, tun->dev->name);
+		netif_get_mac_address(&ifr.ifr_hwaddr, net, tun->dev->name);
 		if (copy_to_user(argp, &ifr, ifreq_len))
 			ret = -EFAULT;
 		break;
@@ -3734,3 +3734,4 @@ MODULE_AUTHOR(DRV_COPYRIGHT);
 MODULE_LICENSE("GPL");
 MODULE_ALIAS_MISCDEV(TUN_MINOR);
 MODULE_ALIAS("devname:net/tun");
+MODULE_IMPORT_NS("NETDEV_INTERNAL");
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index a8a275fb05c3..b4fcf72e4b16 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -4221,7 +4221,7 @@ int dev_set_mac_address(struct net_device *dev, struct sockaddr_storage *ss,
 			struct netlink_ext_ack *extack);
 int dev_set_mac_address_user(struct net_device *dev, struct sockaddr_storage *ss,
 			     struct netlink_ext_ack *extack);
-int dev_get_mac_address(struct sockaddr *sa, struct net *net, char *dev_name);
+int netif_get_mac_address(struct sockaddr *sa, struct net *net, char *dev_name);
 int netif_get_port_parent_id(struct net_device *dev,
 			     struct netdev_phys_item_id *ppid, bool recurse);
 bool netdev_port_same_parent_id(struct net_device *a, struct net_device *b);
diff --git a/net/core/dev.c b/net/core/dev.c
index 3567cd09920e..a0eb0ed3a1bd 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -9736,7 +9736,7 @@ int netif_set_mac_address(struct net_device *dev, struct sockaddr_storage *ss,
 DECLARE_RWSEM(dev_addr_sem);
 
 /* "sa" is a true struct sockaddr with limited "sa_data" member. */
-int dev_get_mac_address(struct sockaddr *sa, struct net *net, char *dev_name)
+int netif_get_mac_address(struct sockaddr *sa, struct net *net, char *dev_name)
 {
 	size_t size = sizeof(sa->sa_data_min);
 	struct net_device *dev;
@@ -9762,7 +9762,7 @@ int dev_get_mac_address(struct sockaddr *sa, struct net *net, char *dev_name)
 	up_read(&dev_addr_sem);
 	return ret;
 }
-EXPORT_SYMBOL(dev_get_mac_address);
+EXPORT_SYMBOL_NS_GPL(netif_get_mac_address, "NETDEV_INTERNAL");
 
 int netif_change_carrier(struct net_device *dev, bool new_carrier)
 {
diff --git a/net/core/dev_ioctl.c b/net/core/dev_ioctl.c
index 616479e71466..ceb2d63a818a 100644
--- a/net/core/dev_ioctl.c
+++ b/net/core/dev_ioctl.c
@@ -728,7 +728,8 @@ int dev_ioctl(struct net *net, unsigned int cmd, struct ifreq *ifr,
 	switch (cmd) {
 	case SIOCGIFHWADDR:
 		dev_load(net, ifr->ifr_name);
-		ret = dev_get_mac_address(&ifr->ifr_hwaddr, net, ifr->ifr_name);
+		ret = netif_get_mac_address(&ifr->ifr_hwaddr, net,
+					    ifr->ifr_name);
 		if (colon)
 			*colon = ':';
 		return ret;
diff --git a/net/core/net-sysfs.c b/net/core/net-sysfs.c
index f7a6cc7aea79..e41ad1890e49 100644
--- a/net/core/net-sysfs.c
+++ b/net/core/net-sysfs.c
@@ -256,7 +256,7 @@ static ssize_t name_assign_type_show(struct device *dev,
 }
 static DEVICE_ATTR_RO(name_assign_type);
 
-/* use same locking rules as GIFHWADDR ioctl's (dev_get_mac_address()) */
+/* use same locking rules as GIFHWADDR ioctl's (netif_get_mac_address()) */
 static ssize_t address_show(struct device *dev, struct device_attribute *attr,
 			    char *buf)
 {
-- 
2.50.0

[PATCH net-next v4 3/7] net: s/dev_pre_changeaddr_notify/netif_pre_changeaddr_notify/

[Stanislav Fomichev]

Commit cc34acd577f1 ("docs: net: document new locking reality")
introduced netif_ vs dev_ function semantics: the former expects locked
netdev, the latter takes care of the locking. We don't strictly
follow this semantics on either side, but there are more dev_xxx handlers
now that don't fit. Rename them to netif_xxx where appropriate.

netif_pre_changeaddr_notify is used only by ipvlan/bond, so move it into
NETDEV_INTERNAL namespace.

Signed-off-by: Stanislav Fomichev <sdf@fomichev.me>
---
 drivers/net/bonding/bond_main.c  |  3 ++-
 drivers/net/ipvlan/ipvlan_main.c |  7 ++++---
 include/linux/netdevice.h        |  4 ++--
 net/bridge/br.c                  |  7 ++++---
 net/bridge/br_if.c               |  3 ++-
 net/core/dev.c                   | 18 ++++++++++--------
 net/core/dev_addr_lists.c        |  2 +-
 7 files changed, 25 insertions(+), 19 deletions(-)

diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c
index 17c7542be6a5..d8281c486a44 100644
--- a/drivers/net/bonding/bond_main.c
+++ b/drivers/net/bonding/bond_main.c
@@ -1040,7 +1040,7 @@ static int bond_set_dev_addr(struct net_device *bond_dev,
 
 	slave_dbg(bond_dev, slave_dev, "bond_dev=%p slave_dev=%p slave_dev->addr_len=%d\n",
 		  bond_dev, slave_dev, slave_dev->addr_len);
-	err = dev_pre_changeaddr_notify(bond_dev, slave_dev->dev_addr, NULL);
+	err = netif_pre_changeaddr_notify(bond_dev, slave_dev->dev_addr, NULL);
 	if (err)
 		return err;
 
