passt: Introduce ICMP echo proxy

It's nice to be able to confirm connectivity using ICMP or ICMPv6
echo requests, and "ping" sockets on Linux (IPPROTO_ICMP datagram)
allow us to do that without any special capability.

Signed-off-by: Stefano Brivio <sbrivio@redhat.com>

Makefile

@@ -2,8 +2,8 @@ CFLAGS += -Wall -Wextra -pedantic
 
 all: passt qrap
 
-passt: passt.c passt.h arp.c arp.h dhcp.c dhcp.h ndp.c ndp.h siphash.c siphash.h tap.c tap.h tcp.c tcp.h udp.c udp.h util.c util.h
+passt: passt.c passt.h arp.c arp.h dhcp.c dhcp.h ndp.c ndp.h siphash.c siphash.h tap.c tap.h icmp.c icmp.h tcp.c tcp.h udp.c udp.h util.c util.h
-	$(CC) $(CFLAGS) passt.c arp.c dhcp.c ndp.c siphash.c tap.c tcp.c udp.c util.c -o passt
+	$(CC) $(CFLAGS) passt.c arp.c dhcp.c ndp.c siphash.c tap.c icmp.c tcp.c udp.c util.c -o passt
 
 qrap: qrap.c passt.h
 	$(CC) $(CFLAGS) qrap.c -o qrap

doc/demo.sh

@@ -74,6 +74,7 @@ sysctl -w net.ipv6.conf.all.forwarding=1
 
 ethtool -K veth_passt tx off
 ip netns exec passt ethtool -K veth_passt tx off
+ip netns exec passt sysctl -w net.ipv4.ping_group_range="0 2147483647"
 ulimit -n 300000
 
 

