2021-03-17 10:57:44 +01:00
|
|
|
// 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"
|
|
|
|
|
2021-05-21 11:14:50 +02:00
|
|
|
/* Indexed by ICMP echo identifier */
|
|
|
|
static int icmp_s_v4[USHRT_MAX];
|
|
|
|
static int icmp_s_v6[USHRT_MAX];
|
|
|
|
|
2021-03-17 10:57:44 +01:00
|
|
|
/**
|
|
|
|
* icmp_sock_handler() - Handle new data from socket
|
|
|
|
* @c: Execution context
|
|
|
|
* @s: File descriptor number for socket
|
|
|
|
* @events: epoll events bitmap
|
2021-04-30 14:52:18 +02:00
|
|
|
* @pkt_buf: Buffer to receive packets, currently unused
|
udp: Connection tracking for ephemeral, local ports, and related fixes
As we support UDP forwarding for packets that are sent to local
ports, we actually need some kind of connection tracking for UDP.
While at it, this commit introduces a number of vaguely related fixes
for issues observed while trying this out. In detail:
- implement an explicit, albeit minimalistic, connection tracking
for UDP, to allow usage of ephemeral ports by the guest and by
the host at the same time, by binding them dynamically as needed,
and to allow mapping address changes for packets with a loopback
address as destination
- set the guest MAC address whenever we receive a packet from tap
instead of waiting for an ARP request, and set it to broadcast on
start, otherwise DHCPv6 might not work if all DHCPv6 requests time
out before the guest starts talking IPv4
- split context IPv6 address into address we assign, global or site
address seen on tap, and link-local address seen on tap, and make
sure we use the addresses we've seen as destination (link-local
choice depends on source address). Similarly, for IPv4, split into
address we assign and address we observe, and use the address we
observe as destination
- introduce a clock_gettime() syscall right after epoll_wait() wakes
up, so that we can remove all the other ones and pass the current
timestamp to tap and socket handlers -- this is additionally needed
by UDP to time out bindings to ephemeral ports and mappings between
loopback address and a local address
- rename sock_l4_add() to sock_l4(), no semantic changes intended
- include <arpa/inet.h> in passt.c before kernel headers so that we
can use <netinet/in.h> macros to check IPv6 address types, and
remove a duplicate <linux/ip.h> inclusion
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
2021-04-29 16:59:20 +02:00
|
|
|
* @now: Current timestamp, unused
|
2021-03-17 10:57:44 +01:00
|
|
|
*/
|
2021-04-30 14:52:18 +02:00
|
|
|
void icmp_sock_handler(struct ctx *c, int s, uint32_t events, char *pkt_buf,
|
udp: Connection tracking for ephemeral, local ports, and related fixes
As we support UDP forwarding for packets that are sent to local
ports, we actually need some kind of connection tracking for UDP.
While at it, this commit introduces a number of vaguely related fixes
for issues observed while trying this out. In detail:
- implement an explicit, albeit minimalistic, connection tracking
for UDP, to allow usage of ephemeral ports by the guest and by
the host at the same time, by binding them dynamically as needed,
and to allow mapping address changes for packets with a loopback
address as destination
- set the guest MAC address whenever we receive a packet from tap
instead of waiting for an ARP request, and set it to broadcast on
start, otherwise DHCPv6 might not work if all DHCPv6 requests time
out before the guest starts talking IPv4
- split context IPv6 address into address we assign, global or site
address seen on tap, and link-local address seen on tap, and make
sure we use the addresses we've seen as destination (link-local
choice depends on source address). Similarly, for IPv4, split into
address we assign and address we observe, and use the address we
observe as destination
- introduce a clock_gettime() syscall right after epoll_wait() wakes
up, so that we can remove all the other ones and pass the current
timestamp to tap and socket handlers -- this is additionally needed
by UDP to time out bindings to ephemeral ports and mappings between
loopback address and a local address
- rename sock_l4_add() to sock_l4(), no semantic changes intended
- include <arpa/inet.h> in passt.c before kernel headers so that we
can use <netinet/in.h> macros to check IPv6 address types, and
remove a duplicate <linux/ip.h> inclusion
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
2021-04-29 16:59:20 +02:00
|
|
|
struct timespec *now)
|
2021-03-17 10:57:44 +01:00
|
|
|
{
|
|
|
|
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;
|
2021-04-30 14:52:18 +02:00
|
|
|
(void)pkt_buf;
|
udp: Connection tracking for ephemeral, local ports, and related fixes
As we support UDP forwarding for packets that are sent to local
ports, we actually need some kind of connection tracking for UDP.
