busybox/networking/udhcp/packet.c
Denys Vlasenko 8f0e4c42c6 udhcpd: optional BOOTP support
from Adam Goldman <adamg@pobox.com>

This patch makes udhcpd respond correctly to queries from BOOTP clients.

It contains the following changes:

The end field, or DHCP_END option, is required in DHCP requests but
optional in BOOTP requests. However, we still send an end
field in all replies, because some BOOTP clients expect one in replies
even if they didn't send one in the request.

Requests without a DHCP_MESSAGE_TYPE are recognized as BOOTP requests
and handled appropriately, instead of being discarded. We still require
an RFC 1048 options field, but we allow it to be empty.

Since a BOOTP client will keep using the assigned IP forever, we only
send a BOOTP reply if a static lease exists for that client.

BOOTP replies shouldn't contain DHCP_* options, so we omit them if there
was no DHCP_MESSAGE_TYPE in the request. Options other than DHCP_*
options are still sent.

The options field of a BOOTP reply must be exactly 64 bytes. If we
construct a reply with more than 64 bytes of options, we give up and log
an error instead of sending it. udhcp_send_raw_packet already pads the
options field to 64 bytes if it is too short.

This implementation has been tested against an HP PA-RISC client.

function                                             old     new   delta
.rodata                                           105247  105321     +74
udhcpd_main                                         1520    1591     +71
send_offer                                           419     470     +51
init_packet                                           81      97     +16
udhcp_init_header                                     75      88     +13
udhcp_scan_options                                   192     203     +11
------------------------------------------------------------------------------
(add/remove: 0/0 grow/shrink: 6/0 up/down: 236/0)             Total: 236 bytes