@@ -6743,3 +6743,4 @@ module_exit(bonding_exit);
 MODULE_LICENSE("GPL");
 MODULE_DESCRIPTION(DRV_DESCRIPTION);
 MODULE_AUTHOR("Thomas Davis, tadavis@lbl.gov and many others");
+MODULE_IMPORT_NS("NETDEV_INTERNAL");
diff --git a/drivers/net/ipvlan/ipvlan_main.c b/drivers/net/ipvlan/ipvlan_main.c
index 0ed2fd833a5d..660f3db11766 100644
--- a/drivers/net/ipvlan/ipvlan_main.c
+++ b/drivers/net/ipvlan/ipvlan_main.c
@@ -784,9 +784,9 @@ static int ipvlan_device_event(struct notifier_block *unused,
 	case NETDEV_PRE_CHANGEADDR:
 		prechaddr_info = ptr;
 		list_for_each_entry(ipvlan, &port->ipvlans, pnode) {
-			err = dev_pre_changeaddr_notify(ipvlan->dev,
-						    prechaddr_info->dev_addr,
-						    extack);
+			err = netif_pre_changeaddr_notify(ipvlan->dev,
+							  prechaddr_info->dev_addr,
+							  extack);
 			if (err)
 				return notifier_from_errno(err);
 		}
@@ -1094,3 +1094,4 @@ MODULE_LICENSE("GPL");
 MODULE_AUTHOR("Mahesh Bandewar <maheshb@google.com>");
 MODULE_DESCRIPTION("Driver for L3 (IPv6/IPv4) based VLANs");
 MODULE_ALIAS_RTNL_LINK("ipvlan");
+MODULE_IMPORT_NS("NETDEV_INTERNAL");
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index b4fcf72e4b16..bae5d80ef6d0 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -4213,8 +4213,8 @@ int dev_change_net_namespace(struct net_device *dev, struct net *net,
 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_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_storage *ss,
 			  struct netlink_ext_ack *extack);
 int dev_set_mac_address(struct net_device *dev, struct sockaddr_storage *ss,
diff --git a/net/bridge/br.c b/net/bridge/br.c
index 0adeafe11a36..1885d0c315f0 100644
--- a/net/bridge/br.c
+++ b/net/bridge/br.c
@@ -74,9 +74,9 @@ static int br_device_event(struct notifier_block *unused, unsigned long event, v
 		if (br->dev->addr_assign_type == NET_ADDR_SET)
 			break;
 		prechaddr_info = ptr;
-		err = dev_pre_changeaddr_notify(br->dev,
-						prechaddr_info->dev_addr,
-						extack);
+		err = netif_pre_changeaddr_notify(br->dev,
+						  prechaddr_info->dev_addr,
+						  extack);
 		if (err)
 			return notifier_from_errno(err);
 		break;
@@ -484,3 +484,4 @@ MODULE_LICENSE("GPL");
 MODULE_VERSION(BR_VERSION);
 MODULE_ALIAS_RTNL_LINK("bridge");
 MODULE_DESCRIPTION("Ethernet bridge driver");
+MODULE_IMPORT_NS("NETDEV_INTERNAL");
diff --git a/net/bridge/br_if.c b/net/bridge/br_if.c
index 2450690f98cf..98c5b9c3145f 100644
--- a/net/bridge/br_if.c
+++ b/net/bridge/br_if.c
@@ -668,7 +668,8 @@ int br_add_if(struct net_bridge *br, struct net_device *dev,
 		/* Ask for permission to use this MAC address now, even if we
 		 * don't end up choosing it below.
 		 */
-		err = dev_pre_changeaddr_notify(br->dev, dev->dev_addr, extack);
+		err = netif_pre_changeaddr_notify(br->dev, dev->dev_addr,
+						  extack);
 		if (err)
 			goto err6;
 	}
diff --git a/net/core/dev.c b/net/core/dev.c
index a0eb0ed3a1bd..95a29ea255e8 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -9687,13 +9687,15 @@ void netif_set_group(struct net_device *dev, int new_group)
 }
 
 /**
- *	dev_pre_changeaddr_notify - Call NETDEV_PRE_CHANGEADDR.
- *	@dev: device
- *	@addr: new address
- *	@extack: netlink extended ack
+ * netif_pre_changeaddr_notify() - Call NETDEV_PRE_CHANGEADDR.
+ * @dev: device
+ * @addr: new address
+ * @extack: netlink extended ack
+ *
+ * Return: 0 on success, -errno on failure.
  */
-int dev_pre_changeaddr_notify(struct net_device *dev, const char *addr,
-			      struct netlink_ext_ack *extack)
+int netif_pre_changeaddr_notify(struct net_device *dev, const char *addr,
+				struct netlink_ext_ack *extack)
 {
 	struct netdev_notifier_pre_changeaddr_info info = {
 		.info.dev = dev,
@@ -9705,7 +9707,7 @@ int dev_pre_changeaddr_notify(struct net_device *dev, const char *addr,
 	rc = call_netdevice_notifiers_info(NETDEV_PRE_CHANGEADDR, &info.info);
 	return notifier_to_errno(rc);
 }
-EXPORT_SYMBOL(dev_pre_changeaddr_notify);
+EXPORT_SYMBOL_NS_GPL(netif_pre_changeaddr_notify, "NETDEV_INTERNAL");
 
 int netif_set_mac_address(struct net_device *dev, struct sockaddr_storage *ss,
 			  struct netlink_ext_ack *extack)
@@ -9719,7 +9721,7 @@ int netif_set_mac_address(struct net_device *dev, struct sockaddr_storage *ss,
 		return -EINVAL;
 	if (!netif_device_present(dev))
 		return -ENODEV;
-	err = dev_pre_changeaddr_notify(dev, ss->__data, extack);
+	err = netif_pre_changeaddr_notify(dev, ss->__data, extack);
 	if (err)
 		return err;
 	if (memcmp(dev->dev_addr, ss->__data, dev->addr_len)) {
diff --git a/net/core/dev_addr_lists.c b/net/core/dev_addr_lists.c
index 90716bd736f3..76c91f224886 100644
--- a/net/core/dev_addr_lists.c
+++ b/net/core/dev_addr_lists.c
@@ -603,7 +603,7 @@ int dev_addr_add(struct net_device *dev, const unsigned char *addr,
 
