static int
libcfs_sock_create (struct socket **sockp, int *fatal,
__u32 local_ip, int local_port)
{
struct sockaddr_in locaddr;
struct socket *so;
struct sockopt sopt;
int option;
int rc;
CFS_DECL_FUNNEL_DATA;
*fatal = 1;
CFS_NET_IN;
rc = socreate(PF_INET, &so, SOCK_STREAM, 0);
CFS_NET_EX;
if (rc != 0) {
CERROR ("Can't create socket: %d\n", rc);
return (-rc);
}
bzero(&sopt, sizeof sopt);
option = 1;
sopt.sopt_level = SOL_SOCKET;
sopt.sopt_name = SO_REUSEADDR;
sopt.sopt_val = &option;
sopt.sopt_valsize = sizeof(option);
CFS_NET_IN;
rc = sosetopt(so, &sopt);
if (rc != 0) {
CFS_NET_EX;
CERROR ("Can't set sock reuse address: %d\n", rc);
goto out;
}
/* can't specify a local port without a local IP */
LASSERT (local_ip == 0 || local_port != 0);
if (local_ip != 0 || local_port != 0) {
bzero (&locaddr, sizeof (locaddr));
locaddr.sin_len = sizeof(struct sockaddr_in);
locaddr.sin_family = AF_INET;
locaddr.sin_port = htons (local_port);
locaddr.sin_addr.s_addr = (local_ip != 0) ? htonl(local_ip) :
INADDR_ANY;
rc = sobind(so, (struct sockaddr *)&locaddr);
if (rc == EADDRINUSE) {
CFS_NET_EX;
CDEBUG(D_NET, "Port %d already in use\n", local_port);
*fatal = 0;
goto out;
}
if (rc != 0) {
CFS_NET_EX;
CERROR ("Can't bind to local IP Address %u.%u.%u.%u: %d\n",
HIPQUAD(local_ip), rc);
goto out;
}
}
*sockp = so;
return 0;
out:
CFS_NET_IN;
soclose(so);
CFS_NET_EX;
return -rc;
}
static int help(struct sk_buff **pskb,
struct ip_conntrack *ct, enum ip_conntrack_info ctinfo)
{
unsigned int dataoff;
struct tcphdr _tcph, *th;
char *data, *data_limit, *ib_ptr;
int dir = CTINFO2DIR(ctinfo);
struct ip_conntrack_expect *exp;
u32 seq;
u_int32_t dcc_ip;
u_int16_t dcc_port;
int i, ret = NF_ACCEPT;
char *addr_beg_p, *addr_end_p;
DEBUGP("entered\n");
/* If packet is coming from IRC server */
if (dir == IP_CT_DIR_REPLY)
return NF_ACCEPT;
/* Until there's been traffic both ways, don't look in packets. */
if (ctinfo != IP_CT_ESTABLISHED
&& ctinfo != IP_CT_ESTABLISHED + IP_CT_IS_REPLY) {
DEBUGP("Conntrackinfo = %u\n", ctinfo);
return NF_ACCEPT;
}
/* Not a full tcp header? */
th = skb_header_pointer(*pskb, (*pskb)->nh.iph->ihl*4,
sizeof(_tcph), &_tcph);
if (th == NULL)
return NF_ACCEPT;
/* No data? */
dataoff = (*pskb)->nh.iph->ihl*4 + th->doff*4;
if (dataoff >= (*pskb)->len)
return NF_ACCEPT;
spin_lock_bh(&irc_buffer_lock);
ib_ptr = skb_header_pointer(*pskb, dataoff,
(*pskb)->len - dataoff, irc_buffer);
BUG_ON(ib_ptr == NULL);
data = ib_ptr;
data_limit = ib_ptr + (*pskb)->len - dataoff;
/* strlen("\1DCC SENT t AAAAAAAA P\1\n")=24
* 5+MINMATCHLEN+strlen("t AAAAAAAA P\1\n")=14 */
while (data < (data_limit - (19 + MINMATCHLEN))) {
if (memcmp(data, "\1DCC ", 5)) {
data++;
continue;
}
data += 5;
/* we have at least (19+MINMATCHLEN)-5 bytes valid data left */
DEBUGP("DCC found in master %u.%u.%u.%u:%u %u.%u.%u.%u:%u...\n",
NIPQUAD(iph->saddr), ntohs(th->source),
NIPQUAD(iph->daddr), ntohs(th->dest));
for (i = 0; i < ARRAY_SIZE(dccprotos); i++) {
if (memcmp(data, dccprotos[i], strlen(dccprotos[i]))) {
/* no match */
continue;
}
DEBUGP("DCC %s detected\n", dccprotos[i]);
data += strlen(dccprotos[i]);
/* we have at least
* (19+MINMATCHLEN)-5-dccprotos[i].matchlen bytes valid
* data left (== 14/13 bytes) */
if (parse_dcc((char *)data, data_limit, &dcc_ip,
&dcc_port, &addr_beg_p, &addr_end_p)) {
/* unable to parse */
DEBUGP("unable to parse dcc command\n");
continue;
}
DEBUGP("DCC bound ip/port: %u.%u.%u.%u:%u\n",
HIPQUAD(dcc_ip), dcc_port);
/* dcc_ip can be the internal OR external (NAT'ed) IP
* Tiago Sousa <*****@*****.**> */
if (ct->tuplehash[dir].tuple.src.ip != htonl(dcc_ip)
&& ct->tuplehash[IP_CT_DIR_REPLY].tuple.dst.ip != htonl(dcc_ip)) {
if (net_ratelimit())
printk(KERN_WARNING
"Forged DCC command from "
"%u.%u.%u.%u: %u.%u.%u.%u:%u\n",
NIPQUAD(ct->tuplehash[dir].tuple.src.ip),
HIPQUAD(dcc_ip), dcc_port);
continue;
}
exp = ip_conntrack_expect_alloc(ct);
if (exp == NULL) {
ret = NF_DROP;
goto out;
}
/* save position of address in dcc string,
* necessary for NAT */
DEBUGP("tcph->seq = %u\n", th->seq);
seq = ntohl(th->seq) + (addr_beg_p - ib_ptr);
/* We refer to the reverse direction ("!dir")
* tuples here, because we're expecting
* something in the other * direction.
* Doesn't matter unless NAT is happening. */
exp->tuple = ((struct ip_conntrack_tuple)
{ { 0, { 0 } },
{ ct->tuplehash[!dir].tuple.dst.ip,
{ .tcp = { htons(dcc_port) } },
IPPROTO_TCP }});
/*
* XXX Liang: timeout for write is not supported yet.
*/
int
libcfs_sock_write (struct socket *sock, void *buffer, int nob, int timeout)
{
int rc;
CFS_DECL_NET_DATA;
while (nob > 0) {
struct iovec iov = {
.iov_base = buffer,
.iov_len = nob
};
struct uio suio = {
.uio_iov = &iov,
.uio_iovcnt = 1,
.uio_offset = 0,
.uio_resid = nob,
.uio_segflg = UIO_SYSSPACE,
.uio_rw = UIO_WRITE,
.uio_procp = NULL
};
CFS_NET_IN;
rc = sosend(sock, NULL, &suio, (struct mbuf *)0, (struct mbuf *)0, 0);
CFS_NET_EX;
if (rc != 0) {
if ( suio.uio_resid != nob && ( rc == ERESTART || rc == EINTR ||\
rc == EWOULDBLOCK))
rc = 0;
if ( rc != 0 )
return -rc;
rc = nob - suio.uio_resid;
buffer = ((char *)buffer) + rc;
nob = suio.uio_resid;
continue;
}
break;
}
return (0);
}
/*
* XXX Liang: timeout for read is not supported yet.
