2005-04-16 18:20:36 -04:00
|
|
|
/*
|
|
|
|
* TCP over IPv6
|
2007-02-09 09:24:49 -05:00
|
|
|
* Linux INET6 implementation
|
2005-04-16 18:20:36 -04:00
|
|
|
*
|
|
|
|
* Authors:
|
2007-02-09 09:24:49 -05:00
|
|
|
* Pedro Roque <roque@di.fc.ul.pt>
|
2005-04-16 18:20:36 -04:00
|
|
|
*
|
2007-02-09 09:24:49 -05:00
|
|
|
* Based on:
|
2005-04-16 18:20:36 -04:00
|
|
|
* linux/net/ipv4/tcp.c
|
|
|
|
* linux/net/ipv4/tcp_input.c
|
|
|
|
* linux/net/ipv4/tcp_output.c
|
|
|
|
*
|
|
|
|
* Fixes:
|
|
|
|
* Hideaki YOSHIFUJI : sin6_scope_id support
|
|
|
|
* YOSHIFUJI Hideaki @USAGI and: Support IPV6_V6ONLY socket option, which
|
|
|
|
* Alexey Kuznetsov allow both IPv4 and IPv6 sockets to bind
|
|
|
|
* a single port at the same time.
|
|
|
|
* YOSHIFUJI Hideaki @USAGI: convert /proc/net/tcp6 to seq_file.
|
|
|
|
*
|
|
|
|
* This program is free software; you can redistribute it and/or
|
|
|
|
* modify it under the terms of the GNU General Public License
|
|
|
|
* as published by the Free Software Foundation; either version
|
|
|
|
* 2 of the License, or (at your option) any later version.
|
|
|
|
*/
|
|
|
|
|
2008-12-30 02:04:08 -05:00
|
|
|
#include <linux/bottom_half.h>
|
2005-04-16 18:20:36 -04:00
|
|
|
#include <linux/module.h>
|
|
|
|
#include <linux/errno.h>
|
|
|
|
#include <linux/types.h>
|
|
|
|
#include <linux/socket.h>
|
|
|
|
#include <linux/sockios.h>
|
|
|
|
#include <linux/net.h>
|
|
|
|
#include <linux/jiffies.h>
|
|
|
|
#include <linux/in.h>
|
|
|
|
#include <linux/in6.h>
|
|
|
|
#include <linux/netdevice.h>
|
|
|
|
#include <linux/init.h>
|
|
|
|
#include <linux/jhash.h>
|
|
|
|
#include <linux/ipsec.h>
|
|
|
|
#include <linux/times.h>
|
|
|
|
|
|
|
|
#include <linux/ipv6.h>
|
|
|
|
#include <linux/icmpv6.h>
|
|
|
|
#include <linux/random.h>
|
|
|
|
|
|
|
|
#include <net/tcp.h>
|
|
|
|
#include <net/ndisc.h>
|
2005-08-12 08:26:18 -04:00
|
|
|
#include <net/inet6_hashtables.h>
|
2005-12-14 02:15:24 -05:00
|
|
|
#include <net/inet6_connection_sock.h>
|
2005-04-16 18:20:36 -04:00
|
|
|
#include <net/ipv6.h>
|
|
|
|
#include <net/transp_v6.h>
|
|
|
|
#include <net/addrconf.h>
|
|
|
|
#include <net/ip6_route.h>
|
|
|
|
#include <net/ip6_checksum.h>
|
|
|
|
#include <net/inet_ecn.h>
|
|
|
|
#include <net/protocol.h>
|
|
|
|
#include <net/xfrm.h>
|
|
|
|
#include <net/snmp.h>
|
|
|
|
#include <net/dsfield.h>
|
2005-12-14 02:25:19 -05:00
|
|
|
#include <net/timewait_sock.h>
|
2007-10-27 01:53:14 -04:00
|
|
|
#include <net/netdma.h>
|
2008-04-03 17:22:32 -04:00
|
|
|
#include <net/inet_common.h>
|
2005-04-16 18:20:36 -04:00
|
|
|
|
|
|
|
#include <asm/uaccess.h>
|
|
|
|
|
|
|
|
#include <linux/proc_fs.h>
|
|
|
|
#include <linux/seq_file.h>
|
|
|
|
|
2006-11-14 22:07:45 -05:00
|
|
|
#include <linux/crypto.h>
|
|
|
|
#include <linux/scatterlist.h>
|
|
|
|
|
|
|
|
static void tcp_v6_send_reset(struct sock *sk, struct sk_buff *skb);
|
2008-08-07 02:50:04 -04:00
|
|
|
static void tcp_v6_reqsk_send_ack(struct sock *sk, struct sk_buff *skb,
|
|
|
|
struct request_sock *req);
|
2005-04-16 18:20:36 -04:00
|
|
|
|
|
|
|
static int tcp_v6_do_rcv(struct sock *sk, struct sk_buff *skb);
|
|
|
|
|
2005-12-14 02:15:52 -05:00
|
|
|
static struct inet_connection_sock_af_ops ipv6_mapped;
|
|
|
|
static struct inet_connection_sock_af_ops ipv6_specific;
|
2006-11-14 22:53:22 -05:00
|
|
|
#ifdef CONFIG_TCP_MD5SIG
|
2006-11-14 22:07:45 -05:00
|
|
|
static struct tcp_sock_af_ops tcp_sock_ipv6_specific;
|
|
|
|
static struct tcp_sock_af_ops tcp_sock_ipv6_mapped_specific;
|
2008-04-17 23:45:16 -04:00
|
|
|
#else
|
|
|
|
static struct tcp_md5sig_key *tcp_v6_md5_do_lookup(struct sock *sk,
|
|
|
|
struct in6_addr *addr)
|
|
|
|
{
|
|
|
|
return NULL;
|
|
|
|
}
|
2006-11-14 22:53:22 -05:00
|
|
|
#endif
|
2005-04-16 18:20:36 -04:00
|
|
|
|
|
|
|
static void tcp_v6_hash(struct sock *sk)
|
|
|
|
{
|
|
|
|
if (sk->sk_state != TCP_CLOSE) {
|
2005-12-14 02:15:52 -05:00
|
|
|
if (inet_csk(sk)->icsk_af_ops == &ipv6_mapped) {
|
2005-04-16 18:20:36 -04:00
|
|
|
tcp_prot.hash(sk);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
local_bh_disable();
|
[SOCK] proto: Add hashinfo member to struct proto
This way we can remove TCP and DCCP specific versions of
sk->sk_prot->get_port: both v4 and v6 use inet_csk_get_port
sk->sk_prot->hash: inet_hash is directly used, only v6 need
a specific version to deal with mapped sockets
sk->sk_prot->unhash: both v4 and v6 use inet_hash directly
struct inet_connection_sock_af_ops also gets a new member, bind_conflict, so
that inet_csk_get_port can find the per family routine.
