net/tls: Use RCU API to access tls_ctx->netdev

Currently, tls_device_down synchronizes with tls_device_resync_rx using
RCU, however, the pointer to netdev is stored using WRITE_ONCE and
loaded using READ_ONCE.

Although such approach is technically correct (rcu_dereference is
essentially a READ_ONCE, and rcu_assign_pointer uses WRITE_ONCE to store
NULL), using special RCU helpers for pointers is more valid, as it
includes additional checks and might change the implementation
transparently to the callers.

Mark the netdev pointer as __rcu and use the correct RCU helpers to
access it. For non-concurrent access pass the right conditions that
guarantee safe access (locks taken, refcount value). Also use the
correct helper in mlx5e, where even READ_ONCE was missing.

The transition to RCU exposes existing issues, fixed by this commit:

1. bond_tls_device_xmit could read netdev twice, and it could become
NULL the second time, after the NULL check passed.

2. Drivers shouldn't stop processing the last packet if tls_device_down
just set netdev to NULL, before tls_dev_del was called. This prevents a
possible packet drop when transitioning to the fallback software mode.

Fixes: 89df6a810470 ("net/bonding: Implement TLS TX device offload")
Fixes: c55dcdd435aa ("net/tls: Fix use-after-free after the TLS device goes down and up")
Signed-off-by: Maxim Mikityanskiy <maximmi@nvidia.com>
Link: https://lore.kernel.org/r/20220810081602.1435800-1-maximmi@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

2 files changed, 31 insertions, 10 deletions

diff --git a/net/tls/tls_device.c b/net/tls/tls_device.c
index 6ed41474bdf8..0f983e5f7dde 100644
--- a/net/tls/tls_device.c
+++ b/net/tls/tls_device.c
@@ -71,7 +71,13 @@ static void tls_device_tx_del_task(struct work_struct *work)
 	struct tls_offload_context_tx *offload_ctx =
 		container_of(work, struct tls_offload_context_tx, destruct_work);
 	struct tls_context *ctx = offload_ctx->ctx;
-	struct net_device *netdev = ctx->netdev;
+	struct net_device *netdev;
+
+	/* Safe, because this is the destroy flow, refcount is 0, so
+	 * tls_device_down can't store this field in parallel.
+	 */
+	netdev = rcu_dereference_protected(ctx->netdev,
+					   !refcount_read(&ctx->refcount));
 
 	netdev->tlsdev_ops->tls_dev_del(netdev, ctx, TLS_OFFLOAD_CTX_DIR_TX);
 	dev_put(netdev);
@@ -81,6 +87,7 @@ static void tls_device_tx_del_task(struct work_struct *work)
 
 static void tls_device_queue_ctx_destruction(struct tls_context *ctx)
 {
+	struct net_device *netdev;
 	unsigned long flags;
 	bool async_cleanup;
 
@@ -91,7 +98,14 @@ static void tls_device_queue_ctx_destruction(struct tls_context *ctx)
 	}
 
 	list_del(&ctx->list); /* Remove from tls_device_list / tls_device_down_list */
-	async_cleanup = ctx->netdev && ctx->tx_conf == TLS_HW;
+
+	/* Safe, because this is the destroy flow, refcount is 0, so
+	 * tls_device_down can't store this field in parallel.
+	 */
+	netdev = rcu_dereference_protected(ctx->netdev,
+					   !refcount_read(&ctx->refcount));
+
+	async_cleanup = netdev && ctx->tx_conf == TLS_HW;
 	if (async_cleanup) {
 		struct tls_offload_context_tx *offload_ctx = tls_offload_ctx_tx(ctx);
 