*/
int
libcfs_sock_read (struct socket *sock, void *buffer, int nob, int timeout)
{
int rc;
CFS_DECL_NET_DATA;
while (nob > 0) {
struct iovec iov = {
.iov_base = buffer,
.iov_len = nob
};
struct uio ruio = {
.uio_iov = &iov,
.uio_iovcnt = 1,
.uio_offset = 0,
.uio_resid = nob,
.uio_segflg = UIO_SYSSPACE,
.uio_rw = UIO_READ,
.uio_procp = NULL
};
CFS_NET_IN;
rc = soreceive(sock, (struct sockaddr **)0, &ruio, (struct mbuf **)0, (struct mbuf **)0, (int *)0);
CFS_NET_EX;
if (rc != 0) {
if ( ruio.uio_resid != nob && ( rc == ERESTART || rc == EINTR ||\
rc == EWOULDBLOCK))
rc = 0;
if (rc != 0)
return -rc;
rc = nob - ruio.uio_resid;
buffer = ((char *)buffer) + rc;
nob = ruio.uio_resid;
continue;
}
break;
}
return (0);
}
int
libcfs_sock_setbuf (struct socket *sock, int txbufsize, int rxbufsize)
{
struct sockopt sopt;
int rc = 0;
int option;
CFS_DECL_NET_DATA;
bzero(&sopt, sizeof sopt);
sopt.sopt_dir = SOPT_SET;
sopt.sopt_level = SOL_SOCKET;
sopt.sopt_val = &option;
sopt.sopt_valsize = sizeof(option);
if (txbufsize != 0) {
option = txbufsize;
if (option > KSOCK_MAX_BUF)
option = KSOCK_MAX_BUF;
sopt.sopt_name = SO_SNDBUF;
CFS_NET_IN;
rc = sosetopt(sock, &sopt);
CFS_NET_EX;
if (rc != 0) {
CERROR ("Can't set send buffer %d: %d\n",
option, rc);
return -rc;
}
}
if (rxbufsize != 0) {
option = rxbufsize;
sopt.sopt_name = SO_RCVBUF;
CFS_NET_IN;
rc = sosetopt(sock, &sopt);
CFS_NET_EX;
if (rc != 0) {
CERROR ("Can't set receive buffer %d: %d\n",
option, rc);
return -rc;
}
}
return 0;
}
int
libcfs_sock_getaddr (struct socket *sock, int remote, __u32 *ip, int *port)
{
struct sockaddr_in *sin;
struct sockaddr *sa = NULL;
int rc;
CFS_DECL_NET_DATA;
if (remote != 0) {
CFS_NET_IN;
rc = sock->so_proto->pr_usrreqs->pru_peeraddr(sock, &sa);
CFS_NET_EX;
if (rc != 0) {
if (sa) FREE(sa, M_SONAME);
CERROR ("Error %d getting sock peer IP\n", rc);
return -rc;
}
} else {
CFS_NET_IN;
rc = sock->so_proto->pr_usrreqs->pru_sockaddr(sock, &sa);
CFS_NET_EX;
if (rc != 0) {
if (sa) FREE(sa, M_SONAME);
CERROR ("Error %d getting sock local IP\n", rc);
return -rc;
}
}
if (sa != NULL) {
sin = (struct sockaddr_in *)sa;
if (ip != NULL)
*ip = ntohl (sin->sin_addr.s_addr);
if (port != NULL)
*port = ntohs (sin->sin_port);
if (sa)
FREE(sa, M_SONAME);
}
return 0;
}
int
libcfs_sock_getbuf (struct socket *sock, int *txbufsize, int *rxbufsize)
{
struct sockopt sopt;
int rc;
CFS_DECL_NET_DATA;
bzero(&sopt, sizeof sopt);
sopt.sopt_dir = SOPT_GET;
sopt.sopt_level = SOL_SOCKET;
if (txbufsize != NULL) {
sopt.sopt_val = txbufsize;
sopt.sopt_valsize = sizeof(*txbufsize);
sopt.sopt_name = SO_SNDBUF;
CFS_NET_IN;
rc = sogetopt(sock, &sopt);
CFS_NET_EX;
if (rc != 0) {
CERROR ("Can't get send buffer size: %d\n", rc);
return -rc;
}
}
if (rxbufsize != NULL) {
sopt.sopt_val = rxbufsize;
sopt.sopt_valsize = sizeof(*rxbufsize);
sopt.sopt_name = SO_RCVBUF;
CFS_NET_IN;
rc = sogetopt(sock, &sopt);
CFS_NET_EX;
if (rc != 0) {
CERROR ("Can't get receive buffer size: %d\n", rc);
return -rc;
}
}
return 0;
}
int
libcfs_sock_connect (struct socket **sockp, int *fatal,
__u32 local_ip, int local_port,
__u32 peer_ip, int peer_port)
{
struct sockaddr_in srvaddr;
struct socket *so;
int s;
int rc;
CFS_DECL_FUNNEL_DATA;
rc = libcfs_sock_create(sockp, fatal, local_ip, local_port);
if (rc != 0)
return rc;
so = *sockp;
bzero(&srvaddr, sizeof(srvaddr));
srvaddr.sin_len = sizeof(struct sockaddr_in);
srvaddr.sin_family = AF_INET;
srvaddr.sin_port = htons (peer_port);
srvaddr.sin_addr.s_addr = htonl (peer_ip);
CFS_NET_IN;
rc = soconnect(so, (struct sockaddr *)&srvaddr);
if (rc != 0) {
CFS_NET_EX;
if (rc != EADDRNOTAVAIL && rc != EADDRINUSE)
CDEBUG(D_NETERROR,
"Error %d connecting %u.%u.%u.%u/%d -> %u.%u.%u.%u/%d\n", rc,
HIPQUAD(local_ip), local_port, HIPQUAD(peer_ip), peer_port);
goto out;
}
s = splnet();
while ((so->so_state & SS_ISCONNECTING) && so->so_error == 0) {
CDEBUG(D_NET, "ksocknal sleep for waiting auto_connect.\n");
(void) tsleep((caddr_t)&so->so_timeo, PSOCK, "ksocknal_conn", hz);
}
if ((rc = so->so_error) != 0) {
so->so_error = 0;
splx(s);
CFS_NET_EX;
CDEBUG(D_NETERROR,
"Error %d connecting %u.%u.%u.%u/%d -> %u.%u.%u.%u/%d\n", rc,
HIPQUAD(local_ip), local_port, HIPQUAD(peer_ip), peer_port);
goto out;
}
LASSERT(so->so_state & SS_ISCONNECTED);
splx(s);
CFS_NET_EX;
if (sockp)
*sockp = so;
return (0);
out:
CFS_NET_IN;
soshutdown(so, 2);
soclose(so);
CFS_NET_EX;
return (-rc);
}
void
libcfs_sock_release (struct socket *sock)
{
CFS_DECL_FUNNEL_DATA;
CFS_NET_IN;
soshutdown(sock, 0);
CFS_NET_EX;
}
int
lnet_acceptor(void *arg)
{
char name[16];
cfs_socket_t *newsock;
int rc;
__u32 magic;
__u32 peer_ip;
int peer_port;
int secure = (int)((long_ptr_t)arg);
LASSERT (lnet_acceptor_state.pta_sock == NULL);
snprintf(name, sizeof(name), "acceptor_%03d", accept_port);
cfs_daemonize(name);
cfs_block_allsigs();
rc = libcfs_sock_listen(&lnet_acceptor_state.pta_sock,
0, accept_port, accept_backlog);
if (rc != 0) {
if (rc == -EADDRINUSE)
LCONSOLE_ERROR_MSG(0x122, "Can't start acceptor on port"
" %d: port already in use\n",
accept_port);
else
LCONSOLE_ERROR_MSG(0x123, "Can't start acceptor on port "
"%d: unexpected error %d\n",
accept_port, rc);
lnet_acceptor_state.pta_sock = NULL;
} else {
LCONSOLE(0, "Accept %s, port %d\n", accept_type, accept_port);
}
/* set init status and unblock parent */
lnet_acceptor_state.pta_shutdown = rc;
cfs_mt_complete(&lnet_acceptor_state.pta_signal);
if (rc != 0)
return rc;
while (!lnet_acceptor_state.pta_shutdown) {
rc = libcfs_sock_accept(&newsock, lnet_acceptor_state.pta_sock);
if (rc != 0) {
if (rc != -EAGAIN) {
CWARN("Accept error %d: pausing...\n", rc);
cfs_pause(cfs_time_seconds(1));
}
continue;
}
/* maybe we're waken up with libcfs_sock_abort_accept() */
if (lnet_acceptor_state.pta_shutdown) {
