rxrpc: Move the client conn cache management to the I/O thread

Move the management of the client connection cache to the I/O thread rather
than managing it from the namespace as an aggregate across all the local
endpoints within the namespace.

This will allow a load of locking to be got rid of in a future patch as
only the I/O thread will be looking at the this.

The downside is that the total number of cached connections on the system
can get higher because the limit is now per-local rather than per-netns.
We can, however, keep the number of client conns in use across the entire
netfs and use that to reduce the expiration time of idle connection.

Signed-off-by: David Howells <dhowells@redhat.com>
cc: Marc Dionne <marc.dionne@auristor.com>
cc: linux-afs@lists.infradead.org

net/rxrpc/ar-internal.h

@@ -76,13 +76,7 @@ struct rxrpc_net {
 
 	bool			live;
 
-	bool			kill_all_client_conns;
 	atomic_t		nr_client_conns;
-	spinlock_t		client_conn_cache_lock; /* Lock for ->*_client_conns */
-	struct mutex		client_conn_discard_lock; /* Prevent multiple discarders */
-	struct list_head	idle_client_conns;
-	struct work_struct	client_conn_reaper;
-	struct timer_list	client_conn_reap_timer;
 
 	struct hlist_head	local_endpoints;
 	struct mutex		local_mutex;	/* Lock for ->local_endpoints */
@@ -294,8 +288,16 @@ struct rxrpc_local {
 	struct sk_buff_head	rx_queue;	/* Received packets */
 	struct list_head	conn_attend_q;	/* Conns requiring immediate attention */
 	struct list_head	call_attend_q;	/* Calls requiring immediate attention */
+
 	struct rb_root		client_bundles;	/* Client connection bundles by socket params */
 	spinlock_t		client_bundles_lock; /* Lock for client_bundles */
+	bool			kill_all_client_conns;
+	spinlock_t		client_conn_cache_lock; /* Lock for ->*_client_conns */
+	struct list_head	idle_client_conns;
+	struct timer_list	client_conn_reap_timer;
+	unsigned long		client_conn_flags;
+#define RXRPC_CLIENT_CONN_REAP_TIMER	0	/* The client conn reap timer expired */
+
 	spinlock_t		lock;		/* access lock */
 	rwlock_t		services_lock;	/* lock for services list */
 	int			debug_id;	/* debug ID for printks */
@@ -946,8 +948,7 @@ void rxrpc_expose_client_call(struct rxrpc_call *);
 void rxrpc_disconnect_client_call(struct rxrpc_bundle *, struct rxrpc_call *);
 void rxrpc_deactivate_bundle(struct rxrpc_bundle *bundle);
 void rxrpc_put_client_conn(struct rxrpc_connection *, enum rxrpc_conn_trace);
-void rxrpc_discard_expired_client_conns(struct work_struct *);
+void rxrpc_discard_expired_client_conns(struct rxrpc_local *local);
-void rxrpc_destroy_all_client_connections(struct rxrpc_net *);
 void rxrpc_clean_up_local_conns(struct rxrpc_local *);
 
 /*

net/rxrpc/conn_client.c

@@ -578,17 +578,17 @@ static void rxrpc_activate_one_channel(struct rxrpc_connection *conn,
  */
 static void rxrpc_unidle_conn(struct rxrpc_bundle *bundle, struct rxrpc_connection *conn)
 {
-	struct rxrpc_net *rxnet = bundle->local->rxnet;
+	struct rxrpc_local *local = bundle->local;
 	bool drop_ref;
 
 	if (!list_empty(&conn->cache_link)) {
 		drop_ref = false;
-		spin_lock(&rxnet->client_conn_cache_lock);
+		spin_lock(&local->client_conn_cache_lock);
 		if (!list_empty(&conn->cache_link)) {
 			list_del_init(&conn->cache_link);
 			drop_ref = true;
 		}
-		spin_unlock(&rxnet->client_conn_cache_lock);
+		spin_unlock(&local->client_conn_cache_lock);
 		if (drop_ref)
 			rxrpc_put_connection(conn, rxrpc_conn_put_unidle);
 	}
@@ -710,14 +710,10 @@ out:
 int rxrpc_connect_call(struct rxrpc_call *call, gfp_t gfp)
 {
 	struct rxrpc_bundle *bundle;
-	struct rxrpc_local *local = call->local;
-	struct rxrpc_net *rxnet = local->rxnet;
 	int ret = 0;
 
