tap: Use iov_tail with tap_add_packet()

Use IOV_PEEK_HEADER() to get the ethernet header from the iovec.

Move the workaround about multiple iovec array from vu_handle_tx() to
tap_add_packet(). Removing the offset out of the iovec array should
reduce the iovec count to 1.

Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>

6 files changed, 48 insertions, 33 deletions

diff --git a/iov.c b/iov.c
index d20fbb8..0329613 100644
--- a/iov.c
+++ b/iov.c
@@ -200,7 +200,6 @@ size_t iov_tail_size(struct iov_tail *tail)
  *
  * Return: true if the item still contains any bytes, otherwise false
  */
-/* cppcheck-suppress unusedFunction */
 bool iov_drop_header(struct iov_tail *tail, size_t len)
 {
 	tail->off = tail->off + len;
diff --git a/iov.h b/iov.h
index 1821b40..ba1fda5 100644
--- a/iov.h
+++ b/iov.h
@@ -70,6 +70,18 @@ struct iov_tail {
 #define IOV_TAIL(iov_, cnt_, off_) \
 	(struct iov_tail){ .iov = (iov_), .cnt = (cnt_), .off = (off_) }
 
+/**
+ * IOV_TAIL_FROM_BUF() - Create a new IOV tail from a buffer
+ * @buf_:	Buffer address to use in the iovec
+ * @len_:	Buffer size
+ * @off_:	Byte offset in the buffer where the tail begins
+ */
+#define IOV_TAIL_FROM_BUF(buf_, len_, off_) \
+	IOV_TAIL((&(const struct iovec){ .iov_base = (buf_),		\
+					 .iov_len = (len_) }),		\
+		 1,							\
+		 (off_))
+
 bool iov_tail_prune(struct iov_tail *tail);
 size_t iov_tail_size(struct iov_tail *tail);
 bool iov_drop_header(struct iov_tail *tail, size_t len);
@@ -114,4 +126,12 @@ ssize_t iov_tail_clone(struct iovec *dst_iov, size_t dst_iov_cnt,
 	((__typeof__(var_) *)(iov_remove_header_((tail_), &(var_),	\
 				    sizeof(var_), __alignof__(var_))))
 
+/** IOV_DROP_HEADER() - Remove a typed header from an IOV tail
+ * @tail_:	IOV tail to remove header from (modified)
+ * @type_:	Data type of the header to remove
+ *
+ * Return: true if the tail still contains any bytes, otherwise false
+ */
+#define IOV_DROP_HEADER(tail_, type_) 	iov_drop_header((tail_), sizeof(type_))
+
 #endif /* IOVEC_H */
diff --git a/pcap.c b/pcap.c
index e9d8d80..54fba5c 100644
--- a/pcap.c
+++ b/pcap.c
@@ -74,6 +74,7 @@ static void pcap_frame(const struct iovec *iov, size_t iovcnt,
  * @pkt:	Pointer to data buffer, including L2 headers
  * @l2len:	L2 frame length
  */
+/* cppcheck-suppress unusedFunction */
 void pcap(const char *pkt, size_t l2len)
 {
 	struct iovec iov = { (char *)pkt, l2len };
diff --git a/tap.c b/tap.c
index cddc164..b1de0c5 100644
--- a/tap.c
+++ b/tap.c
@@ -1070,24 +1070,29 @@ void tap_handler(struct ctx *c, const struct timespec *now)
 /**
  * tap_add_packet() - Queue/capture packet, update notion of guest MAC address
  * @c:		Execution context
- * @l2len:	Total L2 packet length
- * @p:		Packet buffer
+ * @data:	Packet to add to the pool
  * @now:	Current timestamp
  */
-void tap_add_packet(struct ctx *c, ssize_t l2len, char *p,
+void tap_add_packet(struct ctx *c, struct iov_tail *data,
 		    const struct timespec *now)
 {
+	struct ethhdr eh_storage;
 	const struct ethhdr *eh;
 
-	pcap(p, l2len);
+	pcap_iov(data->iov, data->cnt, data->off);
 
-	eh = (struct ethhdr *)p;
+	eh = IOV_PEEK_HEADER(data, eh_storage);
+	if (!eh)
+		return;
 
 	if (memcmp(c->guest_mac, eh->h_source, ETH_ALEN)) {
 		memcpy(c->guest_mac, eh->h_source, ETH_ALEN);
 		proto_update_l2_buf(c->guest_mac, NULL);
 	}
 
+	iov_tail_prune(data);
+	ASSERT(data->cnt == 1); /* packet_add() doesn't support iovec */
+
 	switch (ntohs(eh->h_proto)) {
 	case ETH_P_ARP:
 	case ETH_P_IP:
@@ -1095,14 +1100,16 @@ void tap_add_packet(struct ctx *c, ssize_t l2len, char *p,
 			tap4_handler(c, pool_tap4, now);
 			pool_flush(pool_tap4);
 		}
-		packet_add(pool_tap4, l2len, p);
+		packet_add(pool_tap4, data->iov[0].iov_len - data->off,
+			   (char *)data->iov[0].iov_base + data->off);
 		break;
 	case ETH_P_IPV6:
 		if (pool_full(pool_tap6)) {
 			tap6_handler(c, pool_tap6, now);
 			pool_flush(pool_tap6);
 		}
-		packet_add(pool_tap6, l2len, p);
+		packet_add(pool_tap6, data->iov[0].iov_len - data->off,
+			   (char *)data->iov[0].iov_base + data->off);
 		break;
 	default:
 		break;
@@ -1170,6 +1177,7 @@ static void tap_passt_input(struct ctx *c, const struct timespec *now)
 