libcfs_sock_release(newsock);
break;
}
rc = libcfs_sock_getaddr(newsock, 1, &peer_ip, &peer_port);
if (rc != 0) {
CERROR("Can't determine new connection's address\n");
goto failed;
}
if (secure && peer_port > LNET_ACCEPTOR_MAX_RESERVED_PORT) {
CERROR("Refusing connection from %u.%u.%u.%u: "
"insecure port %d\n",
HIPQUAD(peer_ip), peer_port);
goto failed;
}
rc = libcfs_sock_read(newsock, &magic, sizeof(magic),
accept_timeout);
if (rc != 0) {
CERROR("Error %d reading connection request from "
"%u.%u.%u.%u\n", rc, HIPQUAD(peer_ip));
goto failed;
}
rc = lnet_accept(newsock, magic);
if (rc != 0)
goto failed;
continue;
failed:
libcfs_sock_release(newsock);
}
libcfs_sock_release(lnet_acceptor_state.pta_sock);
lnet_acceptor_state.pta_sock = NULL;
LCONSOLE(0, "Acceptor stopping\n");
/* unblock lnet_acceptor_stop() */
cfs_mt_complete(&lnet_acceptor_state.pta_signal);
return 0;
}
/* FIXME: This should be in userspace. Later. */
static int help(const struct iphdr *iph, size_t len,
struct ip_conntrack *ct, enum ip_conntrack_info ctinfo)
{
/* tcplen not negative guarenteed by ip_conntrack_tcp.c */
struct tcphdr *tcph = (void *) iph + iph->ihl * 4;
const char *data = (const char *) tcph + tcph->doff * 4;
const char *_data = data;
char *data_limit;
u_int32_t tcplen = len - iph->ihl * 4;
u_int32_t datalen = tcplen - tcph->doff * 4;
int dir = CTINFO2DIR(ctinfo);
struct ip_conntrack_expect expect, *exp = &expect;
struct ip_ct_irc_expect *exp_irc_info = &exp->help.exp_irc_info;
u_int32_t dcc_ip;
u_int16_t dcc_port;
int i;
char *addr_beg_p, *addr_end_p;
DEBUGP("entered\n");
/* If packet is coming from IRC server */
if (dir == IP_CT_DIR_REPLY)
return NF_ACCEPT;
/* Until there's been traffic both ways, don't look in packets. */
if (ctinfo != IP_CT_ESTABLISHED
&& ctinfo != IP_CT_ESTABLISHED + IP_CT_IS_REPLY) {
DEBUGP("Conntrackinfo = %u\n", ctinfo);
return NF_ACCEPT;
}
/* Not whole TCP header? */
if (tcplen < sizeof(struct tcphdr) || tcplen < tcph->doff * 4) {
DEBUGP("tcplen = %u\n", (unsigned) tcplen);
return NF_ACCEPT;
}
/* Checksum invalid? Ignore. */
/* FIXME: Source route IP option packets --RR */
if (tcp_v4_check(tcph, tcplen, iph->saddr, iph->daddr,
csum_partial((char *) tcph, tcplen, 0))) {
DEBUGP("bad csum: %p %u %u.%u.%u.%u %u.%u.%u.%u\n",
tcph, tcplen, NIPQUAD(iph->saddr),
NIPQUAD(iph->daddr));
return NF_ACCEPT;
}
data_limit = (char *) data + datalen;
/* strlen("\1DCC SEND t AAAAAAAA P\1\n")=24
* 5+MINMATCHLEN+strlen("t AAAAAAAA P\1\n")=14 */
while (data < (data_limit - (19 + MINMATCHLEN))) {
if (memcmp(data, "\1DCC ", 5)) {
data++;
continue;
}
data += 5;
/* we have at least (19+MINMATCHLEN)-5 bytes valid data left */
DEBUGP("DCC found in master %u.%u.%u.%u:%u %u.%u.%u.%u:%u...\n",
NIPQUAD(iph->saddr), ntohs(tcph->source),
NIPQUAD(iph->daddr), ntohs(tcph->dest));
for (i = 0; i < NUM_DCCPROTO; i++) {
if (memcmp(data, dccprotos[i].match,
dccprotos[i].matchlen)) {
/* no match */
continue;
}
DEBUGP("DCC %s detected\n", dccprotos[i].match);
data += dccprotos[i].matchlen;
/* we have at least
* (19+MINMATCHLEN)-5-dccprotos[i].matchlen bytes valid
* data left (== 14/13 bytes) */
if (parse_dcc((char *) data, data_limit, &dcc_ip,
&dcc_port, &addr_beg_p, &addr_end_p)) {
/* unable to parse */
DEBUGP("unable to parse dcc command\n");
continue;
}
DEBUGP("DCC bound ip/port: %u.%u.%u.%u:%u\n",
HIPQUAD(dcc_ip), dcc_port);
if (ct->tuplehash[dir].tuple.src.ip != htonl(dcc_ip)) {
if (net_ratelimit())
printk(KERN_WARNING
"Forged DCC command from "
"%u.%u.%u.%u: %u.%u.%u.%u:%u\n",
NIPQUAD(ct->tuplehash[dir].tuple.src.ip),
HIPQUAD(dcc_ip), dcc_port);
continue;
}
memset(&expect, 0, sizeof(expect));
LOCK_BH(&ip_irc_lock);
/* save position of address in dcc string,
* neccessary for NAT */
DEBUGP("tcph->seq = %u\n", tcph->seq);
exp->seq = ntohl(tcph->seq) + (addr_beg_p - _data);
exp_irc_info->len = (addr_end_p - addr_beg_p);
exp_irc_info->port = dcc_port;
DEBUGP("wrote info seq=%u (ofs=%u), len=%d\n",
exp->seq, (addr_end_p - _data), exp_irc_info->len);
exp->tuple = ((struct ip_conntrack_tuple)
{ { 0, { 0 } },
{ htonl(dcc_ip), { .tcp = { htons(dcc_port) } },
IPPROTO_TCP }});
int netpoll_setup(struct netpoll *np)
{
struct net_device *ndev = NULL;
struct in_device *in_dev;
struct netpoll_info *npinfo;
unsigned long flags;
int err;
if (np->dev_name)
ndev = dev_get_by_name(np->dev_name);
if (!ndev) {
printk(KERN_ERR "%s: %s doesn't exist, aborting.\n",
np->name, np->dev_name);
return -ENODEV;
}
np->dev = ndev;
if (!ndev->npinfo) {
npinfo = kmalloc(sizeof(*npinfo), GFP_KERNEL);
if (!npinfo) {
err = -ENOMEM;
goto release;
}
npinfo->rx_flags = 0;
npinfo->rx_np = NULL;
spin_lock_init(&npinfo->poll_lock);
npinfo->poll_owner = -1;
spin_lock_init(&npinfo->rx_lock);
skb_queue_head_init(&npinfo->arp_tx);
skb_queue_head_init(&npinfo->txq);
INIT_DELAYED_WORK(&npinfo->tx_work, queue_process);
atomic_set(&npinfo->refcnt, 1);
} else {
npinfo = ndev->npinfo;
atomic_inc(&npinfo->refcnt);
}
if (!ndev->poll_controller) {
printk(KERN_ERR "%s: %s doesn't support polling, aborting.\n",
np->name, np->dev_name);
err = -ENOTSUPP;
goto release;
}
if (!netif_running(ndev)) {
unsigned long atmost, atleast;
printk(KERN_INFO "%s: device %s not up yet, forcing it\n",
np->name, np->dev_name);
rtnl_lock();
err = dev_open(ndev);
rtnl_unlock();
if (err) {
printk(KERN_ERR "%s: failed to open %s\n",
np->name, ndev->name);
goto release;
}
atleast = jiffies + HZ/10;
atmost = jiffies + 4*HZ;
while (!netif_carrier_ok(ndev)) {
if (time_after(jiffies, atmost)) {
printk(KERN_NOTICE
"%s: timeout waiting for carrier\n",
np->name);
break;
}
cond_resched();
}
/* If carrier appears to come up instantly, we don't
* trust it and pause so that we don't pump all our
* queued console messages into the bitbucket.