Now only the lookup routines receive as a parameter a struct inet_hashtable.
With this we further reuse code, reducing the difference among INET transport
protocols.
Eventually work has to be done on UDP and SCTP to make them share this
infrastructure and get as a bonus inet_diag interfaces so that iproute can be
used with these protocols.
net-2.6/net/ipv4/inet_hashtables.c:
struct proto | +8
struct inet_connection_sock_af_ops | +8
2 structs changed
__inet_hash_nolisten | +18
__inet_hash | -210
inet_put_port | +8
inet_bind_bucket_create | +1
__inet_hash_connect | -8
5 functions changed, 27 bytes added, 218 bytes removed, diff: -191
net-2.6/net/core/sock.c:
proto_seq_show | +3
1 function changed, 3 bytes added, diff: +3
net-2.6/net/ipv4/inet_connection_sock.c:
inet_csk_get_port | +15
1 function changed, 15 bytes added, diff: +15
net-2.6/net/ipv4/tcp.c:
tcp_set_state | -7
1 function changed, 7 bytes removed, diff: -7
net-2.6/net/ipv4/tcp_ipv4.c:
tcp_v4_get_port | -31
tcp_v4_hash | -48
tcp_v4_destroy_sock | -7
tcp_v4_syn_recv_sock | -2
tcp_unhash | -179
5 functions changed, 267 bytes removed, diff: -267
net-2.6/net/ipv6/inet6_hashtables.c:
__inet6_hash | +8
1 function changed, 8 bytes added, diff: +8
net-2.6/net/ipv4/inet_hashtables.c:
inet_unhash | +190
inet_hash | +242
2 functions changed, 432 bytes added, diff: +432
vmlinux:
16 functions changed, 485 bytes added, 492 bytes removed, diff: -7
/home/acme/git/net-2.6/net/ipv6/tcp_ipv6.c:
tcp_v6_get_port | -31
tcp_v6_hash | -7
tcp_v6_syn_recv_sock | -9
3 functions changed, 47 bytes removed, diff: -47
/home/acme/git/net-2.6/net/dccp/proto.c:
dccp_destroy_sock | -7
dccp_unhash | -179
dccp_hash | -49
dccp_set_state | -7
dccp_done | +1
5 functions changed, 1 bytes added, 242 bytes removed, diff: -241
/home/acme/git/net-2.6/net/dccp/ipv4.c:
dccp_v4_get_port | -31
dccp_v4_request_recv_sock | -2
2 functions changed, 33 bytes removed, diff: -33
/home/acme/git/net-2.6/net/dccp/ipv6.c:
dccp_v6_get_port | -31
dccp_v6_hash | -7
dccp_v6_request_recv_sock | +5
3 functions changed, 5 bytes added, 38 bytes removed, diff: -33
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-02-03 07:06:04 -05:00
|
|
|
__inet6_hash(sk);
|
2005-04-16 18:20:36 -04:00
|
|
|
local_bh_enable();
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2009-01-08 13:41:23 -05:00
|
|
|
static __inline__ __sum16 tcp_v6_check(int len,
|
2007-02-09 09:24:49 -05:00
|
|
|
struct in6_addr *saddr,
|
|
|
|
struct in6_addr *daddr,
|
2006-11-15 00:35:48 -05:00
|
|
|
__wsum base)
|
2005-04-16 18:20:36 -04:00
|
|
|
{
|
|
|
|
return csum_ipv6_magic(saddr, daddr, len, IPPROTO_TCP, base);
|
|
|
|
}
|
|
|
|
|
2006-11-10 17:06:49 -05:00
|
|
|
static __u32 tcp_v6_init_sequence(struct sk_buff *skb)
|
2005-04-16 18:20:36 -04:00
|
|
|
{
|
2007-04-25 20:54:47 -04:00
|
|
|
return secure_tcpv6_sequence_number(ipv6_hdr(skb)->daddr.s6_addr32,
|
|
|
|
ipv6_hdr(skb)->saddr.s6_addr32,
|
2007-04-11 00:04:22 -04:00
|
|
|
tcp_hdr(skb)->dest,
|
|
|
|
tcp_hdr(skb)->source);
|
2005-04-16 18:20:36 -04:00
|
|
|
}
|
|
|
|
|
2007-02-09 09:24:49 -05:00
|
|
|
static int tcp_v6_connect(struct sock *sk, struct sockaddr *uaddr,
|
2005-04-16 18:20:36 -04:00
|
|
|
int addr_len)
|
|
|
|
{
|
|
|
|
struct sockaddr_in6 *usin = (struct sockaddr_in6 *) uaddr;
|
2007-02-09 09:24:49 -05:00
|
|
|
struct inet_sock *inet = inet_sk(sk);
|
2005-12-14 02:26:10 -05:00
|
|
|
struct inet_connection_sock *icsk = inet_csk(sk);
|
2005-04-16 18:20:36 -04:00
|
|
|
struct ipv6_pinfo *np = inet6_sk(sk);
|
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
|
|
|
struct in6_addr *saddr = NULL, *final_p = NULL, final;
|
|
|
|
struct flowi fl;
|
|
|
|
struct dst_entry *dst;
|
|
|
|
int addr_type;
|
|
|
|
int err;
|
|
|
|
|
2007-02-09 09:24:49 -05:00
|
|
|
if (addr_len < SIN6_LEN_RFC2133)
|
2005-04-16 18:20:36 -04:00
|
|
|
return -EINVAL;
|
|
|
|
|
2007-02-09 09:24:49 -05:00
|
|
|
if (usin->sin6_family != AF_INET6)
|
2005-04-16 18:20:36 -04:00
|
|
|
return(-EAFNOSUPPORT);
|
|
|
|
|
|
|
|
memset(&fl, 0, sizeof(fl));
|
|
|
|
|
|
|
|
if (np->sndflow) {
|
|
|
|
fl.fl6_flowlabel = usin->sin6_flowinfo&IPV6_FLOWINFO_MASK;
|
|
|
|
IP6_ECN_flow_init(fl.fl6_flowlabel);
|
|
|
|
if (fl.fl6_flowlabel&IPV6_FLOWLABEL_MASK) {
|
|
|
|
struct ip6_flowlabel *flowlabel;
|
|
|
|
flowlabel = fl6_sock_lookup(sk, fl.fl6_flowlabel);
|
|
|
|
if (flowlabel == NULL)
|
|
|
|
return -EINVAL;
|
|
|
|
ipv6_addr_copy(&usin->sin6_addr, &flowlabel->dst);
|
|
|
|
fl6_sock_release(flowlabel);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2007-02-09 09:24:49 -05:00
|
|
|
* connect() to INADDR_ANY means loopback (BSD'ism).