 	ASSERT_RTNL();
 
-	err = dev_pre_changeaddr_notify(dev, addr, NULL);
+	err = netif_pre_changeaddr_notify(dev, addr, NULL);
 	if (err)
 		return err;
 	err = __hw_addr_add(&dev->dev_addrs, addr, dev->addr_len, addr_type);
-- 
2.50.0

[PATCH net-next v4 4/7] net: s/__dev_set_mtu/__netif_set_mtu/

[Stanislav Fomichev]

Commit cc34acd577f1 ("docs: net: document new locking reality")
introduced netif_ vs dev_ function semantics: the former expects locked
netdev, the latter takes care of the locking. We don't strictly
follow this semantics on either side, but there are more dev_xxx handlers
now that don't fit. Rename them to netif_xxx where appropriate.

__netif_set_mtu is used only by bond, so move it into
NETDEV_INTERNAL namespace.

Signed-off-by: Stanislav Fomichev <sdf@fomichev.me>
---
 drivers/net/bonding/bond_main.c |  2 +-
 include/linux/netdevice.h       |  2 +-
 net/core/dev.c                  | 22 +++++++++++++---------
 3 files changed, 15 insertions(+), 11 deletions(-)

diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c
index d8281c486a44..257333c88710 100644
--- a/drivers/net/bonding/bond_main.c
+++ b/drivers/net/bonding/bond_main.c
@@ -2669,7 +2669,7 @@ static int __bond_release_one(struct net_device *bond_dev,
 
 	if (unregister) {
 		netdev_lock_ops(slave_dev);
-		__dev_set_mtu(slave_dev, slave->original_mtu);
+		__netif_set_mtu(slave_dev, slave->original_mtu);
 		netdev_unlock_ops(slave_dev);
 	} else {
 		dev_set_mtu(slave_dev, slave->original_mtu);
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index bae5d80ef6d0..2115050e1d25 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -4210,7 +4210,7 @@ int __dev_change_net_namespace(struct net_device *dev, struct net *net,
 			       struct netlink_ext_ack *extack);
 int dev_change_net_namespace(struct net_device *dev, struct net *net,
 			     const char *pat);
-int __dev_set_mtu(struct net_device *, int);
+int __netif_set_mtu(struct net_device *dev, int new_mtu);
 int netif_set_mtu(struct net_device *dev, int new_mtu);
 int dev_set_mtu(struct net_device *, int);
 int netif_pre_changeaddr_notify(struct net_device *dev, const char *addr,
diff --git a/net/core/dev.c b/net/core/dev.c
index 95a29ea255e8..750cf055c090 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -9565,7 +9565,7 @@ int netif_change_flags(struct net_device *dev, unsigned int flags,
 	return ret;
 }
 
-int __dev_set_mtu(struct net_device *dev, int new_mtu)
+int __netif_set_mtu(struct net_device *dev, int new_mtu)
 {
 	const struct net_device_ops *ops = dev->netdev_ops;
 
@@ -9576,7 +9576,7 @@ int __dev_set_mtu(struct net_device *dev, int new_mtu)
 	WRITE_ONCE(dev->mtu, new_mtu);
 	return 0;
 }
-EXPORT_SYMBOL(__dev_set_mtu);
+EXPORT_SYMBOL_NS_GPL(__netif_set_mtu, "NETDEV_INTERNAL");
 
 int dev_validate_mtu(struct net_device *dev, int new_mtu,
 		     struct netlink_ext_ack *extack)
@@ -9595,18 +9595,22 @@ int dev_validate_mtu(struct net_device *dev, int new_mtu,
 }
 
 /**
- *	netif_set_mtu_ext - Change maximum transfer unit
- *	@dev: device
- *	@new_mtu: new transfer unit
- *	@extack: netlink extended ack
+ * netif_set_mtu_ext() - Change maximum transfer unit
+ * @dev: device
+ * @new_mtu: new transfer unit
+ * @extack: netlink extended ack
  *
- *	Change the maximum transfer size of the network device.
+ * Change the maximum transfer size of the network device.
+ *
+ * Return: 0 on success, -errno on failure.
  */
 int netif_set_mtu_ext(struct net_device *dev, int new_mtu,
 		      struct netlink_ext_ack *extack)
 {
 	int err, orig_mtu;
 
+	netdev_ops_assert_locked(dev);
+
 	if (new_mtu == dev->mtu)
 		return 0;
 
@@ -9623,7 +9627,7 @@ int netif_set_mtu_ext(struct net_device *dev, int new_mtu,
 		return err;
 
 	orig_mtu = dev->mtu;
-	err = __dev_set_mtu(dev, new_mtu);
+	err = __netif_set_mtu(dev, new_mtu);
 
 	if (!err) {
 		err = call_netdevice_notifiers_mtu(NETDEV_CHANGEMTU, dev,
@@ -9633,7 +9637,7 @@ int netif_set_mtu_ext(struct net_device *dev, int new_mtu,
 			/* setting mtu back and notifying everyone again,
 			 * so that they have a chance to revert changes.
 			 */
-			__dev_set_mtu(dev, orig_mtu);
+			__netif_set_mtu(dev, orig_mtu);
 			call_netdevice_notifiers_mtu(NETDEV_CHANGEMTU, dev,
 						     new_mtu);
 		}
-- 
2.50.0

[PATCH net-next v4 5/7] net: s/dev_get_flags/netif_get_flags/

[Stanislav Fomichev]

Commit cc34acd577f1 ("docs: net: document new locking reality")
introduced netif_ vs dev_ function semantics: the former expects locked
netdev, the latter takes care of the locking. We don't strictly
follow this semantics on either side, but there are more dev_xxx handlers
now that don't fit. Rename them to netif_xxx where appropriate.

