Split tap_ip_send() into IPv4 and IPv6 specific functions

The IPv4 and IPv6 paths in tap_ip_send() have very little in common, and
it turns out that every caller (statically) knows if it is using IPv4 or
IPv6.  So split into separate tap_ip4_send() and tap_ip6_send() functions.
Use a new tap_l2_hdr() function for the very small common part.

While we're there, make some minor cleanups:
  - We were double writing some fields in the IPv6 header, so that it
    temporary matched the pseudo-header for checksum calculation.  With
    recent checksum reworks, this isn't neccessary any more.
  - We don't use any IPv4 header options, so use some sizeof() constructs
    instead of some open coded values for header length.
  - The comment used to say that the flow label was for TCP over IPv6, but
    in fact the only thing we used it for was DHCPv6 over UDP traffic

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

dhcpv6.c

@@ -531,7 +531,7 @@ int dhcpv6(struct ctx *c, const struct pool *p,
 
 			resp_not_on_link.hdr.xid = mh->xid;
 
-			tap_ip_send(c, src, IPPROTO_UDP,
+			tap_ip6_send(c, src, IPPROTO_UDP,
 				     (char *)&resp_not_on_link, n, mh->xid);
 
 			return 1;
@@ -580,7 +580,7 @@ int dhcpv6(struct ctx *c, const struct pool *p,
 
 	resp.hdr.xid = mh->xid;
 
-	tap_ip_send(c, src, IPPROTO_UDP, (char *)&resp, n, mh->xid);
+	tap_ip6_send(c, src, IPPROTO_UDP, (char *)&resp, n, mh->xid);
 	c->ip6.addr_seen = c->ip6.addr;
 
 	return 1;

icmp.c

@@ -69,10 +69,6 @@ static uint8_t icmp_act[IP_VERSIONS][DIV_ROUND_UP(ICMP_NUM_IDS, 8)];
 void icmp_sock_handler(const struct ctx *c, union epoll_ref ref,
 		       uint32_t events, const struct timespec *now)
 {
-	struct in6_addr a6 = { .s6_addr = {    0,    0,    0,    0,
-					       0,    0,    0,    0,
-					       0,    0, 0xff, 0xff,
-					       0,    0,    0,    0 } };
 	union icmp_epoll_ref *iref = &ref.r.p.icmp;
 	struct sockaddr_storage sr;
 	socklen_t sl = sizeof(sr);
@@ -109,7 +105,7 @@ void icmp_sock_handler(const struct ctx *c, union epoll_ref ref,
 			icmp_id_map[V6][id].seq = seq;
 		}
 
-		tap_ip_send(c, &sr6->sin6_addr, IPPROTO_ICMPV6, buf, n, 0);
+		tap_ip6_send(c, &sr6->sin6_addr, IPPROTO_ICMPV6, buf, n, 0);
 	} else {
 		struct sockaddr_in *sr4 = (struct sockaddr_in *)&sr;
 		struct icmphdr *ih = (struct icmphdr *)buf;
@@ -127,9 +123,7 @@ void icmp_sock_handler(const struct ctx *c, union epoll_ref ref,
 			icmp_id_map[V4][id].seq = seq;
 		}
 
-		memcpy(&a6.s6_addr[12], &sr4->sin_addr, sizeof(sr4->sin_addr));
+		tap_ip4_send(c, sr4->sin_addr.s_addr, IPPROTO_ICMP, buf, n);
-
-		tap_ip_send(c, &a6, IPPROTO_ICMP, buf, n, 0);
 	}
 }
 

tap.c

@@ -113,77 +113,97 @@ const struct in6_addr *tap_ip6_daddr(const struct ctx *c,
 }
 