|
|
|
|
*/
|
|
|
|
|
|
|
|
if(ipv6_addr_any(&usin->sin6_addr))
|
|
|
|
usin->sin6_addr.s6_addr[15] = 0x1;
|
2005-04-16 18:20:36 -04:00
|
|
|
|
|
|
|
addr_type = ipv6_addr_type(&usin->sin6_addr);
|
|
|
|
|
|
|
|
if(addr_type & IPV6_ADDR_MULTICAST)
|
|
|
|
return -ENETUNREACH;
|
|
|
|
|
|
|
|
if (addr_type&IPV6_ADDR_LINKLOCAL) {
|
|
|
|
if (addr_len >= sizeof(struct sockaddr_in6) &&
|
|
|
|
usin->sin6_scope_id) {
|
|
|
|
/* If interface is set while binding, indices
|
|
|
|
* must coincide.
|
|
|
|
*/
|
|
|
|
if (sk->sk_bound_dev_if &&
|
|
|
|
sk->sk_bound_dev_if != usin->sin6_scope_id)
|
|
|
|
return -EINVAL;
|
|
|
|
|
|
|
|
sk->sk_bound_dev_if = usin->sin6_scope_id;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Connect to link-local address requires an interface */
|
|
|
|
if (!sk->sk_bound_dev_if)
|
|
|
|
return -EINVAL;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (tp->rx_opt.ts_recent_stamp &&
|
|
|
|
!ipv6_addr_equal(&np->daddr, &usin->sin6_addr)) {
|
|
|
|
tp->rx_opt.ts_recent = 0;
|
|
|
|
tp->rx_opt.ts_recent_stamp = 0;
|
|
|
|
tp->write_seq = 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
ipv6_addr_copy(&np->daddr, &usin->sin6_addr);
|
|
|
|
np->flow_label = fl.fl6_flowlabel;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* TCP over IPv4
|
|
|
|
*/
|
|
|
|
|
|
|
|
if (addr_type == IPV6_ADDR_MAPPED) {
|
2005-12-14 02:26:10 -05:00
|
|
|
u32 exthdrlen = icsk->icsk_ext_hdr_len;
|
2005-04-16 18:20:36 -04:00
|
|
|
struct sockaddr_in sin;
|
|
|
|
|
|
|
|
SOCK_DEBUG(sk, "connect: ipv4 mapped\n");
|
|
|
|
|
|
|
|
if (__ipv6_only_sock(sk))
|
|
|
|
return -ENETUNREACH;
|
|
|
|
|
|
|
|
sin.sin_family = AF_INET;
|
|
|
|
sin.sin_port = usin->sin6_port;
|
|
|
|
sin.sin_addr.s_addr = usin->sin6_addr.s6_addr32[3];
|
|
|
|
|
2005-12-14 02:26:10 -05:00
|
|
|
icsk->icsk_af_ops = &ipv6_mapped;
|
2005-04-16 18:20:36 -04:00
|
|
|
sk->sk_backlog_rcv = tcp_v4_do_rcv;
|
2006-11-14 22:07:45 -05:00
|
|
|
#ifdef CONFIG_TCP_MD5SIG
|
|
|
|
tp->af_specific = &tcp_sock_ipv6_mapped_specific;
|
|
|
|
#endif
|
2005-04-16 18:20:36 -04:00
|
|
|
|
|
|
|
err = tcp_v4_connect(sk, (struct sockaddr *)&sin, sizeof(sin));
|
|
|
|
|
|
|
|
if (err) {
|
2005-12-14 02:26:10 -05:00
|
|
|
icsk->icsk_ext_hdr_len = exthdrlen;
|
|
|
|
icsk->icsk_af_ops = &ipv6_specific;
|
2005-04-16 18:20:36 -04:00
|
|
|
sk->sk_backlog_rcv = tcp_v6_do_rcv;
|
2006-11-14 22:07:45 -05:00
|
|
|
#ifdef CONFIG_TCP_MD5SIG
|
|
|
|
tp->af_specific = &tcp_sock_ipv6_specific;
|
|
|
|
#endif
|
2005-04-16 18:20:36 -04:00
|
|
|
goto failure;
|
|
|
|
} else {
|
|
|
|
ipv6_addr_set(&np->saddr, 0, 0, htonl(0x0000FFFF),
|
|
|
|
inet->saddr);
|
|
|
|
ipv6_addr_set(&np->rcv_saddr, 0, 0, htonl(0x0000FFFF),
|
|
|
|
inet->rcv_saddr);
|
|
|
|
}
|
|
|
|
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!ipv6_addr_any(&np->rcv_saddr))
|
|
|
|
saddr = &np->rcv_saddr;
|
|
|
|
|
|
|
|
fl.proto = IPPROTO_TCP;
|
|
|
|
ipv6_addr_copy(&fl.fl6_dst, &np->daddr);
|
|
|
|
ipv6_addr_copy(&fl.fl6_src,
|
|
|
|
(saddr ? saddr : &np->saddr));
|
|
|
|
fl.oif = sk->sk_bound_dev_if;
|
|
|
|
fl.fl_ip_dport = usin->sin6_port;
|
|
|
|
fl.fl_ip_sport = inet->sport;
|
|
|
|
|
|
|
|
if (np->opt && np->opt->srcrt) {
|
|
|
|
struct rt0_hdr *rt0 = (struct rt0_hdr *)np->opt->srcrt;
|
|
|
|
ipv6_addr_copy(&final, &fl.fl6_dst);
|
|
|
|
ipv6_addr_copy(&fl.fl6_dst, rt0->addr);
|
|
|
|
final_p = &final;
|
|
|
|
}
|
|
|
|
|
2006-08-05 02:12:42 -04:00
|
|
|
security_sk_classify_flow(sk, &fl);
|
|
|
|
|
2005-04-16 18:20:36 -04:00
|
|
|
err = ip6_dst_lookup(sk, &dst, &fl);
