5 files changed, 46 insertions, 18 deletions

diff --git a/net/xdp/xdp_umem.c b/net/xdp/xdp_umem.c
index 56d052bc65cb..56a28a686988 100644
--- a/net/xdp/xdp_umem.c
+++ b/net/xdp/xdp_umem.c
@@ -66,18 +66,31 @@ static void xdp_umem_release(struct xdp_umem *umem)
 	kfree(umem);
 }
 
+static void xdp_umem_release_deferred(struct work_struct *work)
+{
+	struct xdp_umem *umem = container_of(work, struct xdp_umem, work);
+
+	xdp_umem_release(umem);
+}
+
 void xdp_get_umem(struct xdp_umem *umem)
 {
 	refcount_inc(&umem->users);
 }
 
-void xdp_put_umem(struct xdp_umem *umem)
+void xdp_put_umem(struct xdp_umem *umem, bool defer_cleanup)
 {
 	if (!umem)
 		return;
 
-	if (refcount_dec_and_test(&umem->users))
-		xdp_umem_release(umem);
+	if (refcount_dec_and_test(&umem->users)) {
+		if (defer_cleanup) {
+			INIT_WORK(&umem->work, xdp_umem_release_deferred);
+			schedule_work(&umem->work);
+		} else {
+			xdp_umem_release(umem);
+		}
+	}
 }
 
 static int xdp_umem_pin_pages(struct xdp_umem *umem, unsigned long address)
diff --git a/net/xdp/xdp_umem.h b/net/xdp/xdp_umem.h
index 181fdda2f2a8..aa9fe2780410 100644
--- a/net/xdp/xdp_umem.h
+++ b/net/xdp/xdp_umem.h
@@ -9,7 +9,7 @@
 #include <net/xdp_sock_drv.h>
 
 void xdp_get_umem(struct xdp_umem *umem);
-void xdp_put_umem(struct xdp_umem *umem);
+void xdp_put_umem(struct xdp_umem *umem, bool defer_cleanup);
 struct xdp_umem *xdp_umem_create(struct xdp_umem_reg *mr);
 
 #endif /* XDP_UMEM_H_ */
diff --git a/net/xdp/xsk.c b/net/xdp/xsk.c
index cfbec3989a76..62504471fd20 100644
--- a/net/xdp/xsk.c
+++ b/net/xdp/xsk.c
@@ -211,6 +211,14 @@ static int __xsk_rcv(struct xdp_sock *xs, struct xdp_buff *xdp, u32 len,
 	return 0;
 }
 
+static bool xsk_tx_writeable(struct xdp_sock *xs)
+{
+	if (xskq_cons_present_entries(xs->tx) > xs->tx->nentries / 2)
+		return false;
+
+	return true;
+}
+
 static bool xsk_is_bound(struct xdp_sock *xs)
 {
 	if (READ_ONCE(xs->state) == XSK_BOUND) {
@@ -296,7 +304,8 @@ void xsk_tx_release(struct xsk_buff_pool *pool)
 	rcu_read_lock();
 	list_for_each_entry_rcu(xs, &pool->xsk_tx_list, tx_list) {
 		__xskq_cons_release(xs->tx);
-		xs->sk.sk_write_space(&xs->sk);
+		if (xsk_tx_writeable(xs))
+			xs->sk.sk_write_space(&xs->sk);
 	}
 	rcu_read_unlock();
 }
@@ -411,11 +420,7 @@ static int xsk_generic_xmit(struct sock *sk)
 		skb_shinfo(skb)->destructor_arg = (void *)(long)desc.addr;
 		skb->destructor = xsk_destruct_skb;
 
-		/* Hinder dev_direct_xmit from freeing the packet and
-		 * therefore completing it in the destructor
-		 */
-		refcount_inc(&skb->users);
-		err = dev_direct_xmit(skb, xs->queue_id);
+		err = __dev_direct_xmit(skb, xs->queue_id);
 		if  (err == NETDEV_TX_BUSY) {
 			/* Tell user-space to retry the send */
 			skb->destructor = sock_wfree;
@@ -429,12 +434,10 @@ static int xsk_generic_xmit(struct sock *sk)
 		/* Ignore NET_XMIT_CN as packet might have been sent */
 		if (err == NET_XMIT_DROP) {
 			/* SKB completed but not sent */
-			kfree_skb(skb);
 			err = -EBUSY;
 			goto out;
 		}
 
