Merge branch 'bridge_external_fdb_aging'

Scott Feldman says:

====================
bridge: don't age out externally added FDB entries

v3: Per davem review: add del_timer_sync on rocker port remove.

v2: Per Jiri review comment: add BR_DEFAULT_AGEING_TIME to defines

Siva originally proposed skipping externally added FDB entries in the bridge's
FDB garbage collection func, and moving the ageing of externally added entries
to the port driver/device.  This broke rocker, since rocker didn't have a
hardware (or software) mechanism for ageing out its learned FDB entries.

This patchset reintroduces Siva's bridge driver patch to skip externally added
entries and adds support in rocker so rocker can age out its own entries.
Rocker does this using a software timer similar to the bridge's FDB garbage
collection timer.  Other switchdev devices/drivers can use this software timer
method or program the device to nofity aged-out entries to the driver.

Updated switchdev.txt documentation to reflect current state-of-the-art.  This
removes one more XXX todo comment in switchdev.txt.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

5 files changed, 81 insertions, 23 deletions

diff --git a/Documentation/networking/switchdev.txt b/Documentation/networking/switchdev.txt
index 476df0496686..67e43ee7840a 100644
--- a/Documentation/networking/switchdev.txt
+++ b/Documentation/networking/switchdev.txt
@@ -239,20 +239,20 @@ The driver should initialize the attributes to the hardware defaults.
 FDB Ageing
 ^^^^^^^^^^
 
-There are two FDB ageing models supported: 1) ageing by the device, and 2)
-ageing by the kernel.  Ageing by the device is preferred if many FDB entries
-are supported.  The driver calls call_switchdev_notifiers(SWITCHDEV_FDB_DEL,
-...) to age out the FDB entry.  In this model, ageing by the kernel should be
-turned off.  XXX: how to turn off ageing in kernel on a per-port basis or
-otherwise prevent the kernel from ageing out the FDB entry?
-
-In the kernel ageing model, the standard bridge ageing mechanism is used to age
-out stale FDB entries.  To keep an FDB entry "alive", the driver should refresh
-the FDB entry by calling call_switchdev_notifiers(SWITCHDEV_FDB_ADD, ...).  The
+The bridge will skip ageing FDB entries marked with NTF_EXT_LEARNED and it is
+the responsibility of the port driver/device to age out these entries.  If the
+port device supports ageing, when the FDB entry expires, it will notify the
+driver which in turn will notify the bridge with SWITCHDEV_FDB_DEL.  If the
+device does not support ageing, the driver can simulate ageing using a
+garbage collection timer to monitor FBD entries.  Expired entries will be
+notified to the bridge using SWITCHDEV_FDB_DEL.  See rocker driver for
+example of driver running ageing timer.
+
+To keep an NTF_EXT_LEARNED entry "alive", the driver should refresh the FDB
+entry by calling call_switchdev_notifiers(SWITCHDEV_FDB_ADD, ...).  The
 notification will reset the FDB entry's last-used time to now.  The driver
 should rate limit refresh notifications, for example, no more than once a
-second.  If the FDB entry expires, fdb_delete is called to remove entry from
-the device.
+second.  (The last-used time is visible using the bridge -s fdb option).
 
 STP State Change on Port
 ^^^^^^^^^^^^^^^^^^^^^^^^
diff --git a/drivers/net/ethernet/rocker/rocker.c b/drivers/net/ethernet/rocker/rocker.c
index 34ac41ac9e61..32c5429ea5fe 100644
--- a/drivers/net/ethernet/rocker/rocker.c
+++ b/drivers/net/ethernet/rocker/rocker.c
@@ -152,8 +152,9 @@ struct rocker_fdb_tbl_entry {
 	struct hlist_node entry;
 	u32 key_crc32; /* key */
 	bool learned;
+	unsigned long touched;
 	struct rocker_fdb_tbl_key {
-		u32 pport;
+		struct rocker_port *rocker_port;
 		u8 addr[ETH_ALEN];
 		__be16 vlan_id;
 	} key;
@@ -220,6 +221,7 @@ struct rocker_port {
 	__be16 internal_vlan_id;
 	int stp_state;
 	u32 brport_flags;
+	unsigned long ageing_time;
 	bool ctrls[ROCKER_CTRL_MAX];
 	unsigned long vlan_bitmap[ROCKER_VLAN_BITMAP_LEN];
 	struct napi_struct napi_tx;
@@ -246,6 +248,7 @@ struct rocker {
 	u64 flow_tbl_next_cookie;
 	DECLARE_HASHTABLE(group_tbl, 16);
 	spinlock_t group_tbl_lock;		/* for group tbl accesses */
+	struct timer_list fdb_cleanup_timer;
 	DECLARE_HASHTABLE(fdb_tbl, 16);
 	spinlock_t fdb_tbl_lock;		/* for fdb tbl accesses */
 	unsigned long internal_vlan_bitmap[ROCKER_INTERNAL_VLAN_BITMAP_LEN];
@@ -3629,7 +3632,8 @@ static int rocker_port_fdb(struct rocker_port *rocker_port,
 		return -ENOMEM;
 
