1 files changed, 131 insertions, 102 deletions

diff --git a/drivers/net/xen-netfront.c b/drivers/net/xen-netfront.c
index 8b18246ad999..9af2b027c19c 100644
--- a/drivers/net/xen-netfront.c
+++ b/drivers/net/xen-netfront.c
@@ -66,6 +66,10 @@ module_param_named(max_queues, xennet_max_queues, uint, 0644);
 MODULE_PARM_DESC(max_queues,
 		 "Maximum number of queues per virtual interface");
 
+static bool __read_mostly xennet_trusted = true;
+module_param_named(trusted, xennet_trusted, bool, 0644);
+MODULE_PARM_DESC(trusted, "Is the backend trusted");
+
 #define XENNET_TIMEOUT  (5 * HZ)
 
 static const struct ethtool_ops xennet_ethtool_ops;
@@ -78,8 +82,6 @@ struct netfront_cb {
 
 #define RX_COPY_THRESHOLD 256
 
-#define GRANT_INVALID_REF	0
-
 #define NET_TX_RING_SIZE __CONST_RING_SIZE(xen_netif_tx, XEN_PAGE_SIZE)
 #define NET_RX_RING_SIZE __CONST_RING_SIZE(xen_netif_rx, XEN_PAGE_SIZE)
 
@@ -175,6 +177,9 @@ struct netfront_info {
 	/* Is device behaving sane? */
 	bool broken;
 
+	/* Should skbs be bounced into a zeroed buffer? */
+	bool bounce;
+
 	atomic_t rx_gso_checksum_fixup;
 };
 
@@ -224,7 +229,7 @@ static grant_ref_t xennet_get_rx_ref(struct netfront_queue *queue,
 {
 	int i = xennet_rxidx(ri);
 	grant_ref_t ref = queue->grant_rx_ref[i];
-	queue->grant_rx_ref[i] = GRANT_INVALID_REF;
+	queue->grant_rx_ref[i] = INVALID_GRANT_REF;
 	return ref;
 }
 
@@ -273,7 +278,8 @@ static struct sk_buff *xennet_alloc_one_rx_buffer(struct netfront_queue *queue)
 	if (unlikely(!skb))
 		return NULL;
 
-	page = page_pool_dev_alloc_pages(queue->page_pool);
+	page = page_pool_alloc_pages(queue->page_pool,
+				     GFP_ATOMIC | __GFP_NOWARN | __GFP_ZERO);
 	if (unlikely(!page)) {
 		kfree_skb(skb);
 		return NULL;
@@ -424,17 +430,15 @@ static bool xennet_tx_buf_gc(struct netfront_queue *queue)
 			queue->tx_link[id] = TX_LINK_NONE;
 			skb = queue->tx_skbs[id];
 			queue->tx_skbs[id] = NULL;
-			if (unlikely(gnttab_query_foreign_access(
-				queue->grant_tx_ref[id]) != 0)) {
+			if (unlikely(!gnttab_end_foreign_access_ref(
+				queue->grant_tx_ref[id]))) {
 				dev_alert(dev,
 					  "Grant still in use by backend domain\n");
 				goto err;
 			}
-			gnttab_end_foreign_access_ref(
-				queue->grant_tx_ref[id], GNTMAP_readonly);
 			gnttab_release_grant_reference(
 				&queue->gref_tx_head, queue->grant_tx_ref[id]);
-			queue->grant_tx_ref[id] = GRANT_INVALID_REF;
+			queue->grant_tx_ref[id] = INVALID_GRANT_REF;
 			queue->grant_tx_page[id] = NULL;
 			add_id_to_list(&queue->tx_skb_freelist, queue->tx_link, id);
 			dev_kfree_skb_irq(skb);
@@ -669,6 +673,33 @@ static int xennet_xdp_xmit(struct net_device *dev, int n,
 	return nxmit;
 }
 