*/
if (time_before(jiffies, atleast)) {
printk(KERN_NOTICE "%s: carrier detect appears"
" untrustworthy, waiting 4 seconds\n",
np->name);
msleep(4000);
}
}
if (is_zero_ether_addr(np->local_mac) && ndev->dev_addr)
memcpy(np->local_mac, ndev->dev_addr, 6);
if (!np->local_ip) {
rcu_read_lock();
in_dev = __in_dev_get_rcu(ndev);
if (!in_dev || !in_dev->ifa_list) {
rcu_read_unlock();
printk(KERN_ERR "%s: no IP address for %s, aborting\n",
np->name, np->dev_name);
err = -EDESTADDRREQ;
goto release;
}
np->local_ip = ntohl(in_dev->ifa_list->ifa_local);
rcu_read_unlock();
printk(KERN_INFO "%s: local IP %d.%d.%d.%d\n",
np->name, HIPQUAD(np->local_ip));
}
if (np->rx_hook) {
spin_lock_irqsave(&npinfo->rx_lock, flags);
npinfo->rx_flags |= NETPOLL_RX_ENABLED;
npinfo->rx_np = np;
spin_unlock_irqrestore(&npinfo->rx_lock, flags);
}
/* fill up the skb queue */
refill_skbs();
/* last thing to do is link it to the net device structure */
ndev->npinfo = npinfo;
/* avoid racing with NAPI reading npinfo */
synchronize_rcu();
return 0;
release:
if (!ndev->npinfo)
kfree(npinfo);
np->dev = NULL;
dev_put(ndev);
return err;
}
int
lnet_accept(cfs_socket_t *sock, __u32 magic)
{
lnet_acceptor_connreq_t cr;
__u32 peer_ip;
int peer_port;
int rc;
int flip;
lnet_ni_t *ni;
char *str;
LASSERT (sizeof(cr) <= 16); /* not too big for the stack */
rc = libcfs_sock_getaddr(sock, 1, &peer_ip, &peer_port);
LASSERT (rc == 0); /* we succeeded before */
if (!lnet_accept_magic(magic, LNET_PROTO_ACCEPTOR_MAGIC)) {
if (lnet_accept_magic(magic, LNET_PROTO_MAGIC)) {
/* future version compatibility!
* When LNET unifies protocols over all LNDs, the first
* thing sent will be a version query. I send back
* LNET_PROTO_ACCEPTOR_MAGIC to tell her I'm "old" */
memset (&cr, 0, sizeof(cr));
cr.acr_magic = LNET_PROTO_ACCEPTOR_MAGIC;
cr.acr_version = LNET_PROTO_ACCEPTOR_VERSION;
rc = libcfs_sock_write(sock, &cr, sizeof(cr),
accept_timeout);
if (rc != 0)
CERROR("Error sending magic+version in response"
"to LNET magic from %u.%u.%u.%u: %d\n",
HIPQUAD(peer_ip), rc);
return -EPROTO;
}
if (magic == le32_to_cpu(LNET_PROTO_TCP_MAGIC))
str = "'old' socknal/tcpnal";
else if (lnet_accept_magic(magic, LNET_PROTO_RA_MAGIC))
str = "'old' ranal";
else
str = "unrecognised";
LCONSOLE_ERROR_MSG(0x11f, "Refusing connection from %u.%u.%u.%u"
" magic %08x: %s acceptor protocol\n",
HIPQUAD(peer_ip), magic, str);
return -EPROTO;
}
flip = (magic != LNET_PROTO_ACCEPTOR_MAGIC);
rc = libcfs_sock_read(sock, &cr.acr_version,
sizeof(cr.acr_version),
accept_timeout);
if (rc != 0) {
CERROR("Error %d reading connection request version from "
"%u.%u.%u.%u\n", rc, HIPQUAD(peer_ip));
return -EIO;
}
if (flip)
__swab32s(&cr.acr_version);
if (cr.acr_version != LNET_PROTO_ACCEPTOR_VERSION) {
/* future version compatibility!
* An acceptor-specific protocol rev will first send a version
* query. I send back my current version to tell her I'm
* "old". */
int peer_version = cr.acr_version;
memset (&cr, 0, sizeof(cr));
cr.acr_magic = LNET_PROTO_ACCEPTOR_MAGIC;
cr.acr_version = LNET_PROTO_ACCEPTOR_VERSION;
rc = libcfs_sock_write(sock, &cr, sizeof(cr),
accept_timeout);
if (rc != 0)
CERROR("Error sending magic+version in response"
"to version %d from %u.%u.%u.%u: %d\n",
peer_version, HIPQUAD(peer_ip), rc);
return -EPROTO;
}
rc = libcfs_sock_read(sock, &cr.acr_nid,
sizeof(cr) -
offsetof(lnet_acceptor_connreq_t, acr_nid),
accept_timeout);
if (rc != 0) {
CERROR("Error %d reading connection request from "
"%u.%u.%u.%u\n", rc, HIPQUAD(peer_ip));
return -EIO;
}
if (flip)
__swab64s(&cr.acr_nid);
ni = lnet_net2ni(LNET_NIDNET(cr.acr_nid));
if (ni == NULL || /* no matching net */
ni->ni_nid != cr.acr_nid) { /* right NET, wrong NID! */
if (ni != NULL)
lnet_ni_decref(ni);
LCONSOLE_ERROR_MSG(0x120, "Refusing connection from %u.%u.%u.%u"
" for %s: No matching NI\n",
HIPQUAD(peer_ip), libcfs_nid2str(cr.acr_nid));
return -EPERM;
}
if (ni->ni_lnd->lnd_accept == NULL) {
/* This catches a request for the loopback LND */
lnet_ni_decref(ni);
LCONSOLE_ERROR_MSG(0x121, "Refusing connection from %u.%u.%u.%u"
" for %s: NI doesn not accept IP connections\n",
HIPQUAD(peer_ip), libcfs_nid2str(cr.acr_nid));
return -EPERM;
}
CDEBUG(D_NET, "Accept %s from %u.%u.%u.%u\n",
libcfs_nid2str(cr.acr_nid), HIPQUAD(peer_ip));
rc = ni->ni_lnd->lnd_accept(ni, sock);
lnet_ni_decref(ni);
return rc;
}
/* All actions that we need after receiving hello on passive conn:
* 1) Stash peer's nid, pid, incarnation and conn type
* 2) Cope with easy case: conn[idx] is empty - just save conn there
* 3) Resolve race:
* a) if our nid is higher - reply with CONN_NONE and make us zombie
* b) if peer's nid is higher - postpone race resolution till
* READY state
* 4) Anyhow, send reply hello
*/
int
usocklnd_passiveconn_hellorecv(usock_conn_t *conn)
{
ksock_hello_msg_t *hello = conn->uc_rx_hello;
int type;
int idx;
int rc;
usock_peer_t *peer;
lnet_ni_t *ni = conn->uc_ni;
__u32 peer_ip = conn->uc_peer_ip;
__u16 peer_port = conn->uc_peer_port;
/* don't know parent peer yet and not zombie */
LASSERT (conn->uc_peer == NULL &&
ni != NULL);
/* don't know peer's nid and incarnation yet */
if (peer_port > LNET_ACCEPTOR_MAX_RESERVED_PORT) {
/* do not trust liblustre clients */
conn->uc_peerid.pid = peer_port | LNET_PID_USERFLAG;
conn->uc_peerid.nid = LNET_MKNID(LNET_NIDNET(ni->ni_nid),
peer_ip);
if (hello->kshm_ctype != SOCKLND_CONN_ANY) {