-		consume_skb(skb);
 		sent_frame = true;
 	}
 
@@ -442,7 +445,8 @@ static int xsk_generic_xmit(struct sock *sk)
 
 out:
 	if (sent_frame)
-		sk->sk_write_space(sk);
+		if (xsk_tx_writeable(xs))
+			sk->sk_write_space(sk);
 
 	mutex_unlock(&xs->mutex);
 	return err;
@@ -477,11 +481,13 @@ static int xsk_sendmsg(struct socket *sock, struct msghdr *m, size_t total_len)
 static __poll_t xsk_poll(struct file *file, struct socket *sock,
 			     struct poll_table_struct *wait)
 {
-	__poll_t mask = datagram_poll(file, sock, wait);
+	__poll_t mask = 0;
 	struct sock *sk = sock->sk;
 	struct xdp_sock *xs = xdp_sk(sk);
 	struct xsk_buff_pool *pool;
 
+	sock_poll_wait(file, sock, wait);
+
 	if (unlikely(!xsk_is_bound(xs)))
 		return mask;
 
@@ -497,7 +503,7 @@ static __poll_t xsk_poll(struct file *file, struct socket *sock,
 
 	if (xs->rx && !xskq_prod_is_empty(xs->rx))
 		mask |= EPOLLIN | EPOLLRDNORM;
-	if (xs->tx && !xskq_cons_is_full(xs->tx))
+	if (xs->tx && xsk_tx_writeable(xs))
 		mask |= EPOLLOUT | EPOLLWRNORM;
 
 	return mask;
@@ -1147,7 +1153,7 @@ static void xsk_destruct(struct sock *sk)
 		return;
 
 	if (!xp_put_pool(xs->pool))
-		xdp_put_umem(xs->umem);
+		xdp_put_umem(xs->umem, !xs->pool);
 
 	sk_refcnt_debug_dec(sk);
 }
diff --git a/net/xdp/xsk_buff_pool.c b/net/xdp/xsk_buff_pool.c
index 8a3bf4e1318e..d5adeee9d5d9 100644
--- a/net/xdp/xsk_buff_pool.c
+++ b/net/xdp/xsk_buff_pool.c
@@ -175,6 +175,7 @@ static int __xp_assign_dev(struct xsk_buff_pool *pool,
 
 	if (!pool->dma_pages) {
 		WARN(1, "Driver did not DMA map zero-copy buffers");
+		err = -EINVAL;
 		goto err_unreg_xsk;
 	}
 	pool->umem->zc = true;
@@ -185,8 +186,10 @@ err_unreg_xsk:
 err_unreg_pool:
 	if (!force_zc)
 		err = 0; /* fallback to copy mode */
-	if (err)
+	if (err) {
 		xsk_clear_pool_at_qid(netdev, queue_id);
+		dev_put(netdev);
+	}
 	return err;
 }
 
@@ -242,7 +245,7 @@ static void xp_release_deferred(struct work_struct *work)
 		pool->cq = NULL;
 	}
 
-	xdp_put_umem(pool->umem);
+	xdp_put_umem(pool->umem, false);
 	xp_destroy(pool);
 }
 
diff --git a/net/xdp/xsk_queue.h b/net/xdp/xsk_queue.h
index cdb9cf3cd136..9e71b9f27679 100644
--- a/net/xdp/xsk_queue.h
+++ b/net/xdp/xsk_queue.h
@@ -264,6 +264,12 @@ static inline bool xskq_cons_is_full(struct xsk_queue *q)
 		q->nentries;
 }
 
+static inline u32 xskq_cons_present_entries(struct xsk_queue *q)
+{
+	/* No barriers needed since data is not accessed */
+	return READ_ONCE(q->ring->producer) - READ_ONCE(q->ring->consumer);
+}
+
 /* Functions for producers */
 
 static inline bool xskq_prod_is_full(struct xsk_queue *q)