+static struct sk_buff *bounce_skb(const struct sk_buff *skb)
+{
+	unsigned int headerlen = skb_headroom(skb);
+	/* Align size to allocate full pages and avoid contiguous data leaks */
+	unsigned int size = ALIGN(skb_end_offset(skb) + skb->data_len,
+				  XEN_PAGE_SIZE);
+	struct sk_buff *n = alloc_skb(size, GFP_ATOMIC | __GFP_ZERO);
+
+	if (!n)
+		return NULL;
+
+	if (!IS_ALIGNED((uintptr_t)n->head, XEN_PAGE_SIZE)) {
+		WARN_ONCE(1, "misaligned skb allocated\n");
+		kfree_skb(n);
+		return NULL;
+	}
+
+	/* Set the data pointer */
+	skb_reserve(n, headerlen);
+	/* Set the tail pointer and length */
+	skb_put(n, skb->len);
+
+	BUG_ON(skb_copy_bits(skb, -headerlen, n->head, headerlen + skb->len));
+
+	skb_copy_header(n, skb);
+	return n;
+}
 
 #define MAX_XEN_SKB_FRAGS (65536 / XEN_PAGE_SIZE + 1)
 
@@ -722,9 +753,13 @@ static netdev_tx_t xennet_start_xmit(struct sk_buff *skb, struct net_device *dev
 
 	/* The first req should be at least ETH_HLEN size or the packet will be
 	 * dropped by netback.
+	 *
+	 * If the backend is not trusted bounce all data to zeroed pages to
+	 * avoid exposing contiguous data on the granted page not belonging to
+	 * the skb.
 	 */
-	if (unlikely(PAGE_SIZE - offset < ETH_HLEN)) {
-		nskb = skb_copy(skb, GFP_ATOMIC);
+	if (np->bounce || unlikely(PAGE_SIZE - offset < ETH_HLEN)) {
+		nskb = bounce_skb(skb);
 		if (!nskb)
 			goto drop;
 		dev_consume_skb_any(skb);
@@ -842,13 +877,35 @@ static int xennet_close(struct net_device *dev)
 	return 0;
 }
 
+static void xennet_destroy_queues(struct netfront_info *info)
+{
+	unsigned int i;
+
+	for (i = 0; i < info->netdev->real_num_tx_queues; i++) {
+		struct netfront_queue *queue = &info->queues[i];
+
+		if (netif_running(info->netdev))
+			napi_disable(&queue->napi);
+		netif_napi_del(&queue->napi);
+	}
+
+	kfree(info->queues);
+	info->queues = NULL;
+}
+
+static void xennet_uninit(struct net_device *dev)
+{
+	struct netfront_info *np = netdev_priv(dev);
+	xennet_destroy_queues(np);
+}
+
 static void xennet_set_rx_rsp_cons(struct netfront_queue *queue, RING_IDX val)
 {
 	unsigned long flags;
 
 	spin_lock_irqsave(&queue->rx_cons_lock, flags);
 	queue->rx.rsp_cons = val;
-	queue->rx_rsp_unconsumed = RING_HAS_UNCONSUMED_RESPONSES(&queue->rx);
+	queue->rx_rsp_unconsumed = XEN_RING_NR_UNCONSUMED_RESPONSES(&queue->rx);
 	spin_unlock_irqrestore(&queue->rx_cons_lock, flags);
 }
 
@@ -968,7 +1025,6 @@ static int xennet_get_responses(struct netfront_queue *queue,
 	struct device *dev = &queue->info->netdev->dev;
 	struct bpf_prog *xdp_prog;
 	struct xdp_buff xdp;
-	unsigned long ret;
 	int slots = 1;
 	int err = 0;
 	u32 verdict;
@@ -987,22 +1043,12 @@ static int xennet_get_responses(struct netfront_queue *queue,
 	}
 
 	for (;;) {
-		if (unlikely(rx->status < 0 ||
-			     rx->offset + rx->status > XEN_PAGE_SIZE)) {
-			if (net_ratelimit())
-				dev_warn(dev, "rx->offset: %u, size: %d\n",
-					 rx->offset, rx->status);
-			xennet_move_rx_slot(queue, skb, ref);
-			err = -EINVAL;
-			goto next;
-		}
-
 		/*
 		 * This definitely indicates a bug, either in this driver or in
 		 * the backend driver. In future this should flag the bad
 		 * situation to the system controller to reboot the backend.
 		 */
-		if (ref == GRANT_INVALID_REF) {
+		if (ref == INVALID_GRANT_REF) {
 			if (net_ratelimit())
 				dev_warn(dev, "Bad rx response id %d.\n",
 					 rx->id);
@@ -1010,8 +1056,23 @@ static int xennet_get_responses(struct netfront_queue *queue,
 			goto next;
 		}
 