lnet_ni_decref(ni);
conn->uc_ni = NULL;
CERROR("Refusing to accept connection of type=%d from "
"userspace process %u.%u.%u.%u:%d\n", hello->kshm_ctype,
HIPQUAD(peer_ip), peer_port);
return -EINVAL;
}
} else {
conn->uc_peerid.pid = hello->kshm_src_pid;
conn->uc_peerid.nid = hello->kshm_src_nid;
}
conn->uc_type = type = usocklnd_invert_type(hello->kshm_ctype);
rc = usocklnd_find_or_create_peer(ni, conn->uc_peerid, &peer);
if (rc) {
lnet_ni_decref(ni);
conn->uc_ni = NULL;
return rc;
}
peer->up_last_alive = cfs_time_current();
idx = usocklnd_type2idx(conn->uc_type);
/* safely check whether we're first */
pthread_mutex_lock(&peer->up_lock);
usocklnd_cleanup_stale_conns(peer, hello->kshm_src_incarnation, NULL);
if (peer->up_conns[idx] == NULL) {
peer->up_last_alive = cfs_time_current();
conn->uc_peer = peer;
conn->uc_ni = NULL;
usocklnd_link_conn_to_peer(conn, peer, idx);
usocklnd_conn_addref(conn);
} else {
usocklnd_peer_decref(peer);
/* Resolve race in favour of higher NID */
if (conn->uc_peerid.nid < conn->uc_ni->ni_nid) {
/* make us zombie */
conn->uc_ni = NULL;
type = SOCKLND_CONN_NONE;
}
/* if conn->uc_peerid.nid > conn->uc_ni->ni_nid,
* postpone race resolution till READY state
* (hopefully that conn[idx] will die because of
* incoming hello of CONN_NONE type) */
}
pthread_mutex_unlock(&peer->up_lock);
/* allocate and initialize fake tx with hello */
conn->uc_tx_hello = usocklnd_create_hello_tx(ni, type,
conn->uc_peerid.nid);
if (conn->uc_ni == NULL)
lnet_ni_decref(ni);
if (conn->uc_tx_hello == NULL)
return -ENOMEM;
/* rc == 0 */
pthread_mutex_lock(&conn->uc_lock);
if (conn->uc_state == UC_DEAD)
goto passive_hellorecv_done;
conn->uc_state = UC_SENDING_HELLO;
conn->uc_tx_deadline = cfs_time_shift(usock_tuns.ut_timeout);
conn->uc_tx_flag = 1;
rc = usocklnd_add_pollrequest(conn, POLL_SET_REQUEST, POLLOUT);
passive_hellorecv_done:
pthread_mutex_unlock(&conn->uc_lock);
return rc;
}
int netpoll_parse_options(struct netpoll *np, char *opt)
{
char *cur=opt, *delim;
if (*cur != '@') {
if ((delim = strchr(cur, '@')) == NULL)
goto parse_failed;
*delim = 0;
np->local_port = simple_strtol(cur, NULL, 10);
cur = delim;
}
cur++;
printk(KERN_INFO "%s: local port %d\n", np->name, np->local_port);
if (*cur != '/') {
if ((delim = strchr(cur, '/')) == NULL)
goto parse_failed;
*delim = 0;
np->local_ip = ntohl(in_aton(cur));
cur = delim;
printk(KERN_INFO "%s: local IP %d.%d.%d.%d\n",
np->name, HIPQUAD(np->local_ip));
}
cur++;
if (*cur != ',') {
/* parse out dev name */
if ((delim = strchr(cur, ',')) == NULL)
goto parse_failed;
*delim = 0;
strlcpy(np->dev_name, cur, sizeof(np->dev_name));
cur = delim;
}
cur++;
printk(KERN_INFO "%s: interface %s\n", np->name, np->dev_name);
if (*cur != '@') {
/* dst port */
if ((delim = strchr(cur, '@')) == NULL)
goto parse_failed;
*delim = 0;
np->remote_port = simple_strtol(cur, NULL, 10);
cur = delim;
}
cur++;
printk(KERN_INFO "%s: remote port %d\n", np->name, np->remote_port);
/* dst ip */
if ((delim = strchr(cur, '/')) == NULL)
goto parse_failed;
*delim = 0;
np->remote_ip = ntohl(in_aton(cur));
cur = delim + 1;
printk(KERN_INFO "%s: remote IP %d.%d.%d.%d\n",
np->name, HIPQUAD(np->remote_ip));
if (*cur != 0) {
/* MAC address */
if ((delim = strchr(cur, ':')) == NULL)
goto parse_failed;
*delim = 0;
np->remote_mac[0] = simple_strtol(cur, NULL, 16);
cur = delim + 1;
if ((delim = strchr(cur, ':')) == NULL)
goto parse_failed;
*delim = 0;
np->remote_mac[1] = simple_strtol(cur, NULL, 16);
cur = delim + 1;
if ((delim = strchr(cur, ':')) == NULL)
goto parse_failed;
*delim = 0;
np->remote_mac[2] = simple_strtol(cur, NULL, 16);
cur = delim + 1;
if ((delim = strchr(cur, ':')) == NULL)
goto parse_failed;
*delim = 0;
np->remote_mac[3] = simple_strtol(cur, NULL, 16);
cur = delim + 1;
if ((delim = strchr(cur, ':')) == NULL)
goto parse_failed;
*delim = 0;
np->remote_mac[4] = simple_strtol(cur, NULL, 16);
cur = delim + 1;
np->remote_mac[5] = simple_strtol(cur, NULL, 16);
}
printk(KERN_INFO "%s: remote ethernet address "
"%02x:%02x:%02x:%02x:%02x:%02x\n",
np->name,
np->remote_mac[0],
np->remote_mac[1],
np->remote_mac[2],
np->remote_mac[3],
np->remote_mac[4],
np->remote_mac[5]);
return 0;
parse_failed:
printk(KERN_INFO "%s: couldn't parse config at %s!\n",
np->name, cur);
return -1;
}
static int help(struct sk_buff **pskb, unsigned int protoff,
struct nf_conn *ct, enum ip_conntrack_info ctinfo)
{
unsigned int dataoff;
struct iphdr *iph;
struct tcphdr _tcph, *th;
char *data, *data_limit, *ib_ptr;
int dir = CTINFO2DIR(ctinfo);
struct nf_conntrack_expect *exp;
struct nf_conntrack_tuple *tuple;
u_int32_t dcc_ip;
u_int16_t dcc_port;
__be16 port;
int i, ret = NF_ACCEPT;
char *addr_beg_p, *addr_end_p;
typeof(nf_nat_irc_hook) nf_nat_irc;
/* If packet is coming from IRC server */
if (dir == IP_CT_DIR_REPLY)
return NF_ACCEPT;
/* Until there's been traffic both ways, don't look in packets. */
if (ctinfo != IP_CT_ESTABLISHED &&
ctinfo != IP_CT_ESTABLISHED + IP_CT_IS_REPLY)
return NF_ACCEPT;
/* Not a full tcp header? */
th = skb_header_pointer(*pskb, protoff, sizeof(_tcph), &_tcph);
if (th == NULL)
return NF_ACCEPT;
/* No data? */
dataoff = protoff + th->doff*4;
if (dataoff >= (*pskb)->len)
return NF_ACCEPT;
spin_lock_bh(&irc_buffer_lock);
ib_ptr = skb_header_pointer(*pskb, dataoff, (*pskb)->len - dataoff,
irc_buffer);
BUG_ON(ib_ptr == NULL);