Signed-off-by: Stanislav Fomichev <sdf@fomichev.me>
---
 drivers/infiniband/sw/rxe/rxe_verbs.c |  2 +-
 fs/smb/server/smb2pdu.c               |  2 +-
 include/linux/netdevice.h             |  2 +-
 net/8021q/vlan.c                      |  2 +-
 net/bridge/br_netlink.c               |  2 +-
 net/core/dev.c                        | 10 +++++-----
 net/core/dev_ioctl.c                  |  2 +-
 net/core/rtnetlink.c                  |  4 ++--
 net/ipv4/fib_frontend.c               |  2 +-
 net/ipv4/fib_semantics.c              |  2 +-
 net/ipv4/nexthop.c                    |  2 +-
 net/ipv6/addrconf.c                   |  2 +-
 net/mpls/af_mpls.c                    |  6 +++---
 net/wireless/wext-core.c              |  2 +-
 14 files changed, 21 insertions(+), 21 deletions(-)

diff --git a/drivers/infiniband/sw/rxe/rxe_verbs.c b/drivers/infiniband/sw/rxe/rxe_verbs.c
index 2331e698a65b..4f86b56fee26 100644
--- a/drivers/infiniband/sw/rxe/rxe_verbs.c
+++ b/drivers/infiniband/sw/rxe/rxe_verbs.c
@@ -65,7 +65,7 @@ static int rxe_query_port(struct ib_device *ibdev,
 	attr->state = ib_get_curr_port_state(ndev);
 	if (attr->state == IB_PORT_ACTIVE)
 		attr->phys_state = IB_PORT_PHYS_STATE_LINK_UP;
-	else if (dev_get_flags(ndev) & IFF_UP)
+	else if (netif_get_flags(ndev) & IFF_UP)
 		attr->phys_state = IB_PORT_PHYS_STATE_POLLING;
 	else
 		attr->phys_state = IB_PORT_PHYS_STATE_DISABLED;
diff --git a/fs/smb/server/smb2pdu.c b/fs/smb/server/smb2pdu.c
index fafa86273f12..b837535d2bfc 100644
--- a/fs/smb/server/smb2pdu.c
+++ b/fs/smb/server/smb2pdu.c
@@ -7847,7 +7847,7 @@ static int fsctl_query_iface_info_ioctl(struct ksmbd_conn *conn,
 		if (!ksmbd_find_netdev_name_iface_list(netdev->name))
 			continue;
 
-		flags = dev_get_flags(netdev);
+		flags = netif_get_flags(netdev);
 		if (!(flags & IFF_RUNNING))
 			continue;
 ipv6_retry:
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index 2115050e1d25..d11804be460c 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -4195,7 +4195,7 @@ int generic_hwtstamp_set_lower(struct net_device *dev,
 			       struct kernel_hwtstamp_config *kernel_cfg,
 			       struct netlink_ext_ack *extack);
 int dev_ethtool(struct net *net, struct ifreq *ifr, void __user *userdata);
-unsigned int dev_get_flags(const struct net_device *);
+unsigned int netif_get_flags(const struct net_device *dev);
 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,
diff --git a/net/8021q/vlan.c b/net/8021q/vlan.c
index 06908e37c3d9..df19a9c0c9d9 100644
--- a/net/8021q/vlan.c
+++ b/net/8021q/vlan.c
@@ -459,7 +459,7 @@ static int vlan_device_event(struct notifier_block *unused, unsigned long event,
 	case NETDEV_UP:
 		/* Put all VLANs for this dev in the up state too.  */
 		vlan_group_for_each_dev(grp, i, vlandev) {
-			flgs = dev_get_flags(vlandev);
+			flgs = netif_get_flags(vlandev);
 			if (flgs & IFF_UP)
 				continue;
 
diff --git a/net/bridge/br_netlink.c b/net/bridge/br_netlink.c
index 6e337937d0d7..4e2d53b27221 100644
--- a/net/bridge/br_netlink.c
+++ b/net/bridge/br_netlink.c
@@ -479,7 +479,7 @@ static int br_fill_ifinfo(struct sk_buff *skb,
 	hdr->__ifi_pad = 0;
 	hdr->ifi_type = dev->type;
 	hdr->ifi_index = dev->ifindex;
-	hdr->ifi_flags = dev_get_flags(dev);
+	hdr->ifi_flags = netif_get_flags(dev);
 	hdr->ifi_change = 0;
 
 	if (nla_put_string(skb, IFLA_IFNAME, dev->name) ||
diff --git a/net/core/dev.c b/net/core/dev.c
index 750cf055c090..2d3d9166e7af 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -9425,12 +9425,12 @@ void dev_set_rx_mode(struct net_device *dev)
 }
 
 /**
- *	dev_get_flags - get flags reported to userspace
- *	@dev: device
+ * netif_get_flags() - get flags reported to userspace
+ * @dev: device
  *
- *	Get the combination of flag bits exported through APIs to userspace.
+ * Get the combination of flag bits exported through APIs to userspace.
  */
-unsigned int dev_get_flags(const struct net_device *dev)
+unsigned int netif_get_flags(const struct net_device *dev)
 {
 	unsigned int flags;
 