-		ret = gnttab_end_foreign_access_ref(ref, 0);
-		BUG_ON(!ret);
+		if (unlikely(rx->status < 0 ||
+			     rx->offset + rx->status > XEN_PAGE_SIZE)) {
+			if (net_ratelimit())
+				dev_warn(dev, "rx->offset: %u, size: %d\n",
+					 rx->offset, rx->status);
+			xennet_move_rx_slot(queue, skb, ref);
+			err = -EINVAL;
+			goto next;
+		}
+
+		if (!gnttab_end_foreign_access_ref(ref)) {
+			dev_alert(dev,
+				  "Grant still in use by backend domain\n");
+			queue->info->broken = true;
+			dev_alert(dev, "Disabled for further use\n");
+			return -EINVAL;
+		}
 
 		gnttab_release_grant_reference(&queue->gref_rx_head, ref);
 
@@ -1031,8 +1092,10 @@ static int xennet_get_responses(struct netfront_queue *queue,
 			}
 		}
 		rcu_read_unlock();
-next:
+
 		__skb_queue_tail(list, skb);
+
+next:
 		if (!(rx->flags & XEN_NETRXF_more_data))
 			break;
 
@@ -1232,6 +1295,10 @@ static int xennet_poll(struct napi_struct *napi, int budget)
 					   &need_xdp_flush);
 
 		if (unlikely(err)) {
+			if (queue->info->broken) {
+				spin_unlock(&queue->rx_lock);
+				return 0;
+			}
 err:
 			while ((skb = __skb_dequeue(&tmpq)))
 				__skb_queue_tail(&errq, skb);
@@ -1360,10 +1427,9 @@ static void xennet_release_tx_bufs(struct netfront_queue *queue)
 		queue->tx_skbs[i] = NULL;
 		get_page(queue->grant_tx_page[i]);
 		gnttab_end_foreign_access(queue->grant_tx_ref[i],
-					  GNTMAP_readonly,
-					  (unsigned long)page_address(queue->grant_tx_page[i]));
+					  queue->grant_tx_page[i]);
 		queue->grant_tx_page[i] = NULL;
-		queue->grant_tx_ref[i] = GRANT_INVALID_REF;
+		queue->grant_tx_ref[i] = INVALID_GRANT_REF;
 		add_id_to_list(&queue->tx_skb_freelist, queue->tx_link, i);
 		dev_kfree_skb_irq(skb);
 	}
@@ -1384,7 +1450,7 @@ static void xennet_release_rx_bufs(struct netfront_queue *queue)
 			continue;
 
 		ref = queue->grant_rx_ref[id];
-		if (ref == GRANT_INVALID_REF)
+		if (ref == INVALID_GRANT_REF)
 			continue;
 
 		page = skb_frag_page(&skb_shinfo(skb)->frags[0]);
@@ -1393,9 +1459,8 @@ static void xennet_release_rx_bufs(struct netfront_queue *queue)
 		 * foreign access is ended (which may be deferred).
 		 */
 		get_page(page);
-		gnttab_end_foreign_access(ref, 0,
-					  (unsigned long)page_address(page));
-		queue->grant_rx_ref[id] = GRANT_INVALID_REF;
+		gnttab_end_foreign_access(ref, page);
+		queue->grant_rx_ref[id] = INVALID_GRANT_REF;
 
 		kfree_skb(skb);
 	}
@@ -1473,7 +1538,7 @@ static bool xennet_handle_rx(struct netfront_queue *queue, unsigned int *eoi)
 		return false;
 
 	spin_lock_irqsave(&queue->rx_cons_lock, flags);
-	work_queued = RING_HAS_UNCONSUMED_RESPONSES(&queue->rx);
+	work_queued = XEN_RING_NR_UNCONSUMED_RESPONSES(&queue->rx);
 	if (work_queued > queue->rx_rsp_unconsumed) {
 		queue->rx_rsp_unconsumed = work_queued;
 		*eoi = 0;
@@ -1611,6 +1676,7 @@ static int xennet_xdp(struct net_device *dev, struct netdev_bpf *xdp)
 }
 