 	while (n >= (ssize_t)sizeof(uint32_t)) {
 		uint32_t l2len = ntohl_unaligned(p);
+		struct iov_tail data;
 
 		if (l2len < sizeof(struct ethhdr) || l2len > L2_MAX_LEN_PASST) {
 			err("Bad frame size from guest, resetting connection");
@@ -1184,7 +1192,8 @@ static void tap_passt_input(struct ctx *c, const struct timespec *now)
 		p += sizeof(uint32_t);
 		n -= sizeof(uint32_t);
 
-		tap_add_packet(c, l2len, p, now);
+		data = IOV_TAIL_FROM_BUF(p, l2len, 0);
+		tap_add_packet(c, &data, now);
 
 		p += l2len;
 		n -= l2len;
@@ -1228,6 +1237,8 @@ static void tap_pasta_input(struct ctx *c, const struct timespec *now)
 	for (n = 0;
 	     n <= (ssize_t)(sizeof(pkt_buf) - L2_MAX_LEN_PASTA);
 	     n += len) {
+		struct iov_tail data;
+
 		len = read(c->fd_tap, pkt_buf + n, L2_MAX_LEN_PASTA);
 
 		if (len == 0) {
@@ -1249,7 +1260,8 @@ static void tap_pasta_input(struct ctx *c, const struct timespec *now)
 		    len > (ssize_t)L2_MAX_LEN_PASTA)
 			continue;
 
-		tap_add_packet(c, len, pkt_buf + n, now);
+		data = IOV_TAIL_FROM_BUF(pkt_buf + n, len, 0);
+		tap_add_packet(c, &data, now);
 	}
 
 	tap_handler(c, now);
diff --git a/tap.h b/tap.h
index 936ae93..ce55108 100644
--- a/tap.h
+++ b/tap.h
@@ -119,7 +119,6 @@ void tap_sock_update_pool(void *base, size_t size);
 void tap_backend_init(struct ctx *c);
 void tap_flush_pools(void);
 void tap_handler(struct ctx *c, const struct timespec *now);
-void tap_add_packet(struct ctx *c, ssize_t l2len, char *p,
+void tap_add_packet(struct ctx *c, struct iov_tail *data,
 		    const struct timespec *now);
-
 #endif /* TAP_H */
diff --git a/vu_common.c b/vu_common.c
index 5e6fd4a..b77b214 100644
--- a/vu_common.c
+++ b/vu_common.c
@@ -163,7 +163,6 @@ static void vu_handle_tx(struct vu_dev *vdev, int index,
 	struct vu_virtq_element elem[VIRTQUEUE_MAX_SIZE];
 	struct iovec out_sg[VIRTQUEUE_MAX_SIZE];
 	struct vu_virtq *vq = &vdev->vq[index];
-	int hdrlen = sizeof(struct virtio_net_hdr_mrg_rxbuf);
 	int out_sg_count;
 	int count;
 
@@ -176,6 +175,7 @@ static void vu_handle_tx(struct vu_dev *vdev, int index,
 	while (count < VIRTQUEUE_MAX_SIZE &&
 	       out_sg_count + VU_MAX_TX_BUFFER_NB <= VIRTQUEUE_MAX_SIZE) {
 		int ret;
+		struct iov_tail data;
 
 		elem[count].out_num = VU_MAX_TX_BUFFER_NB;
 		elem[count].out_sg = &out_sg[out_sg_count];
@@ -191,26 +191,10 @@ static void vu_handle_tx(struct vu_dev *vdev, int index,
 			warn("virtio-net transmit queue contains no out buffers");
 			break;
 		}
-		if (elem[count].out_num == 1) {
-			tap_add_packet(vdev->context,
-				       elem[count].out_sg[0].iov_len - hdrlen,
-				       (char *)elem[count].out_sg[0].iov_base +
-				       hdrlen, now);
-		} else {
-			/* vnet header can be in a separate iovec */
-			if (elem[count].out_num != 2) {
-				debug("virtio-net transmit queue contains more than one buffer ([%d]: %u)",
-				      count, elem[count].out_num);
-			} else if (elem[count].out_sg[0].iov_len != (size_t)hdrlen) {
-				debug("virtio-net transmit queue entry not aligned on hdrlen ([%d]: %d != %zu)",
-				      count, hdrlen, elem[count].out_sg[0].iov_len);
-			} else {
-				tap_add_packet(vdev->context,
-					       elem[count].out_sg[1].iov_len,
-					       (char *)elem[count].out_sg[1].iov_base,
-					       now);
-			}
-		}
+
+		data = IOV_TAIL(elem[count].out_sg, elem[count].out_num, 0);
+		if (IOV_DROP_HEADER(&data, struct virtio_net_hdr_mrg_rxbuf))
+			tap_add_packet(vdev->context, &data, now);
 
 		count++;
 	}