 /**
- * tap_ip_send() - Send IP packet, with L2 headers, calculating L3/L4 checksums
+ * tap_push_l2h() - Build an L2 header for an inbound packet
  * @c:		Execution context
- * @src:	IPv6 source address, IPv4-mapped for IPv4 sources
+ * @buf:	Buffer address at which to generate header
- * @proto:	L4 protocol number
+ * @proto:	Ethernet protocol number for L3
- * @in:		Payload
+ *
- * @len:	L4 payload length
+ * Return: pointer at which to write the packet's payload
- * @flow:	Flow label for TCP over IPv6
  */
-void tap_ip_send(const struct ctx *c, const struct in6_addr *src, uint8_t proto,
+static void *tap_push_l2h(const struct ctx *c, void *buf, uint16_t proto)
-		 const char *in, size_t len, uint32_t flow)
 {
-	char buf[USHRT_MAX];
+	struct ethhdr *eh = (struct ethhdr *)buf;
-	struct ethhdr *eh;
-
-	eh = (struct ethhdr *)buf;
 
 	/* TODO: ARP table lookup */
 	memcpy(eh->h_dest, c->mac_guest, ETH_ALEN);
 	memcpy(eh->h_source, c->mac, ETH_ALEN);
+	eh->h_proto = ntohs(proto);
+	return eh + 1;
+}
 
-	if (IN6_IS_ADDR_V4MAPPED(src)) {
+/**
-		struct iphdr *iph = (struct iphdr *)(eh + 1);
+ * tap_ip4_send() - Send IPv4 packet, with L2 headers, calculating L3/L4 checksums
-		char *data = (char *)(iph + 1);
+ * @c:		Execution context
+ * @src:	IPv4 source address
+ * @proto:	L4 protocol number
+ * @in:		Payload
+ * @len:	L4 payload length
+ */
+void tap_ip4_send(const struct ctx *c, in_addr_t src, uint8_t proto,
+		  const char *in, size_t len)
+{
+	char buf[USHRT_MAX];
+	struct iphdr *ip4h = (struct iphdr *)tap_push_l2h(c, buf, ETH_P_IP);
+	char *data = (char *)(ip4h + 1);
 
-		eh->h_proto = ntohs(ETH_P_IP);
+	ip4h->version = 4;
-
+	ip4h->ihl = sizeof(struct iphdr) / 4;
-		iph->version = 4;
+	ip4h->tos = 0;
-		iph->ihl = 5;
+	ip4h->tot_len = htons(len + sizeof(*ip4h));
-		iph->tos = 0;
+	ip4h->id = 0;
-		iph->tot_len = htons(len + 20);
+	ip4h->frag_off = 0;
-		iph->id = 0;
+	ip4h->ttl = 255;
-		iph->frag_off = 0;
+	ip4h->protocol = proto;
-		iph->ttl = 255;
+	ip4h->saddr = src;
-		iph->protocol = proto;
+	ip4h->daddr = tap_ip4_daddr(c);
-		iph->daddr = tap_ip4_daddr(c);
+	csum_ip4_header(ip4h);
-		memcpy(&iph->saddr, &src->s6_addr[12], 4);
-
-		csum_ip4_header(iph);
 
 	memcpy(data, in, len);
 
-		if (iph->protocol == IPPROTO_UDP) {
+	if (ip4h->protocol == IPPROTO_UDP) {
-			struct udphdr *uh = (struct udphdr *)(iph + 1);
+		struct udphdr *uh = (struct udphdr *)(ip4h + 1);
 
-			csum_udp4(uh, iph->saddr, iph->daddr, uh + 1, len - sizeof(*uh));
+		csum_udp4(uh, ip4h->saddr, ip4h->daddr,
-		} else if (iph->protocol == IPPROTO_ICMP) {
+			  uh + 1, len - sizeof(*uh));
-			struct icmphdr *ih = (struct icmphdr *)(iph + 1);
+	} else if (ip4h->protocol == IPPROTO_ICMP) {
+		struct icmphdr *ih = (struct icmphdr *)(ip4h + 1);
 		csum_icmp4(ih, ih + 1, len - sizeof(*ih));
 	}
 