|
|
|
|
if (err)
|
|
|
|
goto failure;
|
|
|
|
if (final_p)
|
|
|
|
ipv6_addr_copy(&fl.fl6_dst, final_p);
|
|
|
|
|
2008-11-25 20:35:18 -05:00
|
|
|
err = __xfrm_lookup(sock_net(sk), &dst, &fl, sk, XFRM_LOOKUP_WAIT);
|
|
|
|
if (err < 0) {
|
2007-05-24 21:17:54 -04:00
|
|
|
if (err == -EREMOTE)
|
|
|
|
err = ip6_dst_blackhole(sk, &dst, &fl);
|
|
|
|
if (err < 0)
|
|
|
|
goto failure;
|
|
|
|
}
|
2005-04-16 18:20:36 -04:00
|
|
|
|
|
|
|
if (saddr == NULL) {
|
|
|
|
saddr = &fl.fl6_src;
|
|
|
|
ipv6_addr_copy(&np->rcv_saddr, saddr);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* set the source address */
|
|
|
|
ipv6_addr_copy(&np->saddr, saddr);
|
|
|
|
inet->rcv_saddr = LOOPBACK4_IPV6;
|
|
|
|
|
2006-06-30 16:37:03 -04:00
|
|
|
sk->sk_gso_type = SKB_GSO_TCPV6;
|
2006-08-29 20:15:09 -04:00
|
|
|
__ip6_dst_store(sk, dst, NULL, NULL);
|
2005-04-16 18:20:36 -04:00
|
|
|
|
2005-12-14 02:26:10 -05:00
|
|
|
icsk->icsk_ext_hdr_len = 0;
|
2005-04-16 18:20:36 -04:00
|
|
|
if (np->opt)
|
2005-12-14 02:26:10 -05:00
|
|
|
icsk->icsk_ext_hdr_len = (np->opt->opt_flen +
|
|
|
|
np->opt->opt_nflen);
|
2005-04-16 18:20:36 -04:00
|
|
|
|
|
|
|
tp->rx_opt.mss_clamp = IPV6_MIN_MTU - sizeof(struct tcphdr) - sizeof(struct ipv6hdr);
|
|
|
|
|
|
|
|
inet->dport = usin->sin6_port;
|
|
|
|
|
|
|
|
tcp_set_state(sk, TCP_SYN_SENT);
|
2005-12-14 02:25:44 -05:00
|
|
|
err = inet6_hash_connect(&tcp_death_row, sk);
|
2005-04-16 18:20:36 -04:00
|
|
|
if (err)
|
|
|
|
goto late_failure;
|
|
|
|
|
|
|
|
if (!tp->write_seq)
|
|
|
|
tp->write_seq = secure_tcpv6_sequence_number(np->saddr.s6_addr32,
|
|
|
|
np->daddr.s6_addr32,
|
|
|
|
inet->sport,
|
|
|
|
inet->dport);
|
|
|
|
|
|
|
|
err = tcp_connect(sk);
|
|
|
|
if (err)
|
|
|
|
goto late_failure;
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
late_failure:
|
|
|
|
tcp_set_state(sk, TCP_CLOSE);
|
|
|
|
__sk_dst_reset(sk);
|
|
|
|
failure:
|
|
|
|
inet->dport = 0;
|
|
|
|
sk->sk_route_caps = 0;
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void tcp_v6_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
|
2006-11-08 03:21:01 -05:00
|
|
|
int type, int code, int offset, __be32 info)
|
2005-04-16 18:20:36 -04:00
|
|
|
{
|
|
|
|
struct ipv6hdr *hdr = (struct ipv6hdr*)skb->data;
|
2005-08-12 08:19:38 -04:00
|
|
|
const struct tcphdr *th = (struct tcphdr *)(skb->data+offset);
|
2005-04-16 18:20:36 -04:00
|
|
|
struct ipv6_pinfo *np;
|
|
|
|
struct sock *sk;
|
|
|
|
int err;
|
2007-02-09 09:24:49 -05:00
|
|
|
struct tcp_sock *tp;
|
2005-04-16 18:20:36 -04:00
|
|
|
__u32 seq;
|
2008-07-16 23:28:42 -04:00
|
|
|
struct net *net = dev_net(skb->dev);
|
2005-04-16 18:20:36 -04:00
|
|
|
|
2008-07-16 23:28:42 -04:00
|
|
|
sk = inet6_lookup(net, &tcp_hashinfo, &hdr->daddr,
|
2008-01-31 08:07:21 -05:00
|
|
|
th->dest, &hdr->saddr, th->source, skb->dev->ifindex);
|
2005-04-16 18:20:36 -04:00
|
|
|
|
|
|
|
if (sk == NULL) {
|
2008-10-08 13:33:26 -04:00
|
|
|
ICMP6_INC_STATS_BH(net, __in6_dev_get(skb->dev),
|
|
|
|
ICMP6_MIB_INERRORS);
|
2005-04-16 18:20:36 -04:00
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (sk->sk_state == TCP_TIME_WAIT) {
|
2006-10-10 22:41:46 -04:00
|
|
|
inet_twsk_put(inet_twsk(sk));
|
2005-04-16 18:20:36 -04:00
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
bh_lock_sock(sk);
|
|
|
|
if (sock_owned_by_user(sk))
|
2008-07-16 23:31:16 -04:00
|
|
|
NET_INC_STATS_BH(net, LINUX_MIB_LOCKDROPPEDICMPS);
|
2005-04-16 18:20:36 -04:00
|
|
|
|
|
|
|
if (sk->sk_state == TCP_CLOSE)
|
|
|
|
goto out;
|
|
|