@@ -9453,7 +9453,7 @@ unsigned int dev_get_flags(const struct net_device *dev)
 
 	return flags;
 }
-EXPORT_SYMBOL(dev_get_flags);
+EXPORT_SYMBOL(netif_get_flags);
 
 int __dev_change_flags(struct net_device *dev, unsigned int flags,
 		       struct netlink_ext_ack *extack)
diff --git a/net/core/dev_ioctl.c b/net/core/dev_ioctl.c
index ceb2d63a818a..9c0ad7f4b5d8 100644
--- a/net/core/dev_ioctl.c
+++ b/net/core/dev_ioctl.c
@@ -147,7 +147,7 @@ static int dev_ifsioc_locked(struct net *net, struct ifreq *ifr, unsigned int cm
 
 	switch (cmd) {
 	case SIOCGIFFLAGS:	/* Get interface flags */
-		ifr->ifr_flags = (short) dev_get_flags(dev);
+		ifr->ifr_flags = (short)netif_get_flags(dev);
 		return 0;
 
 	case SIOCGIFMETRIC:	/* Get the metric on the interface
diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c
index 108995b6eced..094b085cff20 100644
--- a/net/core/rtnetlink.c
+++ b/net/core/rtnetlink.c
@@ -2038,7 +2038,7 @@ static int rtnl_fill_ifinfo(struct sk_buff *skb,
 	ifm->__ifi_pad = 0;
 	ifm->ifi_type = READ_ONCE(dev->type);
 	ifm->ifi_index = READ_ONCE(dev->ifindex);
-	ifm->ifi_flags = dev_get_flags(dev);
+	ifm->ifi_flags = netif_get_flags(dev);
 	ifm->ifi_change = change;
 
 	if (tgt_netnsid >= 0 && nla_put_s32(skb, IFLA_TARGET_NETNSID, tgt_netnsid))
@@ -5227,7 +5227,7 @@ int ndo_dflt_bridge_getlink(struct sk_buff *skb, u32 pid, u32 seq,
 	ifm->__ifi_pad = 0;
 	ifm->ifi_type = dev->type;
 	ifm->ifi_index = dev->ifindex;
-	ifm->ifi_flags = dev_get_flags(dev);
+	ifm->ifi_flags = netif_get_flags(dev);
 	ifm->ifi_change = 0;
 
 
diff --git a/net/ipv4/fib_frontend.c b/net/ipv4/fib_frontend.c
index fd1e1507a224..6e1b94796f67 100644
--- a/net/ipv4/fib_frontend.c
+++ b/net/ipv4/fib_frontend.c
@@ -1524,7 +1524,7 @@ static int fib_netdev_event(struct notifier_block *this, unsigned long event, vo
 		fib_disable_ip(dev, event, false);
 		break;
 	case NETDEV_CHANGE:
-		flags = dev_get_flags(dev);
+		flags = netif_get_flags(dev);
 		if (flags & (IFF_RUNNING | IFF_LOWER_UP))
 			fib_sync_up(dev, RTNH_F_LINKDOWN);
 		else
diff --git a/net/ipv4/fib_semantics.c b/net/ipv4/fib_semantics.c
index a2f04992f579..a5f3c8459758 100644
--- a/net/ipv4/fib_semantics.c
+++ b/net/ipv4/fib_semantics.c
@@ -2087,7 +2087,7 @@ int fib_sync_up(struct net_device *dev, unsigned char nh_flags)
 		return 0;
 
 	if (nh_flags & RTNH_F_DEAD) {
-		unsigned int flags = dev_get_flags(dev);
+		unsigned int flags = netif_get_flags(dev);
 
 		if (flags & (IFF_RUNNING | IFF_LOWER_UP))
 			nh_flags |= RTNH_F_LINKDOWN;
diff --git a/net/ipv4/nexthop.c b/net/ipv4/nexthop.c
index e808801ab9b8..29118c43ebf5 100644
--- a/net/ipv4/nexthop.c
+++ b/net/ipv4/nexthop.c
@@ -3884,7 +3884,7 @@ static int nh_netdev_event(struct notifier_block *this,
 		nexthop_flush_dev(dev, event);
 		break;
 	case NETDEV_CHANGE:
-		if (!(dev_get_flags(dev) & (IFF_RUNNING | IFF_LOWER_UP)))
+		if (!(netif_get_flags(dev) & (IFF_RUNNING | IFF_LOWER_UP)))
 			nexthop_flush_dev(dev, event);
 		break;
 	case NETDEV_CHANGEMTU:
diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index c85b1db74b1a..4f1d7d110302 100644
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c
@@ -6072,7 +6072,7 @@ static int inet6_fill_ifinfo(struct sk_buff *skb, struct inet6_dev *idev,
 	hdr->ifi_type = dev->type;
 	ifindex = READ_ONCE(dev->ifindex);
 	hdr->ifi_index = ifindex;
-	hdr->ifi_flags = dev_get_flags(dev);
+	hdr->ifi_flags = netif_get_flags(dev);
 	hdr->ifi_change = 0;
 
 	iflink = dev_get_iflink(dev);
diff --git a/net/mpls/af_mpls.c b/net/mpls/af_mpls.c
index 47d7dfd9ad09..25c88cba5c48 100644
--- a/net/mpls/af_mpls.c
+++ b/net/mpls/af_mpls.c
@@ -706,7 +706,7 @@ static int mpls_nh_assign_dev(struct net *net, struct mpls_route *rt,
 	} else {
 		unsigned int flags;
 