icmp.c

@@ -0,0 +1,136 @@
+// SPDX-License-Identifier: AGPL-3.0-or-later
+
+/* PASST - Plug A Simple Socket Transport
+ *
+ * icmp.c - ICMP/ICMPv6 echo proxy
+ *
+ * Copyright (c) 2021 Red Hat GmbH
+ * Author: Stefano Brivio <sbrivio@redhat.com>
+ *
+ */
+
+#include <stdio.h>
+#include <errno.h>
+#include <limits.h>
+#include <net/ethernet.h>
+#include <net/if.h>
+#include <netinet/in.h>
+#include <stdint.h>
+#include <stddef.h>
+#include <string.h>
+#include <sys/epoll.h>
+#include <sys/types.h>
+#include <sys/socket.h>
+#include <unistd.h>
+#include <linux/ip.h>
+#include <linux/ipv6.h>
+#include <linux/icmp.h>
+#include <linux/icmpv6.h>
+#include <time.h>
+
+#include "passt.h"
+#include "tap.h"
+#include "util.h"
+#include "icmp.h"
+
+/**
+ * icmp_sock_handler() - Handle new data from socket
+ * @c:		Execution context
+ * @s:		File descriptor number for socket
+ * @events:	epoll events bitmap
+ */
+void icmp_sock_handler(struct ctx *c, int s, uint32_t events)
+{
+	struct in6_addr a6 = { .s6_addr = {    0,    0,    0,    0,
+					       0,    0,    0,    0,
+					       0,    0, 0xff, 0xff,
+					       0,    0,    0,    0 } };
+	struct sockaddr_storage sr, sl;
+	socklen_t slen = sizeof(sr);
+	char buf[USHRT_MAX];
+	ssize_t n;
+
+	(void)events;
+
+	n = recvfrom(s, buf, sizeof(buf), MSG_DONTWAIT,
+		     (struct sockaddr *)&sr, &slen);
+	if (n < 0)
+		return;
+
+	if (getsockname(s, (struct sockaddr *)&sl, &slen))
+		return;
+
+	if (sl.ss_family == AF_INET) {
+		struct sockaddr_in *sr4 = (struct sockaddr_in *)&sr;
+
+		memcpy(&a6.s6_addr[12], &sr4->sin_addr, sizeof(sr4->sin_addr));
+
+		tap_ip_send(c, &a6, IPPROTO_ICMP, buf, n);
+	} else if (sl.ss_family == AF_INET6) {
+		struct sockaddr_in6 *sr6 = (struct sockaddr_in6 *)&sr;
+
+		tap_ip_send(c, &sr6->sin6_addr, IPPROTO_ICMPV6, buf, n);
+	}
+}
+
+/**
+ * icmp_tap_handler() - Handle packets from tap
+ * @c:		Execution context
+ * @af:		Address family, AF_INET or AF_INET6
+ * @in:		Input buffer
+ * @len:	Length, including UDP header
+ */
+void icmp_tap_handler(struct ctx *c, int af, void *addr, char *in, size_t len)
+{
+	if (af == AF_INET) {
+		struct icmphdr *ih = (struct icmphdr *)in;
+		struct sockaddr_in sa = {
+			.sin_family = AF_INET,
+			.sin_addr.s_addr = htonl(INADDR_ANY),
+		};
+
+		if (len < sizeof(*ih) || ih->type != ICMP_ECHO)
+			return;
+
+		sa.sin_port = ih->un.echo.id;
+		bind(c->icmp.s4, (struct sockaddr *)&sa, sizeof(sa));
+
+		sa.sin_addr = *(struct in_addr *)addr;
+		sendto(c->icmp.s4, in, len, MSG_DONTWAIT,
+		       (struct sockaddr *)&sa, sizeof(sa));
+	} else if (af == AF_INET6) {
+		struct sockaddr_in6 sa = {
+			.sin6_family = AF_INET6,
+			.sin6_addr = IN6ADDR_ANY_INIT,
+		};
+		struct icmp6hdr *ih = (struct icmp6hdr *)in;
+
+		if (len < sizeof(*ih) ||
+		    (ih->icmp6_type != 128 && ih->icmp6_type != 129))
+			return;
+
+		sa.sin6_port = ih->icmp6_identifier;
+		bind(c->icmp.s6, (struct sockaddr *)&sa, sizeof(sa));
+
+		sa.sin6_addr = *(struct in6_addr *)addr;
+		sendto(c->icmp.s6, in, len, MSG_DONTWAIT | MSG_NOSIGNAL,
+		       (struct sockaddr *)&sa, sizeof(sa));
+	}
+}
+
+/**
+ * icmp_sock_init() - Create ICMP, ICMPv6 sockets for echo requests and replies
+ * @c:		Execution context
+ *
+ * Return: 0 on success, -1 on failure
+ */
+int icmp_sock_init(struct ctx *c)
+{
+	if (c->v4 && (c->icmp.s4 = sock_l4_add(c, 4, IPPROTO_ICMP, 0)) < 0)
+		return -1;
+
+	if (c->v6 && (c->icmp.s6 = sock_l4_add(c, 6, IPPROTO_ICMPV6, 0)) < 0)
+		return -1;
+
+	return 0;
+}

icmp.h

@@ -0,0 +1,20 @@
+#ifndef ICMP_H
+#define ICMP_H
+
+struct ctx;
+
+void icmp_sock_handler(struct ctx *c, int s, uint32_t events);
+void icmp_tap_handler(struct ctx *c, int af, void *addr, char *in, size_t len);
+int icmp_sock_init(struct ctx *c);
+
+/**
+ * struct icmp_ctx - Execution context for ICMP routines
+ * @s4:		ICMP socket number
+ * @s6:		ICMPv6 socket number
+ */
+struct icmp_ctx {
+	int s4;
+	int s6;
+};
+
+#endif /* ICMP_H */

passt.c

@@ -46,6 +46,7 @@
 #include "dhcp.h"
 #include "ndp.h"
 #include "util.h"
+#include "icmp.h"
 #include "tcp.h"
 #include "udp.h"
 
@@ -335,6 +336,8 @@ static void tap4_handler(struct ctx *c, char *in, size_t len)
 		tcp_tap_handler(c, AF_INET, &iph->daddr, l4h, len);
 	else if (iph->protocol == IPPROTO_UDP)
 		udp_tap_handler(c, AF_INET, &iph->daddr, l4h, len);
+	else if (iph->protocol == IPPROTO_ICMP)
+		icmp_tap_handler(c, AF_INET, &iph->daddr, l4h, len);
 }
 