|
|
|
|
|
tp = tcp_sk(sk);
|
2007-02-09 09:24:49 -05:00
|
|
|
seq = ntohl(th->seq);
|
2005-04-16 18:20:36 -04:00
|
|
|
if (sk->sk_state != TCP_LISTEN &&
|
|
|
|
!between(seq, tp->snd_una, tp->snd_nxt)) {
|
2008-07-16 23:31:16 -04:00
|
|
|
NET_INC_STATS_BH(net, LINUX_MIB_OUTOFWINDOWICMPS);
|
2005-04-16 18:20:36 -04:00
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
|
|
|
np = inet6_sk(sk);
|
|
|
|
|
|
|
|
if (type == ICMPV6_PKT_TOOBIG) {
|
|
|
|
struct dst_entry *dst = NULL;
|
|
|
|
|
|
|
|
if (sock_owned_by_user(sk))
|
|
|
|
goto out;
|
|
|
|
if ((1 << sk->sk_state) & (TCPF_LISTEN | TCPF_CLOSE))
|
|
|
|
goto out;
|
|
|
|
|
|
|
|
/* icmp should have updated the destination cache entry */
|
|
|
|
dst = __sk_dst_check(sk, np->dst_cookie);
|
|
|
|
|
|
|
|
if (dst == NULL) {
|
|
|
|
struct inet_sock *inet = inet_sk(sk);
|
|
|
|
struct flowi fl;
|
|
|
|
|
|
|
|
/* BUGGG_FUTURE: Again, it is not clear how
|
|
|
|
to handle rthdr case. Ignore this complexity
|
|
|
|
for now.
|
|
|
|
*/
|
|
|
|
memset(&fl, 0, sizeof(fl));
|
|
|
|
fl.proto = IPPROTO_TCP;
|
|
|
|
ipv6_addr_copy(&fl.fl6_dst, &np->daddr);
|
|
|
|
ipv6_addr_copy(&fl.fl6_src, &np->saddr);
|
|
|
|
fl.oif = sk->sk_bound_dev_if;
|
|
|
|
fl.fl_ip_dport = inet->dport;
|
|
|
|
fl.fl_ip_sport = inet->sport;
|
2006-08-05 02:12:42 -04:00
|
|
|
security_skb_classify_flow(skb, &fl);
|
2005-04-16 18:20:36 -04:00
|
|
|
|
|
|
|
if ((err = ip6_dst_lookup(sk, &dst, &fl))) {
|
|
|
|
sk->sk_err_soft = -err;
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
2008-11-25 20:35:18 -05:00
|
|
|
if ((err = xfrm_lookup(net, &dst, &fl, sk, 0)) < 0) {
|
2005-04-16 18:20:36 -04:00
|
|
|
sk->sk_err_soft = -err;
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
|
|
|
} else
|
|
|
|
dst_hold(dst);
|
|
|
|
|
2005-12-14 02:26:10 -05:00
|
|
|
if (inet_csk(sk)->icsk_pmtu_cookie > dst_mtu(dst)) {
|
2005-04-16 18:20:36 -04:00
|
|
|
tcp_sync_mss(sk, dst_mtu(dst));
|
|
|
|
tcp_simple_retransmit(sk);
|
|
|
|
} /* else let the usual retransmit timer handle it */
|
|
|
|
dst_release(dst);
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
|
|
|
icmpv6_err_convert(type, code, &err);
|
|
|
|
|
2005-06-19 01:47:21 -04:00
|
|
|
/* Might be for an request_sock */
|
2005-04-16 18:20:36 -04:00
|
|
|
switch (sk->sk_state) {
|
2005-06-19 01:47:21 -04:00
|
|
|
struct request_sock *req, **prev;
|
2005-04-16 18:20:36 -04:00
|
|
|
case TCP_LISTEN:
|
|
|
|
if (sock_owned_by_user(sk))
|
|
|
|
goto out;
|
|
|
|
|
2005-12-14 02:15:24 -05:00
|
|
|
req = inet6_csk_search_req(sk, &prev, th->dest, &hdr->daddr,
|
|
|
|
&hdr->saddr, inet6_iif(skb));
|
2005-04-16 18:20:36 -04:00
|
|
|
if (!req)
|
|
|
|
goto out;
|
|
|
|
|
|
|
|
/* ICMPs are not backlogged, hence we cannot get
|
|
|
|
* an established socket here.
|
|
|
|
*/
|
2008-07-26 00:43:18 -04:00
|
|
|
WARN_ON(req->sk != NULL);
|
2005-04-16 18:20:36 -04:00
|
|
|
|
[NET] Generalise TCP's struct open_request minisock infrastructure
Kept this first changeset minimal, without changing existing names to
ease peer review.
Basicaly tcp_openreq_alloc now receives the or_calltable, that in turn
has two new members:
->slab, that replaces tcp_openreq_cachep
->obj_size, to inform the size of the openreq descendant for
a specific protocol
The protocol specific fields in struct open_request were moved to a
class hierarchy, with the things that are common to all connection
oriented PF_INET protocols in struct inet_request_sock, the TCP ones
in tcp_request_sock, that is an inet_request_sock, that is an
open_request.