-		flags = dev_get_flags(dev);
+		flags = netif_get_flags(dev);
 		if (!(flags & (IFF_RUNNING | IFF_LOWER_UP)))
 			nh->nh_flags |= RTNH_F_LINKDOWN;
 	}
@@ -1616,14 +1616,14 @@ static int mpls_dev_notify(struct notifier_block *this, unsigned long event,
 			return notifier_from_errno(err);
 		break;
 	case NETDEV_UP:
-		flags = dev_get_flags(dev);
+		flags = netif_get_flags(dev);
 		if (flags & (IFF_RUNNING | IFF_LOWER_UP))
 			mpls_ifup(dev, RTNH_F_DEAD | RTNH_F_LINKDOWN);
 		else
 			mpls_ifup(dev, RTNH_F_DEAD);
 		break;
 	case NETDEV_CHANGE:
-		flags = dev_get_flags(dev);
+		flags = netif_get_flags(dev);
 		if (flags & (IFF_RUNNING | IFF_LOWER_UP)) {
 			mpls_ifup(dev, RTNH_F_DEAD | RTNH_F_LINKDOWN);
 		} else {
diff --git a/net/wireless/wext-core.c b/net/wireless/wext-core.c
index bea70eb6f034..c32a7c6903d5 100644
--- a/net/wireless/wext-core.c
+++ b/net/wireless/wext-core.c
@@ -431,7 +431,7 @@ static struct nlmsghdr *rtnetlink_ifinfo_prep(struct net_device *dev,
 	r->__ifi_pad = 0;
 	r->ifi_type = dev->type;
 	r->ifi_index = dev->ifindex;
-	r->ifi_flags = dev_get_flags(dev);
+	r->ifi_flags = netif_get_flags(dev);
 	r->ifi_change = 0;	/* Wireless changes don't affect those flags */
 
 	if (nla_put_string(skb, IFLA_IFNAME, dev->name))
-- 
2.50.0

[PATCH net-next v4 6/7] net: s/dev_set_threaded/netif_set_threaded/

[Stanislav Fomichev]

Commit cc34acd577f1 ("docs: net: document new locking reality")
introduced netif_ vs dev_ function semantics: the former expects locked
netdev, the latter takes care of the locking. We don't strictly
follow this semantics on either side, but there are more dev_xxx handlers
now that don't fit. Rename them to netif_xxx where appropriate.

Note that one dev_set_threaded call still remains in mt76 for debugfs file.

Signed-off-by: Stanislav Fomichev <sdf@fomichev.me>
---
 .../networking/net_cachelines/net_device.rst         |  2 +-
 drivers/net/ethernet/atheros/atl1c/atl1c_main.c      |  2 +-
 drivers/net/ethernet/mellanox/mlxsw/pci.c            |  2 +-
 drivers/net/ethernet/renesas/ravb_main.c             |  2 +-
 drivers/net/wireguard/device.c                       |  2 +-
 drivers/net/wireless/ath/ath10k/snoc.c               |  2 +-
 include/linux/netdevice.h                            |  1 +
 net/core/dev.c                                       |  6 +++---
 net/core/dev_api.c                                   | 12 ++++++++++++
 net/core/net-sysfs.c                                 |  2 +-
 10 files changed, 23 insertions(+), 10 deletions(-)

diff --git a/Documentation/networking/net_cachelines/net_device.rst b/Documentation/networking/net_cachelines/net_device.rst
index c69cc89c958e..2d3dc4692d20 100644
--- a/Documentation/networking/net_cachelines/net_device.rst
+++ b/Documentation/networking/net_cachelines/net_device.rst
@@ -165,7 +165,7 @@ struct sfp_bus*                     sfp_bus
 struct lock_class_key*              qdisc_tx_busylock
 bool                                proto_down
 unsigned:1                          wol_enabled
-unsigned:1                          threaded                                                            napi_poll(napi_enable,dev_set_threaded)
+unsigned:1                          threaded                                                            napi_poll(napi_enable,netif_set_threaded)
 unsigned_long:1                     see_all_hwtstamp_requests
 unsigned_long:1                     change_proto_down
 unsigned_long:1                     netns_immutable
diff --git a/drivers/net/ethernet/atheros/atl1c/atl1c_main.c b/drivers/net/ethernet/atheros/atl1c/atl1c_main.c
index ef1a51347351..3a9ad4a9c1cb 100644
--- a/drivers/net/ethernet/atheros/atl1c/atl1c_main.c
+++ b/drivers/net/ethernet/atheros/atl1c/atl1c_main.c
@@ -2688,7 +2688,7 @@ static int atl1c_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 	adapter->mii.mdio_write = atl1c_mdio_write;
 	adapter->mii.phy_id_mask = 0x1f;
 	adapter->mii.reg_num_mask = MDIO_CTRL_REG_MASK;
-	dev_set_threaded(netdev, true);
+	netif_set_threaded(netdev, true);
 	for (i = 0; i < adapter->rx_queue_count; ++i)
 		netif_napi_add(netdev, &adapter->rrd_ring[i].napi,
 			       atl1c_clean_rx);
diff --git a/drivers/net/ethernet/mellanox/mlxsw/pci.c b/drivers/net/ethernet/mellanox/mlxsw/pci.c
index 058dcabfaa2e..a2e97b712a3d 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/pci.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/pci.c
@@ -156,7 +156,7 @@ static int mlxsw_pci_napi_devs_init(struct mlxsw_pci *mlxsw_pci)
 	}
 	strscpy(mlxsw_pci->napi_dev_rx->name, "mlxsw_rx",
 		sizeof(mlxsw_pci->napi_dev_rx->name));
-	dev_set_threaded(mlxsw_pci->napi_dev_rx, true);
+	netif_set_threaded(mlxsw_pci->napi_dev_rx, true);
 
 	return 0;
 
diff --git a/drivers/net/ethernet/renesas/ravb_main.c b/drivers/net/ethernet/renesas/ravb_main.c
index c9f4976a3527..4e79bf88688a 100644
--- a/drivers/net/ethernet/renesas/ravb_main.c
+++ b/drivers/net/ethernet/renesas/ravb_main.c
@@ -3075,7 +3075,7 @@ static int ravb_probe(struct platform_device *pdev)
 	if (info->coalesce_irqs) {
 		netdev_sw_irq_coalesce_default_on(ndev);
 		if (num_present_cpus() == 1)
-			dev_set_threaded(ndev, true);
+			netif_set_threaded(ndev, true);
 	}
 