 	fdb->learned = (flags & ROCKER_OP_FLAG_LEARNED);
-	fdb->key.pport = rocker_port->pport;
+	fdb->touched = jiffies;
+	fdb->key.rocker_port = rocker_port;
 	ether_addr_copy(fdb->key.addr, addr);
 	fdb->key.vlan_id = vlan_id;
 	fdb->key_crc32 = crc32(~0, &fdb->key, sizeof(fdb->key));
@@ -3638,13 +3642,17 @@ static int rocker_port_fdb(struct rocker_port *rocker_port,
 
 	found = rocker_fdb_tbl_find(rocker, fdb);
 
-	if (removing && found) {
-		rocker_port_kfree(trans, fdb);
-		if (trans != SWITCHDEV_TRANS_PREPARE)
-			hash_del(&found->entry);
-	} else if (!removing && !found) {
+	if (found) {
+		found->touched = jiffies;
+		if (removing) {
+			rocker_port_kfree(trans, fdb);
+			if (trans != SWITCHDEV_TRANS_PREPARE)
+				hash_del(&found->entry);
+		}
+	} else if (!removing) {
 		if (trans != SWITCHDEV_TRANS_PREPARE)
-			hash_add(rocker->fdb_tbl, &fdb->entry, fdb->key_crc32);
+			hash_add(rocker->fdb_tbl, &fdb->entry,
+				 fdb->key_crc32);
 	}
 
 	spin_unlock_irqrestore(&rocker->fdb_tbl_lock, lock_flags);
@@ -3680,7 +3688,7 @@ static int rocker_port_fdb_flush(struct rocker_port *rocker_port,
 	spin_lock_irqsave(&rocker->fdb_tbl_lock, lock_flags);
 
 	hash_for_each_safe(rocker->fdb_tbl, bkt, tmp, found, entry) {
-		if (found->key.pport != rocker_port->pport)
+		if (found->key.rocker_port != rocker_port)
 			continue;
 		if (!found->learned)
 			continue;
@@ -3699,6 +3707,41 @@ err_out:
 	return err;
 }
 
+static void rocker_fdb_cleanup(unsigned long data)
+{
+	struct rocker *rocker = (struct rocker *)data;
+	struct rocker_port *rocker_port;
+	struct rocker_fdb_tbl_entry *entry;
+	struct hlist_node *tmp;
+	unsigned long next_timer = jiffies + BR_MIN_AGEING_TIME;
+	unsigned long expires;
+	unsigned long lock_flags;
+	int flags = ROCKER_OP_FLAG_NOWAIT | ROCKER_OP_FLAG_REMOVE |
+		    ROCKER_OP_FLAG_LEARNED;
+	int bkt;
+
+	spin_lock_irqsave(&rocker->fdb_tbl_lock, lock_flags);
+
+	hash_for_each_safe(rocker->fdb_tbl, bkt, tmp, entry, entry) {
+		if (!entry->learned)
+			continue;
+		rocker_port = entry->key.rocker_port;
+		expires = entry->touched + rocker_port->ageing_time;
+		if (time_before_eq(expires, jiffies)) {
+			rocker_port_fdb_learn(rocker_port, SWITCHDEV_TRANS_NONE,
+					      flags, entry->key.addr,
+					      entry->key.vlan_id);
+			hash_del(&entry->entry);
+		} else if (time_before(expires, next_timer)) {
+			next_timer = expires;
+		}
+	}
+
+	spin_unlock_irqrestore(&rocker->fdb_tbl_lock, lock_flags);
+
+	mod_timer(&rocker->fdb_cleanup_timer, round_jiffies_up(next_timer));
+}
+
 static int rocker_port_router_mac(struct rocker_port *rocker_port,
 				  enum switchdev_trans trans, int flags,
 				  __be16 vlan_id)
@@ -4547,7 +4590,7 @@ static int rocker_port_fdb_dump(const struct rocker_port *rocker_port,
 