 	_enter("{%d,%lx},", call->debug_id, call->user_call_ID);
 
-	rxrpc_discard_expired_client_conns(&rxnet->client_conn_reaper);
-
 	rxrpc_get_call(call, rxrpc_call_get_io_thread);
 
 	bundle = rxrpc_prep_call(call, gfp);
@@ -787,14 +783,14 @@ void rxrpc_expose_client_call(struct rxrpc_call *call)
 /*
  * Set the reap timer.
  */
-static void rxrpc_set_client_reap_timer(struct rxrpc_net *rxnet)
+static void rxrpc_set_client_reap_timer(struct rxrpc_local *local)
 {
-	if (!rxnet->kill_all_client_conns) {
+	if (!local->kill_all_client_conns) {
 		unsigned long now = jiffies;
 		unsigned long reap_at = now + rxrpc_conn_idle_client_expiry;
 
-		if (rxnet->live)
+		if (local->rxnet->live)
-			timer_reduce(&rxnet->client_conn_reap_timer, reap_at);
+			timer_reduce(&local->client_conn_reap_timer, reap_at);
 	}
 }
 
@@ -805,7 +801,7 @@ void rxrpc_disconnect_client_call(struct rxrpc_bundle *bundle, struct rxrpc_call
 {
 	struct rxrpc_connection *conn;
 	struct rxrpc_channel *chan = NULL;
-	struct rxrpc_net *rxnet = bundle->local->rxnet;
+	struct rxrpc_local *local = bundle->local;
 	unsigned int channel;
 	bool may_reuse;
 	u32 cid;
@@ -895,11 +891,11 @@ void rxrpc_disconnect_client_call(struct rxrpc_bundle *bundle, struct rxrpc_call
 		conn->idle_timestamp = jiffies;
 
 		rxrpc_get_connection(conn, rxrpc_conn_get_idle);
-		spin_lock(&rxnet->client_conn_cache_lock);
+		spin_lock(&local->client_conn_cache_lock);
-		list_move_tail(&conn->cache_link, &rxnet->idle_client_conns);
+		list_move_tail(&conn->cache_link, &local->idle_client_conns);
-		spin_unlock(&rxnet->client_conn_cache_lock);
+		spin_unlock(&local->client_conn_cache_lock);
 
-		rxrpc_set_client_reap_timer(rxnet);
+		rxrpc_set_client_reap_timer(local);
 	}
 
 out:
@@ -986,42 +982,34 @@ void rxrpc_kill_client_conn(struct rxrpc_connection *conn)
  * This may be called from conn setup or from a work item so cannot be
  * considered non-reentrant.
  */
-void rxrpc_discard_expired_client_conns(struct work_struct *work)
+void rxrpc_discard_expired_client_conns(struct rxrpc_local *local)
 {
 	struct rxrpc_connection *conn;
-	struct rxrpc_net *rxnet =
-		container_of(work, struct rxrpc_net, client_conn_reaper);
 	unsigned long expiry, conn_expires_at, now;
 	unsigned int nr_conns;
 
 	_enter("");
 
-	if (list_empty(&rxnet->idle_client_conns)) {
+	if (list_empty(&local->idle_client_conns)) {
 		_leave(" [empty]");
 		return;
 	}
 
-	/* Don't double up on the discarding */
-	if (!mutex_trylock(&rxnet->client_conn_discard_lock)) {
-		_leave(" [already]");
-		return;
-	}
-
 	/* We keep an estimate of what the number of conns ought to be after
 	 * we've discarded some so that we don't overdo the discarding.
 	 */
-	nr_conns = atomic_read(&rxnet->nr_client_conns);
+	nr_conns = atomic_read(&local->rxnet->nr_client_conns);
 
 next:
-	spin_lock(&rxnet->client_conn_cache_lock);
+	spin_lock(&local->client_conn_cache_lock);
 