I.e. this uses the same approach used for the struct sock class
hierarchy, with sk_prot indicating if the protocol wants to use the
open_request infrastructure by filling in sk_prot->rsk_prot with an
or_calltable.
Results? Performance is improved and TCP v4 now uses only 64 bytes per
open request minisock, down from 96 without this patch :-)
Next changeset will rename some of the structs, fields and functions
mentioned above, struct or_calltable is way unclear, better name it
struct request_sock_ops, s/struct open_request/struct request_sock/g,
etc.
Signed-off-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-06-19 01:46:52 -04:00
|
|
|
if (seq != tcp_rsk(req)->snt_isn) {
|
2008-07-16 23:31:16 -04:00
|
|
|
NET_INC_STATS_BH(net, LINUX_MIB_OUTOFWINDOWICMPS);
|
2005-04-16 18:20:36 -04:00
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
2005-08-09 23:10:42 -04:00
|
|
|
inet_csk_reqsk_queue_drop(sk, req, prev);
|
2005-04-16 18:20:36 -04:00
|
|
|
goto out;
|
|
|
|
|
|
|
|
case TCP_SYN_SENT:
|
|
|
|
case TCP_SYN_RECV: /* Cannot happen.
|
2007-02-09 09:24:49 -05:00
|
|
|
It can, it SYNs are crossed. --ANK */
|
2005-04-16 18:20:36 -04:00
|
|
|
if (!sock_owned_by_user(sk)) {
|
|
|
|
sk->sk_err = err;
|
|
|
|
sk->sk_error_report(sk); /* Wake people up to see the error (see connect in sock.c) */
|
|
|
|
|
|
|
|
tcp_done(sk);
|
|
|
|
} else
|
|
|
|
sk->sk_err_soft = err;
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!sock_owned_by_user(sk) && np->recverr) {
|
|
|
|
sk->sk_err = err;
|
|
|
|
sk->sk_error_report(sk);
|
|
|
|
} else
|
|
|
|
sk->sk_err_soft = err;
|
|
|
|
|
|
|
|
out:
|
|
|
|
bh_unlock_sock(sk);
|
|
|
|
sock_put(sk);
|
|
|
|
}
|
|
|
|
|
|
|
|
|
2008-02-29 14:43:03 -05:00
|
|
|
static int tcp_v6_send_synack(struct sock *sk, struct request_sock *req)
|
2005-04-16 18:20:36 -04:00
|
|
|
{
|
2005-12-14 02:15:40 -05:00
|
|
|
struct inet6_request_sock *treq = inet6_rsk(req);
|
2005-04-16 18:20:36 -04:00
|
|
|
struct ipv6_pinfo *np = inet6_sk(sk);
|
|
|
|
struct sk_buff * skb;
|
|
|
|
struct ipv6_txoptions *opt = NULL;
|
|
|
|
struct in6_addr * final_p = NULL, final;
|
|
|
|
struct flowi fl;
|
2008-02-29 14:43:03 -05:00
|
|
|
struct dst_entry *dst;
|
2005-04-16 18:20:36 -04:00
|
|
|
int err = -1;
|
|
|
|
|
|
|
|
memset(&fl, 0, sizeof(fl));
|
|
|
|
fl.proto = IPPROTO_TCP;
|
[NET] Generalise TCP's struct open_request minisock infrastructure
Kept this first changeset minimal, without changing existing names to
ease peer review.
Basicaly tcp_openreq_alloc now receives the or_calltable, that in turn
has two new members:
->slab, that replaces tcp_openreq_cachep
->obj_size, to inform the size of the openreq descendant for
a specific protocol
The protocol specific fields in struct open_request were moved to a
class hierarchy, with the things that are common to all connection
oriented PF_INET protocols in struct inet_request_sock, the TCP ones
in tcp_request_sock, that is an inet_request_sock, that is an
open_request.
I.e. this uses the same approach used for the struct sock class
hierarchy, with sk_prot indicating if the protocol wants to use the
open_request infrastructure by filling in sk_prot->rsk_prot with an
or_calltable.
Results? Performance is improved and TCP v4 now uses only 64 bytes per
open request minisock, down from 96 without this patch :-)
Next changeset will rename some of the structs, fields and functions
mentioned above, struct or_calltable is way unclear, better name it
struct request_sock_ops, s/struct open_request/struct request_sock/g,
etc.
Signed-off-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-06-19 01:46:52 -04:00
|
|
|
ipv6_addr_copy(&fl.fl6_dst, &treq->rmt_addr);
|
|
|
|
ipv6_addr_copy(&fl.fl6_src, &treq->loc_addr);
|
2005-04-16 18:20:36 -04:00
|
|
|
fl.fl6_flowlabel = 0;
|
[NET] Generalise TCP's struct open_request minisock infrastructure
Kept this first changeset minimal, without changing existing names to
ease peer review.
Basicaly tcp_openreq_alloc now receives the or_calltable, that in turn
has two new members:
->slab, that replaces tcp_openreq_cachep
->obj_size, to inform the size of the openreq descendant for
a specific protocol
The protocol specific fields in struct open_request were moved to a
class hierarchy, with the things that are common to all connection
oriented PF_INET protocols in struct inet_request_sock, the TCP ones
in tcp_request_sock, that is an inet_request_sock, that is an
open_request.
I.e. this uses the same approach used for the struct sock class
hierarchy, with sk_prot indicating if the protocol wants to use the
open_request infrastructure by filling in sk_prot->rsk_prot with an
or_calltable.
Results? Performance is improved and TCP v4 now uses only 64 bytes per
open request minisock, down from 96 without this patch :-)
Next changeset will rename some of the structs, fields and functions
mentioned above, struct or_calltable is way unclear, better name it
struct request_sock_ops, s/struct open_request/struct request_sock/g,
etc.