-		if (tap_send(c, buf, len + sizeof(*iph) + sizeof(*eh)) < 0)
+	if (tap_send(c, buf, len + (data - buf)) < 0)
 		debug("tap: failed to send %lu bytes (IPv4)", len);
-	} else {
+}
-		struct ipv6hdr *ip6h = (struct ipv6hdr *)(eh + 1);
+
+/**
+ * tap_ip6_send() - Send IPv6 packet, with L2 headers, calculating L3/L4 checksums
+ * @c:		Execution context
+ * @src:	IPv6 source address
+ * @proto:	L4 protocol number
+ * @in:		Payload
+ * @len:	L4 payload length
+ * @flow:	Flow label
+ */
+void tap_ip6_send(const struct ctx *c, const struct in6_addr *src,
+		  uint8_t proto, const char *in, size_t len, uint32_t flow)
+{
+	char buf[USHRT_MAX];
+	struct ipv6hdr *ip6h =
+		(struct ipv6hdr *)tap_push_l2h(c, buf, ETH_P_IPV6);
 	char *data = (char *)(ip6h + 1);
 
-		eh->h_proto = ntohs(ETH_P_IPV6);
-
-		memset(ip6h->flow_lbl, 0, 3);
 	ip6h->payload_len = htons(len);
 	ip6h->priority = 0;
-
+	ip6h->version = 6;
+	ip6h->nexthdr = proto;
+	ip6h->hop_limit = 255;
 	ip6h->saddr = *src;
 	ip6h->daddr = *tap_ip6_daddr(c, src);
+	ip6h->flow_lbl[0] = (flow >> 16) & 0xf;
+	ip6h->flow_lbl[1] = (flow >> 8) & 0xff;
+	ip6h->flow_lbl[2] = (flow >> 0) & 0xff;
 
 	memcpy(data, in, len);
 
-		ip6h->hop_limit = proto;
-		ip6h->version = 0;
-		ip6h->nexthdr = 0;
-
 	if (proto == IPPROTO_UDP) {
 		struct udphdr *uh = (struct udphdr *)(ip6h + 1);
 
@@ -195,19 +215,10 @@ void tap_ip_send(const struct ctx *c, const struct in6_addr *src, uint8_t proto,
 		csum_icmp6(ih, &ip6h->saddr, &ip6h->daddr,
 			   ih + 1, len - sizeof(*ih));
 	}
-		ip6h->version = 6;
-		ip6h->nexthdr = proto;
-		ip6h->hop_limit = 255;
-		if (flow) {
-			ip6h->flow_lbl[0] = (flow >> 16) & 0xf;
-			ip6h->flow_lbl[1] = (flow >> 8) & 0xff;
-			ip6h->flow_lbl[2] = (flow >> 0) & 0xff;
-		}
 
-		if (tap_send(c, buf, len + sizeof(*ip6h) + sizeof(*eh)) < 1)
+	if (tap_send(c, buf, len + (data - buf)) < 1)
 		debug("tap: failed to send %lu bytes (IPv6)", len);
 }
-}
 
 PACKET_POOL_DECL(pool_l4, UIO_MAXIOV, pkt_buf);
 

tap.h

@@ -9,8 +9,10 @@
 in_addr_t tap_ip4_daddr(const struct ctx *c);
 const struct in6_addr *tap_ip6_daddr(const struct ctx *c,
 				     const struct in6_addr *src);
-void tap_ip_send(const struct ctx *c, const struct in6_addr *src, uint8_t proto,
+void tap_ip4_send(const struct ctx *c, in_addr_t src, uint8_t proto,
-		 const char *in, size_t len, uint32_t flow);
+		  const char *in, size_t len);
+void tap_ip6_send(const struct ctx *c, const struct in6_addr *src,
+		  uint8_t proto, const char *in, size_t len, uint32_t flow);
 int tap_send(const struct ctx *c, const void *data, size_t len);
 void tap_handler(struct ctx *c, int fd, uint32_t events,
 		 const struct timespec *now);