-	if (list_empty(&rxnet->idle_client_conns))
+	if (list_empty(&local->idle_client_conns))
 		goto out;
 
-	conn = list_entry(rxnet->idle_client_conns.next,
+	conn = list_entry(local->idle_client_conns.next,
 			  struct rxrpc_connection, cache_link);
 
-	if (!rxnet->kill_all_client_conns) {
+	if (!local->kill_all_client_conns) {
 		/* If the number of connections is over the reap limit, we
 		 * expedite discard by reducing the expiry timeout.  We must,
 		 * however, have at least a short grace period to be able to do
@@ -1044,7 +1032,7 @@ next:
 	trace_rxrpc_client(conn, -1, rxrpc_client_discard);
 	list_del_init(&conn->cache_link);
 
-	spin_unlock(&rxnet->client_conn_cache_lock);
+	spin_unlock(&local->client_conn_cache_lock);
 
 	rxrpc_unbundle_conn(conn);
 	/* Drop the ->cache_link ref */
@@ -1062,32 +1050,11 @@ not_yet_expired:
 	 * then things get messier.
 	 */
 	_debug("not yet");
-	if (!rxnet->kill_all_client_conns)
+	if (!local->kill_all_client_conns)
-		timer_reduce(&rxnet->client_conn_reap_timer, conn_expires_at);
+		timer_reduce(&local->client_conn_reap_timer, conn_expires_at);
 
 out:
-	spin_unlock(&rxnet->client_conn_cache_lock);
+	spin_unlock(&local->client_conn_cache_lock);
-	mutex_unlock(&rxnet->client_conn_discard_lock);
-	_leave("");
-}
-
-/*
- * Preemptively destroy all the client connection records rather than waiting
- * for them to time out
- */
-void rxrpc_destroy_all_client_connections(struct rxrpc_net *rxnet)
-{
-	_enter("");
-
-	spin_lock(&rxnet->client_conn_cache_lock);
-	rxnet->kill_all_client_conns = true;
-	spin_unlock(&rxnet->client_conn_cache_lock);
-
-	del_timer_sync(&rxnet->client_conn_reap_timer);
-
-	if (!rxrpc_queue_work(&rxnet->client_conn_reaper))
-		_debug("destroy: queue failed");
-
 	_leave("");
 }
 
@@ -1097,14 +1064,19 @@ void rxrpc_destroy_all_client_connections(struct rxrpc_net *rxnet)
 void rxrpc_clean_up_local_conns(struct rxrpc_local *local)
 {
 	struct rxrpc_connection *conn, *tmp;
-	struct rxrpc_net *rxnet = local->rxnet;
 	LIST_HEAD(graveyard);
 
 	_enter("");
 
-	spin_lock(&rxnet->client_conn_cache_lock);
+	spin_lock(&local->client_conn_cache_lock);
+	local->kill_all_client_conns = true;
+	spin_unlock(&local->client_conn_cache_lock);
 
-	list_for_each_entry_safe(conn, tmp, &rxnet->idle_client_conns,
+	del_timer_sync(&local->client_conn_reap_timer);
+
+	spin_lock(&local->client_conn_cache_lock);
+
+	list_for_each_entry_safe(conn, tmp, &local->idle_client_conns,
 				 cache_link) {
 		if (conn->local == local) {
 			atomic_dec(&conn->active);
@@ -1113,7 +1085,7 @@ void rxrpc_clean_up_local_conns(struct rxrpc_local *local)
 		}
 	}
 
