Merge tag 'ovpn-net-20260809' of https://github.com/OpenVPN/ovpn-net-next

Antonio Quartulli says:

====================
Included fixes:

* release key slot crypto transforms from a workqueue rather than an RCU
  callback, because crypto_free_aead() may sleep with async or hardware
  implementations
* run all deferred ovpn work on a module-owned workqueue and drain it on
  module exit, so no work item can still be executing module text after
  the module is unloaded
* finish crypto callback cleanup (key slot release and leftover skb)
  before dropping the peer reference that gates netdev unregistration
  and module removal
* avoid dereferencing a NULL key slot when userspace asks to kill a key
  that is not installed on the peer

* tag 'ovpn-net-20260809' of https://github.com/OpenVPN/ovpn-net-next:
  ovpn: defer key slot crypto freeing to workqueue
  ovpn: run deferred work on a module-owned workqueue
  ovpn: finish crypto callback cleanup before peer release
  ovpn: fix NULL dereference when killing missing key
====================

Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20260809212142.2249027-1-antonio@openvpn.net
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

drivers/net/ovpn/crypto.c

@@ -18,20 +18,12 @@
 #include "crypto_aead.h"
 #include "crypto.h"
 
-static void ovpn_ks_destroy_rcu(struct rcu_head *head)
-{
-	struct ovpn_crypto_key_slot *ks;
-
-	ks = container_of(head, struct ovpn_crypto_key_slot, rcu);
-	ovpn_aead_crypto_key_slot_destroy(ks);
-}
-
 void ovpn_crypto_key_slot_release(struct kref *kref)
 {
 	struct ovpn_crypto_key_slot *ks;
 
 	ks = container_of(kref, struct ovpn_crypto_key_slot, refcount);
-	call_rcu(&ks->rcu, ovpn_ks_destroy_rcu);
+	queue_rcu_work(ovpn_wq, &ks->free_work);
 }
 
 /* can only be invoked when all peer references have been dropped (i.e. RCU
@@ -58,15 +50,19 @@ void ovpn_crypto_state_release(struct ovpn_crypto_state *cs)
 bool ovpn_crypto_kill_key(struct ovpn_crypto_state *cs, u8 key_id)
 {
 	struct ovpn_crypto_key_slot *ks = NULL;
+	struct ovpn_crypto_key_slot *tmp;
+	int slot = 0;
 
 	spin_lock_bh(&cs->lock);
-	if (rcu_access_pointer(cs->slots[0])->key_id == key_id) {
-		ks = rcu_replace_pointer(cs->slots[0], NULL,
-					 lockdep_is_held(&cs->lock));
-	} else if (rcu_access_pointer(cs->slots[1])->key_id == key_id) {
-		ks = rcu_replace_pointer(cs->slots[1], NULL,
-					 lockdep_is_held(&cs->lock));
+	tmp = rcu_access_pointer(cs->slots[slot]);
+	if (!tmp || tmp->key_id != key_id) {
+		slot = 1;
+		tmp = rcu_access_pointer(cs->slots[slot]);
 	}
+
+	if (tmp && tmp->key_id == key_id)
+		ks = rcu_replace_pointer(cs->slots[slot], NULL,
+					 lockdep_is_held(&cs->lock));
 	spin_unlock_bh(&cs->lock);
 
 	if (ks)

drivers/net/ovpn/crypto.h

@@ -10,6 +10,8 @@
 #ifndef _NET_OVPN_OVPNCRYPTO_H_
 #define _NET_OVPN_OVPNCRYPTO_H_
 
+#include <linux/workqueue.h>
+
 #include "pktid.h"
 #include "proto.h"
 
@@ -45,8 +47,8 @@ struct ovpn_crypto_key_slot {
 
 	struct ovpn_pktid_recv pid_recv ____cacheline_aligned_in_smp;
 	struct ovpn_pktid_xmit pid_xmit ____cacheline_aligned_in_smp;
+	struct rcu_work free_work;
 	struct kref refcount;
-	struct rcu_head rcu;
 };
 
 struct ovpn_crypto_state {

drivers/net/ovpn/crypto_aead.c

@@ -9,6 +9,7 @@
 
 #include <crypto/aead.h>
 #include <linux/skbuff.h>
+#include <linux/workqueue.h>
 #include <net/ip.h>
 #include <net/ipv6.h>
 #include <net/udp.h>
@@ -380,13 +381,19 @@ error:
 	return ERR_PTR(ret);
 }
 