@@ -229,7 +243,8 @@ static void tls_device_resync_tx(struct sock *sk, struct tls_context *tls_ctx,
 
 	trace_tls_device_tx_resync_send(sk, seq, rcd_sn);
 	down_read(&device_offload_lock);
-	netdev = tls_ctx->netdev;
+	netdev = rcu_dereference_protected(tls_ctx->netdev,
+					   lockdep_is_held(&device_offload_lock));
 	if (netdev)
 		err = netdev->tlsdev_ops->tls_dev_resync(netdev, sk, seq,
 							 rcd_sn,
@@ -710,7 +725,7 @@ static void tls_device_resync_rx(struct tls_context *tls_ctx,
 
 	trace_tls_device_rx_resync_send(sk, seq, rcd_sn, rx_ctx->resync_type);
 	rcu_read_lock();
-	netdev = READ_ONCE(tls_ctx->netdev);
+	netdev = rcu_dereference(tls_ctx->netdev);
 	if (netdev)
 		netdev->tlsdev_ops->tls_dev_resync(netdev, sk, seq, rcd_sn,
 						   TLS_OFFLOAD_CTX_DIR_RX);
@@ -1035,7 +1050,7 @@ static void tls_device_attach(struct tls_context *ctx, struct sock *sk,
 	if (sk->sk_destruct != tls_device_sk_destruct) {
 		refcount_set(&ctx->refcount, 1);
 		dev_hold(netdev);
-		ctx->netdev = netdev;
+		RCU_INIT_POINTER(ctx->netdev, netdev);
 		spin_lock_irq(&tls_device_lock);
 		list_add_tail(&ctx->list, &tls_device_list);
 		spin_unlock_irq(&tls_device_lock);
@@ -1306,7 +1321,8 @@ void tls_device_offload_cleanup_rx(struct sock *sk)
 	struct net_device *netdev;
 
 	down_read(&device_offload_lock);
-	netdev = tls_ctx->netdev;
+	netdev = rcu_dereference_protected(tls_ctx->netdev,
+					   lockdep_is_held(&device_offload_lock));
 	if (!netdev)
 		goto out;
 
@@ -1315,7 +1331,7 @@ void tls_device_offload_cleanup_rx(struct sock *sk)
 
 	if (tls_ctx->tx_conf != TLS_HW) {
 		dev_put(netdev);
-		tls_ctx->netdev = NULL;
+		rcu_assign_pointer(tls_ctx->netdev, NULL);
 	} else {
 		set_bit(TLS_RX_DEV_CLOSED, &tls_ctx->flags);
 	}
@@ -1335,7 +1351,11 @@ static int tls_device_down(struct net_device *netdev)
 
 	spin_lock_irqsave(&tls_device_lock, flags);
 	list_for_each_entry_safe(ctx, tmp, &tls_device_list, list) {
-		if (ctx->netdev != netdev ||
+		struct net_device *ctx_netdev =
+			rcu_dereference_protected(ctx->netdev,
+						  lockdep_is_held(&device_offload_lock));
+
+		if (ctx_netdev != netdev ||
 		    !refcount_inc_not_zero(&ctx->refcount))
 			continue;
 
@@ -1352,7 +1372,7 @@ static int tls_device_down(struct net_device *netdev)
 		/* Stop the RX and TX resync.
 		 * tls_dev_resync must not be called after tls_dev_del.
 		 */
-		WRITE_ONCE(ctx->netdev, NULL);
+		rcu_assign_pointer(ctx->netdev, NULL);
 
 		/* Start skipping the RX resync logic completely. */
 		set_bit(TLS_RX_DEV_DEGRADED, &ctx->flags);
diff --git a/net/tls/tls_device_fallback.c b/net/tls/tls_device_fallback.c
index 618cee704217..7dfc8023e0f1 100644
--- a/net/tls/tls_device_fallback.c
+++ b/net/tls/tls_device_fallback.c
@@ -426,7 +426,8 @@ struct sk_buff *tls_validate_xmit_skb(struct sock *sk,
 				      struct net_device *dev,
 				      struct sk_buff *skb)
 {
-	if (dev == tls_get_ctx(sk)->netdev || netif_is_bond_master(dev))
+	if (dev == rcu_dereference_bh(tls_get_ctx(sk)->netdev) ||
+	    netif_is_bond_master(dev))
 		return skb;
 
 	return tls_sw_fallback(sk, skb);