-	spin_unlock(&rxnet->client_conn_cache_lock);
+	spin_unlock(&local->client_conn_cache_lock);
 
 	while (!list_empty(&graveyard)) {
 		conn = list_entry(graveyard.next,

net/rxrpc/conn_object.c

@@ -470,7 +470,6 @@ void rxrpc_destroy_all_connections(struct rxrpc_net *rxnet)
 	_enter("");
 
 	atomic_dec(&rxnet->nr_conns);
-	rxrpc_destroy_all_client_connections(rxnet);
 
 	del_timer_sync(&rxnet->service_conn_reap_timer);
 	rxrpc_queue_work(&rxnet->service_conn_reaper);

net/rxrpc/io_thread.c

@@ -435,6 +435,10 @@ int rxrpc_io_thread(void *data)
 			continue;
 		}
 
+		if (test_and_clear_bit(RXRPC_CLIENT_CONN_REAP_TIMER,
+				       &local->client_conn_flags))
+			rxrpc_discard_expired_client_conns(local);
+
 		/* Deal with calls that want immediate attention. */
 		if ((call = list_first_entry_or_null(&local->call_attend_q,
 						     struct rxrpc_call,

net/rxrpc/local_object.c

@@ -82,6 +82,16 @@ static long rxrpc_local_cmp_key(const struct rxrpc_local *local,
 	}
 }
 
+static void rxrpc_client_conn_reap_timeout(struct timer_list *timer)
+{
+	struct rxrpc_local *local =
+		container_of(timer, struct rxrpc_local, client_conn_reap_timer);
+
+	if (local->kill_all_client_conns &&
+	    test_and_set_bit(RXRPC_CLIENT_CONN_REAP_TIMER, &local->client_conn_flags))
+		rxrpc_wake_up_io_thread(local);
+}
+
 /*
  * Allocate a new local endpoint.
  */
@@ -103,8 +113,15 @@ static struct rxrpc_local *rxrpc_alloc_local(struct net *net,
 		skb_queue_head_init(&local->rx_queue);
 		INIT_LIST_HEAD(&local->conn_attend_q);
 		INIT_LIST_HEAD(&local->call_attend_q);
+
 		local->client_bundles = RB_ROOT;
 		spin_lock_init(&local->client_bundles_lock);
+		local->kill_all_client_conns = false;
+		spin_lock_init(&local->client_conn_cache_lock);
+		INIT_LIST_HEAD(&local->idle_client_conns);
+		timer_setup(&local->client_conn_reap_timer,
+			    rxrpc_client_conn_reap_timeout, 0);
+
 		spin_lock_init(&local->lock);
 		rwlock_init(&local->services_lock);
 		local->debug_id = atomic_inc_return(&rxrpc_debug_id);

net/rxrpc/net_ns.c

@@ -10,15 +10,6 @@
 
 unsigned int rxrpc_net_id;
 
-static void rxrpc_client_conn_reap_timeout(struct timer_list *timer)
-{
-	struct rxrpc_net *rxnet =
-		container_of(timer, struct rxrpc_net, client_conn_reap_timer);
-
-	if (rxnet->live)
-		rxrpc_queue_work(&rxnet->client_conn_reaper);
-}
-
 static void rxrpc_service_conn_reap_timeout(struct timer_list *timer)
 {
 	struct rxrpc_net *rxnet =
@@ -63,14 +54,6 @@ static __net_init int rxrpc_init_net(struct net *net)
 		    rxrpc_service_conn_reap_timeout, 0);
 
 	atomic_set(&rxnet->nr_client_conns, 0);
-	rxnet->kill_all_client_conns = false;
-	spin_lock_init(&rxnet->client_conn_cache_lock);
-	mutex_init(&rxnet->client_conn_discard_lock);
-	INIT_LIST_HEAD(&rxnet->idle_client_conns);
-	INIT_WORK(&rxnet->client_conn_reaper,
-		  rxrpc_discard_expired_client_conns);
-	timer_setup(&rxnet->client_conn_reap_timer,
-		    rxrpc_client_conn_reap_timeout, 0);
 
 	INIT_HLIST_HEAD(&rxnet->local_endpoints);
 	mutex_init(&rxnet->local_mutex);