data = ib_ptr;
data_limit = ib_ptr + (*pskb)->len - dataoff;
/* strlen("\1DCC SENT t AAAAAAAA P\1\n")=24
* 5+MINMATCHLEN+strlen("t AAAAAAAA P\1\n")=14 */
while (data < data_limit - (19 + MINMATCHLEN)) {
if (memcmp(data, "\1DCC ", 5)) {
data++;
continue;
}
data += 5;
/* we have at least (19+MINMATCHLEN)-5 bytes valid data left */
iph = ip_hdr(*pskb);
pr_debug("DCC found in master %u.%u.%u.%u:%u %u.%u.%u.%u:%u\n",
NIPQUAD(iph->saddr), ntohs(th->source),
NIPQUAD(iph->daddr), ntohs(th->dest));
for (i = 0; i < ARRAY_SIZE(dccprotos); i++) {
if (memcmp(data, dccprotos[i], strlen(dccprotos[i]))) {
/* no match */
continue;
}
data += strlen(dccprotos[i]);
pr_debug("DCC %s detected\n", dccprotos[i]);
/* we have at least
* (19+MINMATCHLEN)-5-dccprotos[i].matchlen bytes valid
* data left (== 14/13 bytes) */
if (parse_dcc((char *)data, data_limit, &dcc_ip,
&dcc_port, &addr_beg_p, &addr_end_p)) {
pr_debug("unable to parse dcc command\n");
continue;
}
pr_debug("DCC bound ip/port: %u.%u.%u.%u:%u\n",
HIPQUAD(dcc_ip), dcc_port);
/* dcc_ip can be the internal OR external (NAT'ed) IP */
tuple = &ct->tuplehash[dir].tuple;
if (tuple->src.u3.ip != htonl(dcc_ip) &&
tuple->dst.u3.ip != htonl(dcc_ip)) {
if (net_ratelimit())
printk(KERN_WARNING
"Forged DCC command from "
"%u.%u.%u.%u: %u.%u.%u.%u:%u\n",
NIPQUAD(tuple->src.u3.ip),
HIPQUAD(dcc_ip), dcc_port);
continue;
}
exp = nf_ct_expect_alloc(ct);
if (exp == NULL) {
ret = NF_DROP;
goto out;
}
tuple = &ct->tuplehash[!dir].tuple;
port = htons(dcc_port);
nf_ct_expect_init(exp, tuple->src.l3num,
NULL, &tuple->dst.u3,
IPPROTO_TCP, NULL, &port);
nf_nat_irc = rcu_dereference(nf_nat_irc_hook);
if (nf_nat_irc && ct->status & IPS_NAT_MASK)
ret = nf_nat_irc(pskb, ctinfo,
addr_beg_p - ib_ptr,
addr_end_p - addr_beg_p,
exp);
else if (nf_ct_expect_related(exp) != 0)
ret = NF_DROP;
nf_ct_expect_put(exp);
goto out;
}
}
out:
spin_unlock_bh(&irc_buffer_lock);
return ret;
}
/* Switch on rx_state.
* Return 0 on success, else return <0
* Always set cont_flag: 1 if we're ready to continue reading, else 0
*/
int
usocklnd_read_hello(usock_conn_t *conn, int *cont_flag)
{
int rc = 0;
ksock_hello_msg_t *hello = conn->uc_rx_hello;
*cont_flag = 0;
/* smth. new emerged in hello - let's process it */
switch (conn->uc_rx_state) {
case UC_RX_HELLO_MAGIC:
if (hello->kshm_magic == LNET_PROTO_MAGIC)
conn->uc_flip = 0;
else if (hello->kshm_magic == __swab32(LNET_PROTO_MAGIC))
conn->uc_flip = 1;
else
return -EPROTO;
usocklnd_rx_helloversion_state_transition(conn);
*cont_flag = 1;
break;
case UC_RX_HELLO_VERSION:
if ((!conn->uc_flip &&
(hello->kshm_version != KSOCK_PROTO_V2)) ||
(conn->uc_flip &&
(hello->kshm_version != __swab32(KSOCK_PROTO_V2))))
return -EPROTO;
usocklnd_rx_hellobody_state_transition(conn);
*cont_flag = 1;
break;
case UC_RX_HELLO_BODY:
if (conn->uc_flip) {
ksock_hello_msg_t *hello = conn->uc_rx_hello;
__swab32s(&hello->kshm_src_pid);
__swab64s(&hello->kshm_src_nid);
__swab32s(&hello->kshm_dst_pid);
__swab64s(&hello->kshm_dst_nid);
__swab64s(&hello->kshm_src_incarnation);
__swab64s(&hello->kshm_dst_incarnation);
__swab32s(&hello->kshm_ctype);
__swab32s(&hello->kshm_nips);
}
if (conn->uc_rx_hello->kshm_nips > LNET_MAX_INTERFACES) {
CERROR("Bad nips %d from ip %u.%u.%u.%u port %d\n",
conn->uc_rx_hello->kshm_nips,
HIPQUAD(conn->uc_peer_ip), conn->uc_peer_port);
return -EPROTO;
}
if (conn->uc_rx_hello->kshm_nips) {
usocklnd_rx_helloIPs_state_transition(conn);
*cont_flag = 1;
break;
}
/* fall through */
case UC_RX_HELLO_IPS:
if (conn->uc_activeflag == 1) /* active conn */
rc = usocklnd_activeconn_hellorecv(conn);
else /* passive conn */
rc = usocklnd_passiveconn_hellorecv(conn);
break;
default:
LBUG(); /* unknown state */
}
return rc;
}
static int help(struct sk_buff *skb,
struct ip_conntrack *ct, enum ip_conntrack_info ctinfo)
{
unsigned int dataoff;
struct tcphdr tcph;
char *data, *data_limit;
int dir = CTINFO2DIR(ctinfo);
struct ip_conntrack_expect *exp;
struct ip_ct_irc_expect *exp_irc_info = NULL;
u_int32_t dcc_ip;
u_int16_t dcc_port;
int i;
char *addr_beg_p, *addr_end_p;
DEBUGP("entered\n");
/* If packet is coming from IRC server */
if (dir == IP_CT_DIR_REPLY)
return NF_ACCEPT;
/* Until there's been traffic both ways, don't look in packets. */
if (ctinfo != IP_CT_ESTABLISHED
&& ctinfo != IP_CT_ESTABLISHED + IP_CT_IS_REPLY) {
DEBUGP("Conntrackinfo = %u\n", ctinfo);
return NF_ACCEPT;
}
/* Not a full tcp header? */
if (skb_copy_bits(skb, skb->nh.iph->ihl*4, &tcph, sizeof(tcph)) != 0)
return NF_ACCEPT;
/* No data? */
dataoff = skb->nh.iph->ihl*4 + tcph.doff*4;
if (dataoff >= skb->len)
return NF_ACCEPT;
LOCK_BH(&ip_irc_lock);
skb_copy_bits(skb, dataoff, irc_buffer, skb->len - dataoff);
data = irc_buffer;
data_limit = irc_buffer + skb->len - dataoff;
/* strlen("\1DCC SENT t AAAAAAAA P\1\n")=24
* 5+MINMATCHLEN+strlen("t AAAAAAAA P\1\n")=14 */
while (data < (data_limit - (19 + MINMATCHLEN))) {
if (memcmp(data, "\1DCC ", 5)) {
data++;
continue;
}
data += 5;
/* we have at least (19+MINMATCHLEN)-5 bytes valid data left */
DEBUGP("DCC found in master %u.%u.%u.%u:%u %u.%u.%u.%u:%u...\n",
NIPQUAD(iph->saddr), ntohs(tcph.source),
NIPQUAD(iph->daddr), ntohs(tcph.dest));
for (i = 0; i < ARRAY_SIZE(dccprotos); i++) {
if (memcmp(data, dccprotos[i], strlen(dccprotos[i]))) {
/* no match */
continue;
}
DEBUGP("DCC %s detected\n", dccprotos[i]);