 	/* Network device register */
diff --git a/drivers/net/wireguard/device.c b/drivers/net/wireguard/device.c
index 4a529f1f9bea..5afec5a865f4 100644
--- a/drivers/net/wireguard/device.c
+++ b/drivers/net/wireguard/device.c
@@ -366,7 +366,7 @@ static int wg_newlink(struct net_device *dev,
 	if (ret < 0)
 		goto err_free_handshake_queue;
 
-	dev_set_threaded(dev, true);
+	netif_set_threaded(dev, true);
 	ret = register_netdevice(dev);
 	if (ret < 0)
 		goto err_uninit_ratelimiter;
diff --git a/drivers/net/wireless/ath/ath10k/snoc.c b/drivers/net/wireless/ath/ath10k/snoc.c
index d51f2e5a79a4..0ee68d3dad12 100644
--- a/drivers/net/wireless/ath/ath10k/snoc.c
+++ b/drivers/net/wireless/ath/ath10k/snoc.c
@@ -936,7 +936,7 @@ static int ath10k_snoc_hif_start(struct ath10k *ar)
 
 	bitmap_clear(ar_snoc->pending_ce_irqs, 0, CE_COUNT_MAX);
 
-	dev_set_threaded(ar->napi_dev, true);
+	netif_set_threaded(ar->napi_dev, true);
 	ath10k_core_napi_enable(ar);
 	/* IRQs are left enabled when we restart due to a firmware crash */
 	if (!test_bit(ATH10K_SNOC_FLAG_RECOVERY, &ar_snoc->flags))
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index d11804be460c..295457fbcfd1 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -588,6 +588,7 @@ static inline bool napi_complete(struct napi_struct *n)
 	return napi_complete_done(n, 0);
 }
 
+int netif_set_threaded(struct net_device *dev, bool threaded);
 int dev_set_threaded(struct net_device *dev, bool threaded);
 
 void napi_disable(struct napi_struct *n);
diff --git a/net/core/dev.c b/net/core/dev.c
index 2d3d9166e7af..dce23b445ad1 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -4797,7 +4797,7 @@ static inline void ____napi_schedule(struct softnet_data *sd,
 
 	if (test_bit(NAPI_STATE_THREADED, &napi->state)) {
 		/* Paired with smp_mb__before_atomic() in
-		 * napi_enable()/dev_set_threaded().
+		 * napi_enable()/netif_set_threaded().
 		 * Use READ_ONCE() to guarantee a complete
 		 * read on napi->thread. Only call
 		 * wake_up_process() when it's not NULL.
@@ -6960,7 +6960,7 @@ static void napi_stop_kthread(struct napi_struct *napi)
 	napi->thread = NULL;
 }
 
-int dev_set_threaded(struct net_device *dev, bool threaded)
+int netif_set_threaded(struct net_device *dev, bool threaded)
 {
 	struct napi_struct *napi;
 	int err = 0;
@@ -7004,7 +7004,7 @@ int dev_set_threaded(struct net_device *dev, bool threaded)
 
 	return err;
 }
-EXPORT_SYMBOL(dev_set_threaded);
+EXPORT_SYMBOL(netif_set_threaded);
 
 /**
  * netif_queue_set_napi - Associate queue with the napi
diff --git a/net/core/dev_api.c b/net/core/dev_api.c
index 1bf0153195f2..dd7f57013ce5 100644
--- a/net/core/dev_api.c
+++ b/net/core/dev_api.c
@@ -367,3 +367,15 @@ void netdev_state_change(struct net_device *dev)
 	netdev_unlock_ops(dev);
 }
 EXPORT_SYMBOL(netdev_state_change);
+
+int dev_set_threaded(struct net_device *dev, bool threaded)
+{
+	int ret;
+
+	netdev_lock(dev);
+	ret = netif_set_threaded(dev, threaded);
+	netdev_unlock(dev);
+
+	return ret;
+}
+EXPORT_SYMBOL(dev_set_threaded);
diff --git a/net/core/net-sysfs.c b/net/core/net-sysfs.c
index e41ad1890e49..c28cd6665444 100644
--- a/net/core/net-sysfs.c
+++ b/net/core/net-sysfs.c
@@ -757,7 +757,7 @@ static int modify_napi_threaded(struct net_device *dev, unsigned long val)
 	if (val != 0 && val != 1)
 		return -EOPNOTSUPP;
 
-	ret = dev_set_threaded(dev, val);
+	ret = netif_set_threaded(dev, val);
 
 	return ret;
 }
-- 
2.50.0

[PATCH net-next v4 7/7] net: s/dev_close_many/netif_close_many/

[Stanislav Fomichev]

Commit cc34acd577f1 ("docs: net: document new locking reality")
introduced netif_ vs dev_ function semantics: the former expects locked
netdev, the latter takes care of the locking. We don't strictly
follow this semantics on either side, but there are more dev_xxx handlers
now that don't fit. Rename them to netif_xxx where appropriate.

netif_close_many is used only by vlan/dsa and one mtk driver, so move it into
NETDEV_INTERNAL namespace.

Signed-off-by: Stanislav Fomichev <sdf@fomichev.me>
---
 drivers/net/ethernet/mediatek/mtk_eth_soc.c |  3 ++-
 include/linux/netdevice.h                   |  2 +-
 net/8021q/vlan.c                            |  3 ++-
 net/core/dev.c                              | 10 +++++-----
 net/dsa/dsa.c                               |  3 ++-
 net/dsa/user.c                              |  2 +-
 6 files changed, 13 insertions(+), 10 deletions(-)

diff --git a/drivers/net/ethernet/mediatek/mtk_eth_soc.c b/drivers/net/ethernet/mediatek/mtk_eth_soc.c
index 11ee7e1829bf..5a5fcde76dc0 100644
--- a/drivers/net/ethernet/mediatek/mtk_eth_soc.c
+++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.c
@@ -4967,7 +4967,7 @@ void mtk_eth_set_dma_device(struct mtk_eth *eth, struct device *dma_dev)
 		list_add_tail(&dev->close_list, &dev_list);
 	}
 