 /**
@@ -391,6 +394,8 @@ static void tap6_handler(struct ctx *c, char *in, size_t len)
 		tcp_tap_handler(c, AF_INET6, &ip6h->daddr, l4h, len);
 	else if (proto == IPPROTO_UDP)
 		udp_tap_handler(c, AF_INET6, &ip6h->daddr, l4h, len);
+	else if (proto == IPPROTO_ICMPV6)
+		icmp_tap_handler(c, AF_INET6, &ip6h->daddr, l4h, len);
 }
 
 /**
@@ -449,15 +454,18 @@ static void sock_handler(struct ctx *c, int fd, uint32_t events)
 
 	sl = sizeof(so);
 
-	if (getsockopt(fd, SOL_SOCKET, SO_TYPE, &so, &sl) ||
+	if (getsockopt(fd, SOL_SOCKET, SO_PROTOCOL, &so, &sl))
-	    so == SOCK_STREAM) {
+		return;
-		fprintf(stderr, "TCP: packet from socket %i\n", fd);
+
+	fprintf(stderr, "%s: packet from socket %i\n",
+		getprotobynumber(so)->p_name, fd);
+
+	if (so == IPPROTO_ICMP || so == IPPROTO_ICMPV6)
+		icmp_sock_handler(c, fd, events);
+	else if (so == IPPROTO_TCP)
 		tcp_sock_handler(c, fd, events);
-	}
+	else if (so == IPPROTO_UDP)
-	else if (so == SOCK_DGRAM) {
 		udp_sock_handler(c, fd, events);
-		fprintf(stderr, "UDP: packet from socket %i\n", fd);
-	}
 }
 
 /**
@@ -546,7 +554,7 @@ int main(int argc, char **argv)
 		exit(EXIT_FAILURE);
 	}
 
-	if (tcp_sock_init(&c) || udp_sock_init(&c))
+	if (icmp_sock_init(&c) || tcp_sock_init(&c) || udp_sock_init(&c))
 		exit(EXIT_FAILURE);
 
 	fd_unix = sock_unix();

passt.h

@@ -1,5 +1,6 @@
 #define UNIX_SOCK_PATH	"/tmp/passt.socket"
 
+#include "icmp.h"
 #include "tcp.h"
 
 /**
@@ -39,5 +40,6 @@ struct ctx {
 
 	char ifn[IF_NAMESIZE];
 
+	struct icmp_ctx icmp;
 	struct tcp_ctx tcp;
 };

util.c

@@ -160,7 +160,8 @@ int sock_l4_add(struct ctx *c, int v, uint16_t proto, uint16_t port)
 	const struct sockaddr *sa;
 	int fd, sl, one = 1;
 
-	if (proto != IPPROTO_TCP && proto != IPPROTO_UDP)
+	if (proto != IPPROTO_TCP && proto != IPPROTO_UDP &&
+	    proto != IPPROTO_ICMP && proto != IPPROTO_ICMPV6)
 		return -1;	/* Not implemented. */
 
 	fd = socket(v == 4 ? AF_INET : AF_INET6,
@@ -170,6 +171,9 @@ int sock_l4_add(struct ctx *c, int v, uint16_t proto, uint16_t port)
 		return -1;
 	}
 
+	if (proto == IPPROTO_ICMP || proto == IPPROTO_ICMPV6)
+		goto epoll_add;
+
 	if (v == 4) {
 		sa = (const struct sockaddr *)&addr4;
 		sl = sizeof(addr4);
@@ -195,6 +199,7 @@ int sock_l4_add(struct ctx *c, int v, uint16_t proto, uint16_t port)
 		return -1;
 	}
 
+epoll_add:
 	ev.events = EPOLLIN;
 	ev.data.fd = fd;
 	if (epoll_ctl(c->epollfd, EPOLL_CTL_ADD, fd, &ev) == -1) {