While at it, this commit introduces a number of vaguely related fixes
for issues observed while trying this out. In detail:
- implement an explicit, albeit minimalistic, connection tracking
for UDP, to allow usage of ephemeral ports by the guest and by
the host at the same time, by binding them dynamically as needed,
and to allow mapping address changes for packets with a loopback
address as destination
- set the guest MAC address whenever we receive a packet from tap
instead of waiting for an ARP request, and set it to broadcast on
start, otherwise DHCPv6 might not work if all DHCPv6 requests time
out before the guest starts talking IPv4
- split context IPv6 address into address we assign, global or site
address seen on tap, and link-local address seen on tap, and make
sure we use the addresses we've seen as destination (link-local
choice depends on source address). Similarly, for IPv4, split into
address we assign and address we observe, and use the address we
observe as destination
- introduce a clock_gettime() syscall right after epoll_wait() wakes
up, so that we can remove all the other ones and pass the current
timestamp to tap and socket handlers -- this is additionally needed
by UDP to time out bindings to ephemeral ports and mappings between
loopback address and a local address
- rename sock_l4_add() to sock_l4(), no semantic changes intended
- include <arpa/inet.h> in passt.c before kernel headers so that we
can use <netinet/in.h> macros to check IPv6 address types, and
remove a duplicate <linux/ip.h> inclusion
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
2021-04-29 16:59:20 +02:00
|
|
|
(void)now;
|
2021-03-17 10:57:44 +01:00
|
|
|
|
|
|
|
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
|
2021-04-22 13:39:36 +02:00
|
|
|
* @msg: Input message
|
|
|
|
* @count: Message count (always 1 for ICMP)
|
udp: Connection tracking for ephemeral, local ports, and related fixes
As we support UDP forwarding for packets that are sent to local
ports, we actually need some kind of connection tracking for UDP.
While at it, this commit introduces a number of vaguely related fixes
for issues observed while trying this out. In detail:
- implement an explicit, albeit minimalistic, connection tracking
for UDP, to allow usage of ephemeral ports by the guest and by
the host at the same time, by binding them dynamically as needed,
and to allow mapping address changes for packets with a loopback
address as destination
- set the guest MAC address whenever we receive a packet from tap
instead of waiting for an ARP request, and set it to broadcast on
start, otherwise DHCPv6 might not work if all DHCPv6 requests time
out before the guest starts talking IPv4
- split context IPv6 address into address we assign, global or site
address seen on tap, and link-local address seen on tap, and make
sure we use the addresses we've seen as destination (link-local
choice depends on source address). Similarly, for IPv4, split into
address we assign and address we observe, and use the address we
observe as destination
- introduce a clock_gettime() syscall right after epoll_wait() wakes
up, so that we can remove all the other ones and pass the current
timestamp to tap and socket handlers -- this is additionally needed
by UDP to time out bindings to ephemeral ports and mappings between
loopback address and a local address
- rename sock_l4_add() to sock_l4(), no semantic changes intended
- include <arpa/inet.h> in passt.c before kernel headers so that we
can use <netinet/in.h> macros to check IPv6 address types, and
remove a duplicate <linux/ip.h> inclusion
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
2021-04-29 16:59:20 +02:00
|
|
|
* @now: Current timestamp, unused
|
2021-04-22 13:39:36 +02:00
|
|
|
*
|
|
|
|
* Return: count of consumed packets (always 1, even if malformed)
|
2021-03-17 10:57:44 +01:00
|
|
|
*/
|
2021-04-22 13:39:36 +02:00
|
|
|
int icmp_tap_handler(struct ctx *c, int af, void *addr,
|
udp: Connection tracking for ephemeral, local ports, and related fixes
As we support UDP forwarding for packets that are sent to local
ports, we actually need some kind of connection tracking for UDP.
While at it, this commit introduces a number of vaguely related fixes
for issues observed while trying this out. In detail:
- implement an explicit, albeit minimalistic, connection tracking
for UDP, to allow usage of ephemeral ports by the guest and by
the host at the same time, by binding them dynamically as needed,
and to allow mapping address changes for packets with a loopback
address as destination
- set the guest MAC address whenever we receive a packet from tap
instead of waiting for an ARP request, and set it to broadcast on
start, otherwise DHCPv6 might not work if all DHCPv6 requests time
out before the guest starts talking IPv4
- split context IPv6 address into address we assign, global or site
address seen on tap, and link-local address seen on tap, and make
sure we use the addresses we've seen as destination (link-local
choice depends on source address). Similarly, for IPv4, split into
address we assign and address we observe, and use the address we
observe as destination
- introduce a clock_gettime() syscall right after epoll_wait() wakes
up, so that we can remove all the other ones and pass the current
timestamp to tap and socket handlers -- this is additionally needed
by UDP to time out bindings to ephemeral ports and mappings between
loopback address and a local address
- rename sock_l4_add() to sock_l4(), no semantic changes intended
- include <arpa/inet.h> in passt.c before kernel headers so that we
can use <netinet/in.h> macros to check IPv6 address types, and
remove a duplicate <linux/ip.h> inclusion
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
2021-04-29 16:59:20 +02:00
|
|
|
struct tap_msg *msg, int count, struct timespec *now)
|
2021-03-17 10:57:44 +01:00
|
|
|
{
|
2021-05-21 11:14:50 +02:00
|
|
|
int s;
|
|
|
|
|
2021-04-22 13:39:36 +02:00
|
|
|
(void)count;
|
udp: Connection tracking for ephemeral, local ports, and related fixes
As we support UDP forwarding for packets that are sent to local
ports, we actually need some kind of connection tracking for UDP.