-void ovpn_aead_crypto_key_slot_destroy(struct ovpn_crypto_key_slot *ks)
+static void ovpn_aead_crypto_key_slot_free(struct ovpn_crypto_key_slot *ks)
 {
-	if (!ks)
-		return;
-
 	crypto_free_aead(ks->encrypt);
 	crypto_free_aead(ks->decrypt);
+}
+
+static void ovpn_aead_crypto_key_slot_free_work(struct work_struct *work)
+{
+	struct ovpn_crypto_key_slot *ks;
+
+	ks = container_of(to_rcu_work(work), struct ovpn_crypto_key_slot,
+			  free_work);
+	ovpn_aead_crypto_key_slot_free(ks);
 	kfree(ks);
 }
 
@@ -420,6 +427,7 @@ ovpn_aead_crypto_key_slot_new(const struct ovpn_key_config *kc)
 
 	ks->encrypt = NULL;
 	ks->decrypt = NULL;
+	INIT_RCU_WORK(&ks->free_work, ovpn_aead_crypto_key_slot_free_work);
 	kref_init(&ks->refcount);
 	ks->key_id = kc->key_id;
 
@@ -453,7 +461,8 @@ ovpn_aead_crypto_key_slot_new(const struct ovpn_key_config *kc)
 	return ks;
 
 destroy_ks:
-	ovpn_aead_crypto_key_slot_destroy(ks);
+	ovpn_aead_crypto_key_slot_free(ks);
+	kfree(ks);
 	return ERR_PTR(ret);
 }
 

drivers/net/ovpn/crypto_aead.h

@@ -22,7 +22,6 @@ int ovpn_aead_decrypt(struct ovpn_peer *peer, struct ovpn_crypto_key_slot *ks,
 
 struct ovpn_crypto_key_slot *
 ovpn_aead_crypto_key_slot_new(const struct ovpn_key_config *kc);
-void ovpn_aead_crypto_key_slot_destroy(struct ovpn_crypto_key_slot *ks);
 
 enum ovpn_cipher_alg ovpn_aead_crypto_alg(struct ovpn_crypto_key_slot *ks);
 

drivers/net/ovpn/io.c

@@ -204,10 +204,10 @@ drop:
 		ovpn_dev_dstats_rx_dropped(peer->ovpn->dev);
 	kfree_skb(skb);
 drop_nocount:
-	if (likely(peer))
-		ovpn_peer_put(peer);
 	if (likely(ks))
 		ovpn_crypto_key_slot_put(ks);
+	if (likely(peer))
+		ovpn_peer_put(peer);
 }
 
 /* RX path entry point: decrypt packet and forward it to the device */
@@ -302,11 +302,11 @@ err_unlock:
 err:
 	if (unlikely(skb))
 		ovpn_dev_dstats_tx_dropped(peer->ovpn->dev);
-	if (likely(peer))
-		ovpn_peer_put(peer);
+	kfree_skb(skb);
 	if (likely(ks))
 		ovpn_crypto_key_slot_put(ks);
-	kfree_skb(skb);
+	if (likely(peer))
+		ovpn_peer_put(peer);
 }
 
 static bool ovpn_encrypt_one(struct ovpn_peer *peer, struct sk_buff *skb)

drivers/net/ovpn/main.c

@@ -12,6 +12,7 @@
 #include <linux/module.h>
 #include <linux/netdevice.h>
 #include <linux/inetdevice.h>
+#include <linux/workqueue.h>
 #include <net/gro_cells.h>
 #include <net/ip.h>
 #include <net/rtnetlink.h>
@@ -26,6 +27,9 @@
 #include "tcp.h"
 #include "udp.h"
 
+/* module-owned workqueue on which all ovpn-specific work is queued */
+struct workqueue_struct *ovpn_wq;
+
 static void ovpn_priv_free(struct net_device *net)
 {
 	struct ovpn_priv *ovpn = netdev_priv(net);
@@ -264,10 +268,16 @@ static int __init ovpn_init(void)
 
 	ovpn_tcp_init();
 
+	ovpn_wq = alloc_workqueue("ovpn", WQ_PERCPU, 0);
+	if (!ovpn_wq) {
+		pr_err("ovpn: cannot allocate workqueue\n");
+		return -ENOMEM;
+	}
+
 	err = rtnl_link_register(&ovpn_link_ops);
 	if (err) {
 		pr_err("ovpn: can't register rtnl link ops: %d\n", err);
-		return err;
+		goto destroy_wq;
 	}
 
 	err = ovpn_nl_register();
@@ -280,6 +290,9 @@ static int __init ovpn_init(void)
 
 unreg_rtnl:
 	rtnl_link_unregister(&ovpn_link_ops);
+destroy_wq:
+	destroy_workqueue(ovpn_wq);
+	ovpn_wq = NULL;
 	return err;
 }
 