 	spin_lock_irqsave(&rocker->fdb_tbl_lock, lock_flags);
 	hash_for_each_safe(rocker->fdb_tbl, bkt, tmp, found, entry) {
-		if (found->key.pport != rocker_port->pport)
+		if (found->key.rocker_port != rocker_port)
 			continue;
 		fdb->addr = found->key.addr;
 		fdb->ndm_state = NUD_REACHABLE;
@@ -4969,6 +5012,7 @@ static int rocker_probe_port(struct rocker *rocker, unsigned int port_number)
 	rocker_port->port_number = port_number;
 	rocker_port->pport = port_number + 1;
 	rocker_port->brport_flags = BR_LEARNING | BR_LEARNING_SYNC;
+	rocker_port->ageing_time = BR_DEFAULT_AGEING_TIME;
 	INIT_LIST_HEAD(&rocker_port->trans_mem);
 
 	rocker_port_dev_addr_init(rocker_port);
@@ -5183,6 +5227,10 @@ static int rocker_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 		goto err_init_tbls;
 	}
 
+	setup_timer(&rocker->fdb_cleanup_timer, rocker_fdb_cleanup,
+		    (unsigned long) rocker);
+	mod_timer(&rocker->fdb_cleanup_timer, jiffies);
+
 	err = rocker_probe_ports(rocker);
 	if (err) {
 		dev_err(&pdev->dev, "failed to probe ports\n");
@@ -5195,6 +5243,7 @@ static int rocker_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 	return 0;
 
 err_probe_ports:
+	del_timer_sync(&rocker->fdb_cleanup_timer);
 	rocker_free_tbls(rocker);
 err_init_tbls:
 	free_irq(rocker_msix_vector(rocker, ROCKER_MSIX_VEC_EVENT), rocker);
@@ -5222,6 +5271,7 @@ static void rocker_remove(struct pci_dev *pdev)
 {
 	struct rocker *rocker = pci_get_drvdata(pdev);
 
+	del_timer_sync(&rocker->fdb_cleanup_timer);
 	rocker_free_tbls(rocker);
 	rocker_write32(rocker, CONTROL, ROCKER_CONTROL_RESET);
 	rocker_remove_ports(rocker);
diff --git a/include/linux/if_bridge.h b/include/linux/if_bridge.h
index dad8b00beed2..a338a688ee4a 100644
--- a/include/linux/if_bridge.h
+++ b/include/linux/if_bridge.h
@@ -46,6 +46,12 @@ struct br_ip_list {
 #define BR_LEARNING_SYNC	BIT(9)
 #define BR_PROXYARP_WIFI	BIT(10)
 
+/* values as per ieee8021QBridgeFdbAgingTime */
+#define BR_MIN_AGEING_TIME	(10 * HZ)
+#define BR_MAX_AGEING_TIME	(1000000 * HZ)
+
+#define BR_DEFAULT_AGEING_TIME	(300 * HZ)
+
 extern void brioctl_set(int (*ioctl_hook)(struct net *, unsigned int, void __user *));
 
 typedef int br_should_route_hook_t(struct sk_buff *skb);
diff --git a/net/bridge/br_device.c b/net/bridge/br_device.c
index 6ed2feb51e3c..2f81624a8257 100644
--- a/net/bridge/br_device.c
+++ b/net/bridge/br_device.c
@@ -391,7 +391,7 @@ void br_dev_setup(struct net_device *dev)
 	br->bridge_max_age = br->max_age = 20 * HZ;
 	br->bridge_hello_time = br->hello_time = 2 * HZ;
 	br->bridge_forward_delay = br->forward_delay = 15 * HZ;
-	br->ageing_time = 300 * HZ;
+	br->ageing_time = BR_DEFAULT_AGEING_TIME;
 
 	br_netfilter_rtable_init(br);
 	br_stp_timer_init(br);
diff --git a/net/bridge/br_fdb.c b/net/bridge/br_fdb.c
index 9e9875da0a4f..6663cc0789a6 100644
--- a/net/bridge/br_fdb.c
+++ b/net/bridge/br_fdb.c
@@ -299,6 +299,8 @@ void br_fdb_cleanup(unsigned long _data)
 			unsigned long this_timer;
 			if (f->is_static)
 				continue;
+			if (f->added_by_external_learn)
+				continue;
 			this_timer = f->updated + delay;
 			if (time_before_eq(this_timer, jiffies))
 				fdb_delete(br, f);