While at it, this commit introduces a number of vaguely related fixes
for issues observed while trying this out. In detail:
- implement an explicit, albeit minimalistic, connection tracking
for UDP, to allow usage of ephemeral ports by the guest and by
the host at the same time, by binding them dynamically as needed,
and to allow mapping address changes for packets with a loopback
address as destination
- set the guest MAC address whenever we receive a packet from tap
instead of waiting for an ARP request, and set it to broadcast on
start, otherwise DHCPv6 might not work if all DHCPv6 requests time
out before the guest starts talking IPv4
- split context IPv6 address into address we assign, global or site
address seen on tap, and link-local address seen on tap, and make
sure we use the addresses we've seen as destination (link-local
choice depends on source address). Similarly, for IPv4, split into
address we assign and address we observe, and use the address we
observe as destination
- introduce a clock_gettime() syscall right after epoll_wait() wakes
up, so that we can remove all the other ones and pass the current
timestamp to tap and socket handlers -- this is additionally needed
by UDP to time out bindings to ephemeral ports and mappings between
loopback address and a local address
- rename sock_l4_add() to sock_l4(), no semantic changes intended
- include <arpa/inet.h> in passt.c before kernel headers so that we
can use <netinet/in.h> macros to check IPv6 address types, and
remove a duplicate <linux/ip.h> inclusion
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
2021-04-29 16:59:20 +02:00
|
|
|
(void)now;
|
2021-05-21 11:14:50 +02:00
|
|
|
(void)c;
|
2021-04-22 13:39:36 +02:00
|
|
|
|
2021-03-17 10:57:44 +01:00
|
|
|
if (af == AF_INET) {
|
2021-04-22 13:39:36 +02:00
|
|
|
struct icmphdr *ih = (struct icmphdr *)msg[0].l4h;
|
2021-03-17 10:57:44 +01:00
|
|
|
struct sockaddr_in sa = {
|
|
|
|
.sin_family = AF_INET,
|
2021-05-21 11:14:53 +02:00
|
|
|
.sin_addr = { .s_addr = INADDR_ANY },
|
|
|
|
.sin_port = ih->un.echo.id,
|
2021-03-17 10:57:44 +01:00
|
|
|
};
|
|
|
|
|
2021-04-22 13:39:36 +02:00
|
|
|
if (msg[0].l4_len < sizeof(*ih) || ih->type != ICMP_ECHO)
|
|
|
|
return 1;
|
2021-03-17 10:57:44 +01:00
|
|
|
|
2021-05-21 11:14:50 +02:00
|
|
|
if ((s = icmp_s_v4[ntohs(ih->un.echo.id)]) < 0)
|
|
|
|
return 1;
|
2021-03-17 10:57:44 +01:00
|
|
|
|
2021-05-21 11:14:53 +02:00
|
|
|
bind(s, (struct sockaddr *)&sa, sizeof(sa));
|
|
|
|
|
|
|
|
sa.sin_addr = *(struct in_addr *)addr;
|
2021-05-21 11:14:50 +02:00
|
|
|
sendto(s, msg[0].l4h, msg[0].l4_len,
|
2021-04-22 13:39:36 +02:00
|
|
|
MSG_DONTWAIT | MSG_NOSIGNAL,
|
2021-03-17 10:57:44 +01:00
|
|
|
(struct sockaddr *)&sa, sizeof(sa));
|
|
|
|
} else if (af == AF_INET6) {
|
2021-05-21 11:14:50 +02:00
|
|
|
struct icmp6hdr *ih = (struct icmp6hdr *)msg[0].l4h;
|
2021-03-17 10:57:44 +01:00
|
|
|
struct sockaddr_in6 sa = {
|
|
|
|
.sin6_family = AF_INET6,
|
2021-05-21 11:14:53 +02:00
|
|
|
.sin6_addr = IN6ADDR_ANY_INIT,
|
|
|
|
.sin6_port = ih->icmp6_identifier,
|
2021-03-17 10:57:44 +01:00
|
|
|
};
|
2021-05-10 08:30:38 +02:00
|
|
|
|