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2023-06-12 13:22:47 +02:00

259 lines
7 KiB
C

/* vi: set sw=4 ts=4: */
/*
* Packet ops
*
* Rewrite by Russ Dill <Russ.Dill@asu.edu> July 2001
*
* Licensed under GPLv2, see file LICENSE in this source tree.
*/
#include "common.h"
#include "dhcpd.h"
#include <netinet/in.h>
#include <netinet/if_ether.h>
#include <netpacket/packet.h>
#if ENABLE_UDHCPC || ENABLE_UDHCPD
void FAST_FUNC udhcp_init_header(struct dhcp_packet *packet, char type)
{
memset(packet, 0, sizeof(*packet));
packet->op = BOOTREQUEST; /* if client to a server */
switch (type) {
IF_FEATURE_UDHCPD_BOOTP(case MSGTYPE_BOOTP:)
/* reply to a BOOTP (not DHCP) client */
case DHCPOFFER:
case DHCPACK:
case DHCPNAK:
packet->op = BOOTREPLY; /* if server to client */
}
packet->htype = 1; /* ethernet */
packet->hlen = 6;
packet->cookie = htonl(RFC1048_MAGIC);
if (DHCP_END != 0)
packet->options[0] = DHCP_END;
IF_FEATURE_UDHCPD_BOOTP(if (type != MSGTYPE_BOOTP))
udhcp_add_simple_option(packet, DHCP_MESSAGE_TYPE, type);
}
#endif
#if defined CONFIG_UDHCP_DEBUG && CONFIG_UDHCP_DEBUG >= 2
void FAST_FUNC udhcp_dump_packet(struct dhcp_packet *packet)
{
char buf[sizeof(packet->chaddr)*2 + 1];
if (dhcp_verbose < 2)
return;
bb_info_msg(
//" op %x"
//" htype %x"
" hlen %x"
//" hops %x"
" xid %x"
//" secs %x"
//" flags %x"
" ciaddr %x"
" yiaddr %x"
" siaddr %x"
" giaddr %x"
//" sname %s"
//" file %s"
//" cookie %x"
//" options %s"
//, packet->op
//, packet->htype
, packet->hlen
//, packet->hops
, packet->xid
//, packet->secs
//, packet->flags
, packet->ciaddr
, packet->yiaddr
, packet->siaddr_nip
, packet->gateway_nip
//, packet->sname[64]
//, packet->file[128]
//, packet->cookie
//, packet->options[]
);
*bin2hex(buf, (void *) packet->chaddr, sizeof(packet->chaddr)) = '\0';
bb_info_msg(" chaddr %s", buf);
}
#endif
/* Read a packet from socket fd, return -1 on read error, -2 on packet error */
int FAST_FUNC udhcp_recv_kernel_packet(struct dhcp_packet *packet, int fd)
{
int bytes;
memset(packet, 0, sizeof(*packet));
bytes = safe_read(fd, packet, sizeof(*packet));
if (bytes < 0) {
log1s("packet read error, ignoring");
return bytes; /* returns -1 */
}
if (bytes < offsetof(struct dhcp_packet, options)
|| packet->cookie != htonl(RFC1048_MAGIC)
) {
bb_simple_info_msg("packet with bad magic, ignoring");
return -2;
}
log2("received %s", "a packet");
/* log2 because more informative msg for valid packets is printed later at log1 level */
udhcp_dump_packet(packet);
return bytes;
}
/* Construct a ip/udp header for a packet, send packet */
int FAST_FUNC udhcp_send_raw_packet(struct dhcp_packet *dhcp_pkt,
uint32_t source_nip, int source_port,
uint32_t dest_nip, int dest_port, const uint8_t *dest_arp,
int ifindex)
{
struct sockaddr_ll dest_sll;
struct ip_udp_dhcp_packet packet;
unsigned padding;
int fd;
int result = -1;
const char *msg;
fd = socket(PF_PACKET, SOCK_DGRAM, htons(ETH_P_IP));
if (fd < 0) {
msg = "socket(%s)";
goto ret_msg;
}
memset(&dest_sll, 0, sizeof(dest_sll));
memset(&packet, 0, offsetof(struct ip_udp_dhcp_packet, data));
packet.data = *dhcp_pkt; /* struct copy */
dest_sll.sll_family = AF_PACKET;
dest_sll.sll_protocol = htons(ETH_P_IP);
dest_sll.sll_ifindex = ifindex;
/*dest_sll.sll_hatype = ARPHRD_???;*/
/*dest_sll.sll_pkttype = PACKET_???;*/
dest_sll.sll_halen = 6;
memcpy(dest_sll.sll_addr, dest_arp, 6);
//TODO: is bind() necessary? we sendto() to this destination, should work anyway
if (bind(fd, (struct sockaddr *)&dest_sll, sizeof(dest_sll)) < 0) {
msg = "bind(%s)";
goto ret_close;
}
/* We were sending full-sized DHCP packets (zero padded),
* but some badly configured servers were seen dropping them.
* Apparently they drop all DHCP packets >576 *ethernet* octets big,
* whereas they may only drop packets >576 *IP* octets big
* (which for typical Ethernet II means 590 octets: 6+6+2 + 576).
*
* In order to work with those buggy servers,
* we truncate packets after end option byte.
*
* However, RFC 1542 says "The IP Total Length and UDP Length
* must be large enough to contain the minimal BOOTP header of 300 octets".
* Thus, we retain enough padding to not go below 300 BOOTP bytes.
* Some devices have filters which drop DHCP packets shorter than that.
*/
padding = DHCP_OPTIONS_BUFSIZE - 1 - udhcp_end_option(packet.data.options);
if (padding > DHCP_SIZE - 300)
padding = DHCP_SIZE - 300;
packet.ip.protocol = IPPROTO_UDP;
packet.ip.saddr = source_nip;
packet.ip.daddr = dest_nip;
packet.udp.source = htons(source_port);
packet.udp.dest = htons(dest_port);
/* size, excluding IP header: */
packet.udp.len = htons(UDP_DHCP_SIZE - padding);
/* for UDP checksumming, ip.len is set to UDP packet len */
packet.ip.tot_len = packet.udp.len;
packet.udp.check = inet_cksum(&packet,
IP_UDP_DHCP_SIZE - padding);
/* but for sending, it is set to IP packet len */
packet.ip.tot_len = htons(IP_UDP_DHCP_SIZE - padding);
packet.ip.ihl = sizeof(packet.ip) >> 2;
packet.ip.version = IPVERSION;
packet.ip.ttl = IPDEFTTL;
packet.ip.check = inet_cksum(&packet.ip, sizeof(packet.ip));
udhcp_dump_packet(dhcp_pkt);
result = sendto(fd, &packet, IP_UDP_DHCP_SIZE - padding, /*flags:*/ 0,
(struct sockaddr *) &dest_sll, sizeof(dest_sll));
msg = "sendto";
ret_close:
close(fd);
if (result < 0) {
ret_msg:
bb_perror_msg(msg, "PACKET");
}
return result;
}
/* Let the kernel do all the work for packet generation */
int FAST_FUNC udhcp_send_kernel_packet(struct dhcp_packet *dhcp_pkt,
uint32_t source_nip, int source_port,
uint32_t dest_nip, int dest_port,
const char *ifname)
{
struct sockaddr_in sa;
unsigned padding;
int fd;
int result = -1;
const char *msg;
fd = socket(PF_INET, SOCK_DGRAM, IPPROTO_UDP);
if (fd < 0) {
msg = "socket(%s)";
goto ret_msg;
}
setsockopt_reuseaddr(fd);
/* If interface carrier goes down, unless we
* bind socket to a particular netdev, the packet
* can go out through another interface, eg. via
* default route despite being bound to a specific
* source IP. As such, bind to device hard and fail
* otherwise. Sending renewal packets on foreign
* interfaces makes no sense.
*/
if (ifname) {
if (setsockopt_bindtodevice(fd, ifname) < 0) {
msg = "bindtodevice";
goto ret_close;
}
}
memset(&sa, 0, sizeof(sa));
sa.sin_family = AF_INET;
sa.sin_port = htons(source_port);
sa.sin_addr.s_addr = source_nip;
if (bind(fd, (struct sockaddr *)&sa, sizeof(sa)) == -1) {
msg = "bind(%s)";
goto ret_close;
}
memset(&sa, 0, sizeof(sa));
sa.sin_family = AF_INET;
sa.sin_port = htons(dest_port);
sa.sin_addr.s_addr = dest_nip;
if (connect(fd, (struct sockaddr *)&sa, sizeof(sa)) == -1) {
msg = "connect";
goto ret_close;
}
udhcp_dump_packet(dhcp_pkt);
padding = DHCP_OPTIONS_BUFSIZE - 1 - udhcp_end_option(dhcp_pkt->options);
if (padding > DHCP_SIZE - 300)
padding = DHCP_SIZE - 300;
result = safe_write(fd, dhcp_pkt, DHCP_SIZE - padding);
msg = "write";
ret_close:
close(fd);
if (result < 0) {
ret_msg:
bb_perror_msg(msg, "UDP");
}
return result;
}