@@ -288,7 +301,11 @@ static __exit void ovpn_cleanup(void)
 	ovpn_nl_unregister();
 	rtnl_link_unregister(&ovpn_link_ops);
 
+	flush_workqueue(ovpn_wq);
 	rcu_barrier();
+
+	destroy_workqueue(ovpn_wq);
+	ovpn_wq = NULL;
 }
 
 module_init(ovpn_init);

drivers/net/ovpn/ovpnpriv.h

@@ -15,6 +15,10 @@
 #include <uapi/linux/if_link.h>
 #include <uapi/linux/ovpn.h>
 
+struct workqueue_struct;
+
+extern struct workqueue_struct *ovpn_wq;
+
 /**
  * struct ovpn_peer_collection - container of peers for MultiPeer mode
  * @by_id: table of peers index by ID

drivers/net/ovpn/peer.c

@@ -62,7 +62,7 @@ void ovpn_peer_keepalive_set(struct ovpn_peer *peer, u32 interval, u32 timeout)
 	/* now that interval and timeout have been changed, kick
 	 * off the worker so that the next delay can be recomputed
 	 */
-	mod_delayed_work(system_percpu_wq, &peer->ovpn->keepalive_work, 0);
+	mod_delayed_work(ovpn_wq, &peer->ovpn->keepalive_work, 0);
 }
 
 /**
@@ -1371,7 +1371,7 @@ static time64_t ovpn_peer_keepalive_work_single(struct ovpn_peer *peer,
 			   peer->id);
 		if (WARN_ON(!ovpn_peer_hold(peer)))
 			return 0;
-		if (!schedule_work(&peer->keepalive_work))
+		if (!queue_work(ovpn_wq, &peer->keepalive_work))
 			ovpn_peer_put(peer);
 	}
 
@@ -1463,8 +1463,8 @@ void ovpn_peer_keepalive_work(struct work_struct *work)
 		netdev_dbg(ovpn->dev,
 			   "scheduling keepalive work: now=%llu next_run=%llu delta=%llu\n",
 			   next_run, now, next_run - now);
-		schedule_delayed_work(&ovpn->keepalive_work,
-				      (next_run - now) * HZ);
+		queue_delayed_work(ovpn_wq, &ovpn->keepalive_work,
+				   (next_run - now) * HZ);
 	}
 	unlock_ovpn(ovpn, &release_list);
 }

drivers/net/ovpn/tcp.c

@@ -151,7 +151,7 @@ err:
 	/* take reference for deferred peer deletion. should never fail */
 	if (WARN_ON(!ovpn_peer_hold(peer)))
 		goto err_nopeer;
-	if (!schedule_work(&peer->tcp.defer_del_work))
+	if (!queue_work(ovpn_wq, &peer->tcp.defer_del_work))
 		ovpn_peer_put(peer);
 	ovpn_dev_dstats_rx_dropped(peer->ovpn->dev);
 err_nopeer:
@@ -284,13 +284,12 @@ static void ovpn_tcp_send_sock(struct ovpn_peer *peer, struct sock *sk)
 			 * stream therefore we abort the connection
 			 */
 			ovpn_peer_hold(peer);
-			if (!schedule_work(&peer->tcp.defer_del_work))
+			if (!queue_work(ovpn_wq, &peer->tcp.defer_del_work))
 				ovpn_peer_put(peer);
 
 			/* we bail out immediately and keep tx_in_progress set
 			 * to true. This way we prevent more TX attempts
-			 * which would lead to more invocations of
-			 * schedule_work()
+			 * which would lead to more invocations of queue_work()
 			 */
 			return;
 		}
@@ -487,7 +486,7 @@ static void ovpn_tcp_write_space(struct sock *sk)
 	rcu_read_lock();
 	sock = rcu_dereference_sk_user_data(sk);
 	if (likely(sock && sock->peer)) {
-		schedule_work(&sock->tcp_tx_work);
+		queue_work(ovpn_wq, &sock->tcp_tx_work);
 		sock->peer->tcp.sk_cb.sk_write_space(sk);
 	}
 	rcu_read_unlock();