2021-04-22 13:39:36 +02:00
|
|
|
if (msg[0].l4_len < sizeof(*ih) ||
|
2021-03-17 10:57:44 +01:00
|
|
|
(ih->icmp6_type != 128 && ih->icmp6_type != 129))
|
2021-04-22 13:39:36 +02:00
|
|
|
return 1;
|
2021-03-17 10:57:44 +01:00
|
|
|
|
2021-05-21 11:14:50 +02:00
|
|
|
if ((s = icmp_s_v6[ntohs(ih->icmp6_identifier)]) < 0)
|
|
|
|
return 1;
|
2021-03-17 10:57:44 +01:00
|
|
|
|
2021-05-21 11:14:53 +02:00
|
|
|
bind(s, (struct sockaddr *)&sa, sizeof(sa));
|
|
|
|
|
|
|
|
sa.sin6_addr = *(struct in6_addr *)addr;
|
2021-05-21 11:14:50 +02:00
|
|
|
sendto(s, msg[0].l4h, msg[0].l4_len,
|
2021-04-22 13:39:36 +02:00
|
|
|
MSG_DONTWAIT | MSG_NOSIGNAL,
|
2021-03-17 10:57:44 +01:00
|
|
|
(struct sockaddr *)&sa, sizeof(sa));
|
|
|
|
}
|
2021-04-22 13:39:36 +02:00
|
|
|
|
|
|
|
return 1;
|
2021-03-17 10:57:44 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* 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)
|
|
|
|
{
|
2021-05-21 11:14:50 +02:00
|
|
|
int i, fail = 0;
|
2021-05-10 08:30:38 +02:00
|
|
|
|
passt: Spare some syscalls, add some optimisations from profiling
Avoid a bunch of syscalls on forwarding paths by:
- storing minimum and maximum file descriptor numbers for each
protocol, fall back to SO_PROTOCOL query only on overlaps
- allocating a larger receive buffer -- this can result in more
coalesced packets than sendmmsg() can take (UIO_MAXIOV, i.e. 1024),
so make sure we don't exceed that within a single call to protocol
tap handlers
- nesting the handling loop in tap_handler() in the receive loop,
so that we have better chances of filling our receive buffer in
fewer calls
- skipping the recvfrom() in the UDP handler on EPOLLERR -- there's
nothing to be done in that case
and while at it:
- restore the 20ms timer interval for periodic (TCP) events, I
accidentally changed that to 100ms in an earlier commit
- attempt using SO_ZEROCOPY for UDP -- if it's not available,
sendmmsg() will succeed anyway
- fix the handling of the status code from sendmmsg(), if it fails,
we'll try to discard the first message, hence return 1 from the
UDP handler
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
2021-04-23 22:22:37 +02:00
|
|
|
c->icmp.fd_min = INT_MAX;
|
|
|
|
c->icmp.fd_max = 0;
|
|
|
|
|
2021-05-21 11:14:50 +02:00
|
|
|
if (c->v4) {
|
|
|
|
for (i = 0; i < USHRT_MAX; i++) {
|
|
|
|
icmp_s_v4[i] = sock_l4(c, AF_INET, IPPROTO_ICMP, i);
|
|
|
|
if (icmp_s_v4[i] < 0)
|
|
|
|
fail = 1;
|
|
|
|
}
|
|
|
|
}
|
2021-03-17 10:57:44 +01:00
|
|
|
|
2021-05-21 11:14:50 +02:00
|
|
|
if (c->v6) {
|
|
|
|
for (i = 0; i < USHRT_MAX; i++) {
|
|
|
|
icmp_s_v6[i] = sock_l4(c, AF_INET6, IPPROTO_ICMPV6, i);
|
|
|
|
if (icmp_s_v6[i] < 0)
|
|
|
|
fail = 1;
|
|
|
|
}
|
|
|
|
}
|
2021-05-10 08:30:38 +02:00
|
|
|
|
|
|
|
if (fail) {
|
2021-05-21 11:14:50 +02:00
|
|
|
warn("Cannot open some \"ping\" sockets. You might need to:");
|
2021-05-10 08:30:38 +02:00
|
|
|
warn(" sysctl -w net.ipv4.ping_group_range=\"0 2147483647\"");
|
2021-05-21 11:14:50 +02:00
|
|
|
warn("...echo requests/replies might fail.");
|
2021-05-10 08:30:38 +02:00
|
|
|
}
|
2021-03-17 10:57:44 +01:00
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|