Signed-off-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-06-19 01:46:52 -04:00
|
|
|
fl.oif = treq->iif;
|
|
|
|
fl.fl_ip_dport = inet_rsk(req)->rmt_port;
|
2008-10-20 02:35:58 -04:00
|
|
|
fl.fl_ip_sport = inet_rsk(req)->loc_port;
|
2006-07-25 02:32:50 -04:00
|
|
|
security_req_classify_flow(req, &fl);
|
2005-04-16 18:20:36 -04:00
|
|
|
|
2008-02-29 14:43:03 -05:00
|
|
|
opt = np->opt;
|
|
|
|
if (opt && opt->srcrt) {
|
|
|
|
struct rt0_hdr *rt0 = (struct rt0_hdr *) opt->srcrt;
|
|
|
|
ipv6_addr_copy(&final, &fl.fl6_dst);
|
|
|
|
ipv6_addr_copy(&fl.fl6_dst, rt0->addr);
|
|
|
|
final_p = &final;
|
2005-04-16 18:20:36 -04:00
|
|
|
}
|
|
|
|
|
2008-02-29 14:43:03 -05:00
|
|
|
err = ip6_dst_lookup(sk, &dst, &fl);
|
|
|
|
if (err)
|
|
|
|
goto done;
|
|
|
|
if (final_p)
|
|
|
|
ipv6_addr_copy(&fl.fl6_dst, final_p);
|
2008-11-25 20:35:18 -05:00
|
|
|
if ((err = xfrm_lookup(sock_net(sk), &dst, &fl, sk, 0)) < 0)
|
2008-02-29 14:43:03 -05:00
|
|
|
goto done;
|
|
|
|
|
2005-04-16 18:20:36 -04:00
|
|
|
skb = tcp_make_synack(sk, dst, req);
|
|
|
|
if (skb) {
|
2007-04-11 00:04:22 -04:00
|
|
|
struct tcphdr *th = tcp_hdr(skb);
|
2005-04-16 18:20:36 -04:00
|
|
|
|
2009-01-08 13:41:23 -05:00
|
|
|
th->check = tcp_v6_check(skb->len,
|
[NET] Generalise TCP's struct open_request minisock infrastructure
Kept this first changeset minimal, without changing existing names to
ease peer review.
Basicaly tcp_openreq_alloc now receives the or_calltable, that in turn
has two new members:
->slab, that replaces tcp_openreq_cachep
->obj_size, to inform the size of the openreq descendant for
a specific protocol
The protocol specific fields in struct open_request were moved to a
class hierarchy, with the things that are common to all connection
oriented PF_INET protocols in struct inet_request_sock, the TCP ones
in tcp_request_sock, that is an inet_request_sock, that is an
open_request.
I.e. this uses the same approach used for the struct sock class
hierarchy, with sk_prot indicating if the protocol wants to use the
open_request infrastructure by filling in sk_prot->rsk_prot with an
or_calltable.
Results? Performance is improved and TCP v4 now uses only 64 bytes per
open request minisock, down from 96 without this patch :-)
Next changeset will rename some of the structs, fields and functions
mentioned above, struct or_calltable is way unclear, better name it
struct request_sock_ops, s/struct open_request/struct request_sock/g,
etc.
Signed-off-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-06-19 01:46:52 -04:00
|
|
|
&treq->loc_addr, &treq->rmt_addr,
|
2008-11-19 18:44:53 -05:00
|
|
|
csum_partial(th, skb->len, skb->csum));
|
2005-04-16 18:20:36 -04:00
|
|
|
|
[NET] Generalise TCP's struct open_request minisock infrastructure
Kept this first changeset minimal, without changing existing names to
ease peer review.
Basicaly tcp_openreq_alloc now receives the or_calltable, that in turn
has two new members:
->slab, that replaces tcp_openreq_cachep
->obj_size, to inform the size of the openreq descendant for
a specific protocol
The protocol specific fields in struct open_request were moved to a
class hierarchy, with the things that are common to all connection
oriented PF_INET protocols in struct inet_request_sock, the TCP ones
in tcp_request_sock, that is an inet_request_sock, that is an
open_request.
I.e. this uses the same approach used for the struct sock class
hierarchy, with sk_prot indicating if the protocol wants to use the
open_request infrastructure by filling in sk_prot->rsk_prot with an
or_calltable.
Results? Performance is improved and TCP v4 now uses only 64 bytes per
open request minisock, down from 96 without this patch :-)
Next changeset will rename some of the structs, fields and functions
mentioned above, struct or_calltable is way unclear, better name it
struct request_sock_ops, s/struct open_request/struct request_sock/g,
etc.