 static const struct net_device_ops xennet_netdev_ops = {
+	.ndo_uninit          = xennet_uninit,
 	.ndo_open            = xennet_open,
 	.ndo_stop            = xennet_close,
 	.ndo_start_xmit      = xennet_start_xmit,
@@ -1733,8 +1799,8 @@ static int netfront_probe(struct xenbus_device *dev,
 static void xennet_end_access(int ref, void *page)
 {
 	/* This frees the page as a side-effect */
-	if (ref != GRANT_INVALID_REF)
-		gnttab_end_foreign_access(ref, 0, (unsigned long)page);
+	if (ref != INVALID_GRANT_REF)
+		gnttab_end_foreign_access(ref, virt_to_page(page));
 }
 
 static void xennet_disconnect_backend(struct netfront_info *info)
@@ -1770,8 +1836,8 @@ static void xennet_disconnect_backend(struct netfront_info *info)
 		xennet_end_access(queue->tx_ring_ref, queue->tx.sring);
 		xennet_end_access(queue->rx_ring_ref, queue->rx.sring);
 
-		queue->tx_ring_ref = GRANT_INVALID_REF;
-		queue->rx_ring_ref = GRANT_INVALID_REF;
+		queue->tx_ring_ref = INVALID_GRANT_REF;
+		queue->rx_ring_ref = INVALID_GRANT_REF;
 		queue->tx.sring = NULL;
 		queue->rx.sring = NULL;
 
@@ -1896,41 +1962,26 @@ static int setup_netfront(struct xenbus_device *dev,
 {
 	struct xen_netif_tx_sring *txs;
 	struct xen_netif_rx_sring *rxs;
-	grant_ref_t gref;
 	int err;
 
-	queue->tx_ring_ref = GRANT_INVALID_REF;
-	queue->rx_ring_ref = GRANT_INVALID_REF;
+	queue->tx_ring_ref = INVALID_GRANT_REF;
+	queue->rx_ring_ref = INVALID_GRANT_REF;
 	queue->rx.sring = NULL;
 	queue->tx.sring = NULL;
 
-	txs = (struct xen_netif_tx_sring *)get_zeroed_page(GFP_NOIO | __GFP_HIGH);
-	if (!txs) {
-		err = -ENOMEM;
-		xenbus_dev_fatal(dev, err, "allocating tx ring page");
+	err = xenbus_setup_ring(dev, GFP_NOIO | __GFP_HIGH, (void **)&txs,
+				1, &queue->tx_ring_ref);
+	if (err)
 		goto fail;
-	}
-	SHARED_RING_INIT(txs);
-	FRONT_RING_INIT(&queue->tx, txs, XEN_PAGE_SIZE);
 
-	err = xenbus_grant_ring(dev, txs, 1, &gref);
-	if (err < 0)
-		goto grant_tx_ring_fail;
-	queue->tx_ring_ref = gref;
+	XEN_FRONT_RING_INIT(&queue->tx, txs, XEN_PAGE_SIZE);
 
-	rxs = (struct xen_netif_rx_sring *)get_zeroed_page(GFP_NOIO | __GFP_HIGH);
-	if (!rxs) {
-		err = -ENOMEM;
-		xenbus_dev_fatal(dev, err, "allocating rx ring page");
-		goto alloc_rx_ring_fail;
-	}
-	SHARED_RING_INIT(rxs);
-	FRONT_RING_INIT(&queue->rx, rxs, XEN_PAGE_SIZE);
+	err = xenbus_setup_ring(dev, GFP_NOIO | __GFP_HIGH, (void **)&rxs,
+				1, &queue->rx_ring_ref);
+	if (err)
+		goto fail;
 
-	err = xenbus_grant_ring(dev, rxs, 1, &gref);
-	if (err < 0)
-		goto grant_rx_ring_fail;
-	queue->rx_ring_ref = gref;
+	XEN_FRONT_RING_INIT(&queue->rx, rxs, XEN_PAGE_SIZE);
 
 	if (feature_split_evtchn)
 		err = setup_netfront_split(queue);
@@ -1942,22 +1993,14 @@ static int setup_netfront(struct xenbus_device *dev,
 		err = setup_netfront_single(queue);
 
 	if (err)
-		goto alloc_evtchn_fail;
+		goto fail;
 
 	return 0;
 