-	dev_close_many(&dev_list, false);
+	netif_close_many(&dev_list, false);
 
 	eth->dma_dev = dma_dev;
 
@@ -5610,3 +5610,4 @@ module_platform_driver(mtk_driver);
 MODULE_LICENSE("GPL");
 MODULE_AUTHOR("John Crispin <blogic@openwrt.org>");
 MODULE_DESCRIPTION("Ethernet driver for MediaTek SoC");
+MODULE_IMPORT_NS("NETDEV_INTERNAL");
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index 295457fbcfd1..e6131c529af4 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -3344,7 +3344,7 @@ 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);
+void netif_close_many(struct list_head *head, bool unlink);
 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);
diff --git a/net/8021q/vlan.c b/net/8021q/vlan.c
index df19a9c0c9d9..7b29e0be74bd 100644
--- a/net/8021q/vlan.c
+++ b/net/8021q/vlan.c
@@ -446,7 +446,7 @@ static int vlan_device_event(struct notifier_block *unused, unsigned long event,
 				list_add(&vlandev->close_list, &close_list);
 		}
 
-		dev_close_many(&close_list, false);
+		netif_close_many(&close_list, false);
 
 		list_for_each_entry_safe(vlandev, tmp, &close_list, close_list) {
 			vlan_stacked_transfer_operstate(dev, vlandev,
@@ -741,3 +741,4 @@ module_exit(vlan_cleanup_module);
 MODULE_DESCRIPTION("802.1Q/802.1ad VLAN Protocol");
 MODULE_LICENSE("GPL");
 MODULE_VERSION(DRV_VERSION);
+MODULE_IMPORT_NS("NETDEV_INTERNAL");
diff --git a/net/core/dev.c b/net/core/dev.c
index dce23b445ad1..26253802f6cd 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -1767,7 +1767,7 @@ static void __dev_close(struct net_device *dev)
 	list_del(&single);
 }
 
-void dev_close_many(struct list_head *head, bool unlink)
+void netif_close_many(struct list_head *head, bool unlink)
 {
 	struct net_device *dev, *tmp;
 
@@ -1785,7 +1785,7 @@ void dev_close_many(struct list_head *head, bool unlink)
 			list_del_init(&dev->close_list);
 	}
 }
-EXPORT_SYMBOL(dev_close_many);
+EXPORT_SYMBOL_NS_GPL(netif_close_many, "NETDEV_INTERNAL");
 
 void netif_close(struct net_device *dev)
 {
@@ -1793,7 +1793,7 @@ void netif_close(struct net_device *dev)
 		LIST_HEAD(single);
 
 		list_add(&dev->close_list, &single);
-		dev_close_many(&single, true);
+		netif_close_many(&single, true);
 		list_del(&single);
 	}
 }
@@ -12070,7 +12070,7 @@ void unregister_netdevice_many_notify(struct list_head *head,
 			netdev_lock(dev);
 		}
 	}
-	dev_close_many(&close_head, true);
+	netif_close_many(&close_head, true);
 	/* ... now unlock them and go over the rest. */
 	list_for_each_entry(dev, head, unreg_list) {
 		if (netdev_need_ops_lock(dev))
@@ -12078,7 +12078,7 @@ void unregister_netdevice_many_notify(struct list_head *head,
 		else
 			list_add_tail(&dev->close_list, &close_head);
 	}
-	dev_close_many(&close_head, true);
+	netif_close_many(&close_head, true);
 
 	list_for_each_entry(dev, head, unreg_list) {
 		/* And unlink it from device chain. */
diff --git a/net/dsa/dsa.c b/net/dsa/dsa.c
index 436a7e1b412a..5b01a0e43ebe 100644
--- a/net/dsa/dsa.c
+++ b/net/dsa/dsa.c
@@ -1621,7 +1621,7 @@ void dsa_switch_shutdown(struct dsa_switch *ds)
 	dsa_switch_for_each_cpu_port(dp, ds)
 		list_add(&dp->conduit->close_list, &close_list);
 
-	dev_close_many(&close_list, true);
+	netif_close_many(&close_list, true);
 
 	dsa_switch_for_each_user_port(dp, ds) {
 		conduit = dsa_port_to_conduit(dp);
@@ -1829,3 +1829,4 @@ MODULE_AUTHOR("Lennert Buytenhek <buytenh@wantstofly.org>");
 MODULE_DESCRIPTION("Driver for Distributed Switch Architecture switch chips");
 MODULE_LICENSE("GPL");
 MODULE_ALIAS("platform:dsa");
+MODULE_IMPORT_NS("NETDEV_INTERNAL");
diff --git a/net/dsa/user.c b/net/dsa/user.c
index e9334520c54a..f59d66f0975d 100644
--- a/net/dsa/user.c
+++ b/net/dsa/user.c
@@ -3604,7 +3604,7 @@ static int dsa_user_netdevice_event(struct notifier_block *nb,
 			list_add(&dp->user->close_list, &close_list);
 		}
 
-		dev_close_many(&close_list, true);
+		netif_close_many(&close_list, true);
 
 		return NOTIFY_OK;
 	}
-- 
2.50.0

Re: [PATCH net-next v4 6/7] net: s/dev_set_threaded/netif_set_threaded/

[Jakub Kicinski]

On Mon, 14 Jul 2025 15:18:54 -0700 Stanislav Fomichev wrote:
> Commit cc34acd577f1 ("docs: net: document new locking reality")
> introduced netif_ vs dev_ function semantics: the former expects locked
> netdev, the latter takes care of the locking. We don't strictly
> follow this semantics on either side, but there are more dev_xxx handlers
> now that don't fit. Rename them to netif_xxx where appropriate.
> 
> Note that one dev_set_threaded call still remains in mt76 for debugfs file.

I applied Sami's patch yesterday, this needs a respin, sorry.
-- 
pw-bot: cr