data += strlen(dccprotos[i]);
/* we have at least
* (19+MINMATCHLEN)-5-dccprotos[i].matchlen bytes valid
* data left (== 14/13 bytes) */
if (parse_dcc((char *)data, data_limit, &dcc_ip,
&dcc_port, &addr_beg_p, &addr_end_p)) {
/* unable to parse */
DEBUGP("unable to parse dcc command\n");
continue;
}
DEBUGP("DCC bound ip/port: %u.%u.%u.%u:%u\n",
HIPQUAD(dcc_ip), dcc_port);
/* dcc_ip can be the internal OR external (NAT'ed) IP
* Tiago Sousa <*****@*****.**> */
if (ct->tuplehash[dir].tuple.src.ip != htonl(dcc_ip)
&& ct->tuplehash[IP_CT_DIR_REPLY].tuple.dst.ip != htonl(dcc_ip)) {
if (net_ratelimit())
printk(KERN_WARNING
"Forged DCC command from "
"%u.%u.%u.%u: %u.%u.%u.%u:%u\n",
NIPQUAD(ct->tuplehash[dir].tuple.src.ip),
HIPQUAD(dcc_ip), dcc_port);
continue;
}
exp = ip_conntrack_expect_alloc();
if (exp == NULL)
goto out;
exp_irc_info = &exp->help.exp_irc_info;
/* save position of address in dcc string,
* necessary for NAT */
DEBUGP("tcph->seq = %u\n", tcph.seq);
exp->seq = ntohl(tcph.seq) + (addr_beg_p - irc_buffer);
exp_irc_info->len = (addr_end_p - addr_beg_p);
exp_irc_info->port = dcc_port;
DEBUGP("wrote info seq=%u (ofs=%u), len=%d\n",
exp->seq, (addr_end_p - _data), exp_irc_info->len);
exp->tuple = ((struct ip_conntrack_tuple)
{ { 0, { 0 } },
{ ct->tuplehash[dir].tuple.src.ip, { .tcp = { htons(dcc_port) } },
IPPROTO_TCP }});
static int create(struct ip_set *set, const void *data, size_t size)
{
int newbytes;
struct ip_set_req_ipmap_create *req =
(struct ip_set_req_ipmap_create *) data;
struct ip_set_ipmap *map;
if (size != sizeof(struct ip_set_req_ipmap_create)) {
ip_set_printk("data length wrong (want %zu, have %zu)",
sizeof(struct ip_set_req_ipmap_create),
size);
return -EINVAL;
}
DP("from %u.%u.%u.%u to %u.%u.%u.%u",
HIPQUAD(req->from), HIPQUAD(req->to));
if (req->from > req->to) {
DP("bad ip range");
return -ENOEXEC;
}
map = kmalloc(sizeof(struct ip_set_ipmap), GFP_KERNEL);
if (!map) {
DP("out of memory for %d bytes",
sizeof(struct ip_set_ipmap));
return -ENOMEM;
}
map->first_ip = req->from;
map->last_ip = req->to;
map->netmask = req->netmask;
if (req->netmask == 0xFFFFFFFF) {
map->hosts = 1;
map->sizeid = map->last_ip - map->first_ip + 1;
} else {
unsigned int mask_bits, netmask_bits;
ip_set_ip_t mask;
map->first_ip &= map->netmask; /* Should we better bark? */
mask = range_to_mask(map->first_ip, map->last_ip, &mask_bits);
netmask_bits = mask_to_bits(map->netmask);
if ((!mask && (map->first_ip || map->last_ip != 0xFFFFFFFF))
|| netmask_bits <= mask_bits)
return -ENOEXEC;
DP("mask_bits %u, netmask_bits %u",
mask_bits, netmask_bits);
map->hosts = 2 << (32 - netmask_bits - 1);
map->sizeid = 2 << (netmask_bits - mask_bits - 1);
}
if (map->sizeid > MAX_RANGE + 1) {
ip_set_printk("range too big (max %d addresses)",
MAX_RANGE+1);
kfree(map);
return -ENOEXEC;
}
DP("hosts %u, sizeid %u", map->hosts, map->sizeid);
newbytes = bitmap_bytes(0, map->sizeid - 1);
map->members = kmalloc(newbytes, GFP_KERNEL);
if (!map->members) {
DP("out of memory for %d bytes", newbytes);
kfree(map);
return -ENOMEM;
}
memset(map->members, 0, newbytes);
set->data = map;
return 0;
}
//start_listening
void moon_listen(void)
{
//For test purposes
moonraker_socket_t *listen;
listen = kmalloc(sizeof(moonraker_socket_t),GFP_KERNEL);
listen->proto = kmalloc(sizeof(moonraker_proto_t),GFP_KERNEL);
listen->proto->name = kmalloc(5,GFP_KERNEL);
listen->ip = 2130706433;
listen->port = 80;
listen->proto->defer_accept=0;
listen->keepalive_timeout=0;
listen->ack_pingpong=1;
listen->max_backlog=2048;
listen->defer_accept=1;
strcpy(listen->proto->name,"http");
//end for test purpose
struct sockaddr_in sin;
struct socket *sock = NULL;
struct sock *sk;
struct tcp_sock *tp;
struct inet_connection_sock *icsk;
moonraker_proto_t *proto = listen->proto;
u16 port = listen->port;
u32 addr = listen->ip; //127.0.0.1
int err = 0;
err = sock_create_kern(PF_INET, SOCK_STREAM, IPPROTO_TCP, &sock);
if (err) {
printk(KERN_ERR "Moonraker: error %d creating socket.\n", err);
goto error;
}
sin.sin_family = AF_INET;
sin.sin_addr.s_addr = htonl(addr);
sin.sin_port = htons(port);
sk = sock->sk;
icsk = inet_csk(sk);
sk->sk_reuse = 1;
sock_set_flag(sk, SOCK_URGINLINE);
err = sock->ops->bind(sock, (struct sockaddr*)&sin, sizeof(sin));
if (err){
printk(KERN_ERR "Moonraker: error %d binding socket. This means that probably some other process is (or was a short time ago) using addr %d\n",err,sin.sin_addr.s_addr);
goto error;
}
tp = tcp_sk(sk);
printk("listen sk accept_queue: %d.\n",!reqsk_queue_empty(&icsk->icsk_accept_queue));
icsk->icsk_ack.pingpong = listen->ack_pingpong;
sock_reset_flag(sk, SOCK_LINGER);
sk->sk_lingertime = 0;
tp->linger2 = listen->keepalive_timeout * HZ;
if (proto->defer_accept && !listen->keepalive_timeout && listen->defer_accept)
icsk->icsk_accept_queue.rskq_defer_accept = 1;
err = sock->ops->listen(sock, listen->max_backlog);
if (err) {
printk(KERN_ERR "Moonraker: error %d listening on socket.\n", err);
goto error;
}
printk(KERN_NOTICE "Moonraker: thread %d listens on %s://%d.%d.%d.%d:%d.\n",
1, proto->name, HIPQUAD(addr), port);
// return sock;
return;
error:
if (sock)
sock_release(sock);
return;
return NULL;
}
void print_skb_data(const char *msg, int count, struct sk_buff *skb, u8 *data, int len)
{
static int skb_count = 1000000;
u8 *ptr, *end;
u32 src_addr, dst_addr;
// Transport layer header.