-	/* If we fail to setup netfront, it is safe to just revoke access to
-	 * granted pages because backend is not accessing it at this point.
-	 */
-alloc_evtchn_fail:
-	gnttab_end_foreign_access_ref(queue->rx_ring_ref, 0);
-grant_rx_ring_fail:
-	free_page((unsigned long)rxs);
-alloc_rx_ring_fail:
-	gnttab_end_foreign_access_ref(queue->tx_ring_ref, 0);
-grant_tx_ring_fail:
-	free_page((unsigned long)txs);
-fail:
+ fail:
+	xenbus_teardown_ring((void **)&queue->rx.sring, 1, &queue->rx_ring_ref);
+	xenbus_teardown_ring((void **)&queue->tx.sring, 1, &queue->tx_ring_ref);
+
 	return err;
 }
 
@@ -1986,7 +2029,7 @@ static int xennet_init_queue(struct netfront_queue *queue)
 	queue->tx_pend_queue = TX_LINK_NONE;
 	for (i = 0; i < NET_TX_RING_SIZE; i++) {
 		queue->tx_link[i] = i + 1;
-		queue->grant_tx_ref[i] = GRANT_INVALID_REF;
+		queue->grant_tx_ref[i] = INVALID_GRANT_REF;
 		queue->grant_tx_page[i] = NULL;
 	}
 	queue->tx_link[NET_TX_RING_SIZE - 1] = TX_LINK_NONE;
@@ -1994,7 +2037,7 @@ static int xennet_init_queue(struct netfront_queue *queue)
 	/* Clear out rx_skbs */
 	for (i = 0; i < NET_RX_RING_SIZE; i++) {
 		queue->rx_skbs[i] = NULL;
-		queue->grant_rx_ref[i] = GRANT_INVALID_REF;
+		queue->grant_rx_ref[i] = INVALID_GRANT_REF;
 	}
 
 	/* A grant for every tx ring slot */
@@ -2103,22 +2146,6 @@ error:
 	return err;
 }
 
-static void xennet_destroy_queues(struct netfront_info *info)
-{
-	unsigned int i;
-
-	for (i = 0; i < info->netdev->real_num_tx_queues; i++) {
-		struct netfront_queue *queue = &info->queues[i];
-
-		if (netif_running(info->netdev))
-			napi_disable(&queue->napi);
-		netif_napi_del(&queue->napi);
-	}
-
-	kfree(info->queues);
-	info->queues = NULL;
-}
-
 
 
 static int xennet_create_page_pool(struct netfront_queue *queue)
@@ -2197,8 +2224,7 @@ static int xennet_create_queues(struct netfront_info *info,
 			return ret;
 		}
 
-		netif_napi_add(queue->info->netdev, &queue->napi,
-			       xennet_poll, 64);
+		netif_napi_add(queue->info->netdev, &queue->napi, xennet_poll);
 		if (netif_running(info->netdev))
 			napi_enable(&queue->napi);
 	}
@@ -2228,6 +2254,10 @@ static int talk_to_netback(struct xenbus_device *dev,
 
 	info->netdev->irq = 0;
 
+	/* Check if backend is trusted. */
+	info->bounce = !xennet_trusted ||
+		       !xenbus_read_unsigned(dev->nodename, "trusted", 1);
+
 	/* Check if backend supports multiple queues */
 	max_queues = xenbus_read_unsigned(info->xbdev->otherend,
 					  "multi-queue-max-queues", 1);
@@ -2395,6 +2425,9 @@ static int xennet_connect(struct net_device *dev)
 		return err;
 	if (np->netback_has_xdp_headroom)
 		pr_info("backend supports XDP headroom\n");
+	if (np->bounce)
+		dev_info(&np->xbdev->dev,
+			 "bouncing transmitted data to zeroed pages\n");
 
 	/* talk_to_netback() sets the correct number of queues */
 	num_queues = dev->real_num_tx_queues;
@@ -2430,10 +2463,6 @@ static int xennet_connect(struct net_device *dev)
 		if (queue->tx_irq != queue->rx_irq)
 			notify_remote_via_irq(queue->rx_irq);
 
-		spin_lock_irq(&queue->tx_lock);
-		xennet_tx_buf_gc(queue);
-		spin_unlock_irq(&queue->tx_lock);
-
 		spin_lock_bh(&queue->rx_lock);
 		xennet_alloc_rx_buffers(queue);
 		spin_unlock_bh(&queue->rx_lock);