union {
struct tcphdr *th;
struct udphdr *uh;
struct icmphdr *icmph;
struct igmphdr *igmph;
struct iphdr *ipiph;
unsigned char *raw;
} h;
// Network layer header.
union {
struct iphdr *iph;
struct ipv6hdr *ipv6h;
struct arpheader *arph;
struct ipxhdr *ipxh;
unsigned char *raw;
} nh;
// Link layer header.
union {
struct ethhdr *ethernet;
unsigned char *raw;
} mac;
int protocol;
if(!count) count = ++skb_count;
if(!msg) msg = (char *)__FUNCTION__;
if(!data){
printk("%s.%d> null data\n", msg, count);
return;
}
ptr = data;
end = data + len;
mac.raw = ptr;
ptr += sizeof(struct ethhdr);
if(ptr > end){ printk("***MAC:"); goto exit; }
protocol = ntohs(mac.ethernet->h_proto);
nh.raw = ptr;
printk("%s.%d> type=%d protocol=0x%x\n",
msg, count, skb->pkt_type, htons(skb->protocol));
if(1){
printk("%s.%d> %p mac src=" MACFMT " dst=" MACFMT "\n",
msg, count, data,
MAC6TUPLE(mac.ethernet->h_source),
MAC6TUPLE(mac.ethernet->h_dest));
}
switch(protocol){
case ETH_P_ARP:
ptr += sizeof(struct arpheader);
if(ptr > end){ printk("***ARP:"); goto exit; }
if(0){
printk("%s.%d> ARP hrd=%d, pro=%d, hln=%d, pln=%d, op=%d\n",
msg, count,
nh.arph->ar_hrd, nh.arph->ar_pro, nh.arph->ar_hln,
nh.arph->ar_pln, nh.arph->ar_op);
}
memcpy(&src_addr, nh.arph->ar_sip, 4);
src_addr = ntohl(src_addr);
memcpy(&dst_addr, nh.arph->ar_tip, 4);
dst_addr = ntohl(dst_addr);
printk("%s.%d> ARP HW src=" MACFMT " dst=" MACFMT "\n",
msg, count, MAC6TUPLE(nh.arph->ar_sha), MAC6TUPLE(nh.arph->ar_tha));
printk("%s.%d> ARP IP src=" IPFMT " dst=" IPFMT "\n",
msg, count, HIPQUAD(src_addr), HIPQUAD(dst_addr));
break;
case ETH_P_IP: {
u16 src_port, dst_port;
if(ptr + sizeof(struct iphdr) > end){ printk("***IP:"); goto exit; }
src_addr = ntohl(nh.iph->saddr);
dst_addr = ntohl(nh.iph->daddr);
if(1){
printk("%s.%d> IP proto=%d src=" IPFMT " dst=" IPFMT "\n",
msg, count, nh.iph->protocol,
HIPQUAD(src_addr), HIPQUAD(dst_addr));
printk("%s.%d> IP tot_len=%u len=%d\n",
msg, count, ntohs(nh.iph->tot_len), len - ETH_HLEN);
}
ptr += (nh.iph->ihl * 4);
if(ptr > end){ printk ("***IP: len"); goto exit; }
h.raw = ptr;
switch(nh.iph->protocol){
case IPPROTO_TCP:
ptr += sizeof(struct tcphdr);
if(ptr > end){ printk("***TCP:"); goto exit; }
src_port = ntohs(h.th->source);
dst_port = ntohs(h.th->dest);
printk("%s.%d> TCP src=" IPFMT ":%u dst=" IPFMT ":%u\n",
msg, count,
HIPQUAD(src_addr), src_port,
HIPQUAD(dst_addr), dst_port);
break;
case IPPROTO_UDP:
ptr += sizeof(struct udphdr);
if(ptr > end){ printk("***UDP:"); goto exit; }
src_port = ntohs(h.uh->source);
dst_port = ntohs(h.uh->dest);
printk("%s.%d> UDP src=" IPFMT ":%u dst=" IPFMT ":%u\n",
msg, count,
HIPQUAD(src_addr), src_port,
HIPQUAD(dst_addr), dst_port);
break;
default:
printk("%s.%d> IP %d src=" IPFMT " dst=" IPFMT "\n",
msg, count,
nh.iph->protocol, HIPQUAD(src_addr), HIPQUAD(dst_addr));
break;
}
break; }
case ETH_P_IPV6:
printk("%s.%d> IPv6\n", msg, count);
break;
case ETH_P_IPX:
printk("%s.%d> IPX\n", msg, count);
break;
default:
printk("%s.%d> protocol=%d\n", msg, count, protocol);
break;
}
return;
exit:
printk("%s.%d> %s: skb problem\n", msg, count, __FUNCTION__);
printk("%s.%d> %s: data=%p end=%p(%d) ptr=%p(%d) eth=%d ip=%d\n",
msg, count, __FUNCTION__,
data, end, end - data, ptr, ptr - data,
sizeof(struct ethhdr),
sizeof(struct iphdr));
return;
}
static void
isdn_net_ciscohdlck_slarp_in(isdn_net_dev *idev, struct sk_buff *skb)
{
isdn_net_local *mlp = idev->mlp;
struct inl_cisco *cisco = mlp->inl_priv;
unsigned char *p;
int period;
u32 code;
u32 my_seq, addr;
u32 your_seq, mask;
u32 local;
u16 unused;
if (skb->len < 14)
return;
p = skb->data;
p += get_u32(p, &code);
switch (code) {
case CISCO_SLARP_REQUEST:
cisco->yourseq = 0;
isdn_net_ciscohdlck_slarp_send_reply(idev);
break;
case CISCO_SLARP_REPLY:
addr = ntohl(*(u32 *)p);
mask = ntohl(*(u32 *)(p+4));
if (mask != 0xfffffffc)
goto slarp_reply_out;
if ((addr & 3) == 0 || (addr & 3) == 3)
goto slarp_reply_out;
local = addr ^ 3;
printk(KERN_INFO "%s: got slarp reply: "
"remote ip: %d.%d.%d.%d, "
"local ip: %d.%d.%d.%d "
"mask: %d.%d.%d.%d\n",
idev->name,
HIPQUAD(addr),
HIPQUAD(local),
HIPQUAD(mask));
break;
slarp_reply_out:
printk(KERN_INFO "%s: got invalid slarp "
"reply (%d.%d.%d.%d/%d.%d.%d.%d) "
"- ignored\n", idev->name,
HIPQUAD(addr), HIPQUAD(mask));
break;
case CISCO_SLARP_KEEPALIVE:
period = (jiffies - cisco->last_slarp_in + HZ/2 - 1) / HZ;
if (cisco->debserint &&
(period != cisco->keepalive_period) &&
cisco->last_slarp_in) {
printk(KERN_DEBUG "%s: Keepalive period mismatch - "
"is %d but should be %d.\n",
idev->name, period, cisco->keepalive_period);
}
cisco->last_slarp_in = jiffies;
p += get_u32(p, &my_seq);
p += get_u32(p, &your_seq);
p += get_u16(p, &unused);
cisco->yourseq = my_seq;
cisco->mineseen = your_seq;
break;
}
}