Signed-off-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-06-19 01:46:52 -04:00
|
|
|
ipv6_addr_copy(&fl.fl6_dst, &treq->rmt_addr);
|
2005-04-16 18:20:36 -04:00
|
|
|
err = ip6_xmit(sk, skb, &fl, opt, 0);
|
2006-11-14 08:21:36 -05:00
|
|
|
err = net_xmit_eval(err);
|
2005-04-16 18:20:36 -04:00
|
|
|
}
|
|
|
|
|
|
|
|
done:
|
2007-02-09 09:24:49 -05:00
|
|
|
if (opt && opt != np->opt)
|
2005-04-16 18:20:36 -04:00
|
|
|
sock_kfree_s(sk, opt, opt->tot_len);
|
2006-01-31 20:51:44 -05:00
|
|
|
dst_release(dst);
|
2005-04-16 18:20:36 -04:00
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
2008-02-08 00:49:26 -05:00
|
|
|
static inline void syn_flood_warning(struct sk_buff *skb)
|
|
|
|
{
|
|
|
|
#ifdef CONFIG_SYN_COOKIES
|
|
|
|
if (sysctl_tcp_syncookies)
|
|
|
|
printk(KERN_INFO
|
|
|
|
"TCPv6: Possible SYN flooding on port %d. "
|
|
|
|
"Sending cookies.\n", ntohs(tcp_hdr(skb)->dest));
|
|
|
|
else
|
|
|
|
#endif
|
|
|
|
printk(KERN_INFO
|
|
|
|
"TCPv6: Possible SYN flooding on port %d. "
|
|
|
|
"Dropping request.\n", ntohs(tcp_hdr(skb)->dest));
|
|
|
|
}
|
|
|
|
|
2005-06-19 01:47:21 -04:00
|
|
|
static void tcp_v6_reqsk_destructor(struct request_sock *req)
|
2005-04-16 18:20:36 -04:00
|
|
|
{
|
2009-02-23 16:45:33 -05:00
|
|
|
kfree_skb(inet6_rsk(req)->pktopts);
|
2005-04-16 18:20:36 -04:00
|
|
|
}
|
|
|
|
|
2006-11-14 22:07:45 -05:00
|
|
|
#ifdef CONFIG_TCP_MD5SIG
|
|
|
|
static struct tcp_md5sig_key *tcp_v6_md5_do_lookup(struct sock *sk,
|
|
|
|
struct in6_addr *addr)
|
|
|
|
{
|
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
|
|
|
int i;
|
|
|
|
|
|
|
|
BUG_ON(tp == NULL);
|
|
|
|
|
|
|
|
if (!tp->md5sig_info || !tp->md5sig_info->entries6)
|
|
|
|
return NULL;
|
|
|
|
|
|
|
|
for (i = 0; i < tp->md5sig_info->entries6; i++) {
|
2008-04-10 02:42:07 -04:00
|
|
|
if (ipv6_addr_equal(&tp->md5sig_info->keys6[i].addr, addr))
|
2007-09-28 18:18:35 -04:00
|
|
|
return &tp->md5sig_info->keys6[i].base;
|
2006-11-14 22:07:45 -05:00
|
|
|
}
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
static struct tcp_md5sig_key *tcp_v6_md5_lookup(struct sock *sk,
|
|
|
|
struct sock *addr_sk)
|
|
|
|
{
|
|
|
|
return tcp_v6_md5_do_lookup(sk, &inet6_sk(addr_sk)->daddr);
|
|
|
|
}
|
|
|
|
|
|
|
|
static struct tcp_md5sig_key *tcp_v6_reqsk_md5_lookup(struct sock *sk,
|
|
|
|
struct request_sock *req)
|
|
|
|
{
|
|
|
|
return tcp_v6_md5_do_lookup(sk, &inet6_rsk(req)->rmt_addr);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int tcp_v6_md5_do_add(struct sock *sk, struct in6_addr *peer,
|
|
|
|
char *newkey, u8 newkeylen)
|
|
|
|
{
|
|
|
|
/* Add key to the list */
|
2007-10-29 23:55:27 -04:00
|
|
|
struct tcp_md5sig_key *key;
|
2006-11-14 22:07:45 -05:00
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
|
|
|
struct tcp6_md5sig_key *keys;
|
|
|
|
|
2007-10-29 23:55:27 -04:00
|
|
|
key = tcp_v6_md5_do_lookup(sk, peer);
|
2006-11-14 22:07:45 -05:00
|
|
|
if (key) {
|
|
|
|
/* modify existing entry - just update that one */
|
2007-10-29 23:55:27 -04:00
|
|
|
kfree(key->key);
|
|
|
|
key->key = newkey;
|
|
|
|
key->keylen = newkeylen;
|
2006-11-14 22:07:45 -05:00
|
|
|
} else {
|
|
|
|
/* reallocate new list if current one is full. */
|
|
|
|
if (!tp->md5sig_info) {
|
|
|
|
tp->md5sig_info = kzalloc(sizeof(*tp->md5sig_info), GFP_ATOMIC);
|
|
|
|
if (!tp->md5sig_info) {
|
|
|
|
kfree(newkey);
|
|
|
|
return -ENOMEM;
|
|
|
|
}
|
2007-06-12 17:36:42 -04:00
|
|
|
sk->sk_route_caps &= ~NETIF_F_GSO_MASK;
|
2006-11-14 22:07:45 -05:00
|
|
|
}
|
2007-11-20 20:30:56 -05:00
|
|
|
if (tcp_alloc_md5sig_pool() == NULL) {
|
|
|
|
kfree(newkey);
|
|
|
|
return -ENOMEM;
|
|
|
|
}
|
2006-11-14 22:07:45 -05:00
|
|
|
if (tp->md5sig_info->alloced6 == tp->md5sig_info->entries6) {
|
|
|
|
keys = kmalloc((sizeof (tp->md5sig_info->keys6[0]) *
|
|
|
|
(tp->md5sig_info->entries6 + 1)), GFP_ATOMIC);
|
|
|
|
|
|
|
|
if (!keys) {
|
|
|
|
tcp_free_md5sig_pool();
|
|
|
|
kfree(newkey);
|
|
|
|
return -ENOMEM;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (tp->md5sig_info->entries6)
|
|
|
|
memmove(keys, tp->md5sig_info->keys6,
|
|
|
|
(sizeof (tp->md5sig_info->keys6[0]) *
|
|
|
|
tp->md5sig_info->entries6));
|
|
|
|
|
|
|
|
kfree(tp->md5sig_info->keys6);
|
|
|
|
tp->md5sig_info->keys6 = keys;
|
|
|
|
tp->md5sig_info->alloced6++;
|
|
|
|
}
|
|
|
|
|
|
|
|
ipv6_addr_copy(&tp->md5sig_info->keys6[tp->md5sig_info->entries6].addr,
|
|
|
|
peer);
|
2007-09-28 18:18:35 -04:00
|
|
|
tp->md5sig_info->keys6[tp->md5sig_info->entries6].base.key = newkey;
|
|
|
|
tp->md5sig_info->keys6[tp->md5sig_info->entries6].base.keylen = newkeylen;
|
2006-11-14 22:07:45 -05:00
|
|
|
|
|
|
|
tp->md5sig_info->entries6++;
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int tcp_v6_md5_add_func(struct sock *sk, struct sock *addr_sk,
|
|
|
|
u8 *newkey, __u8 newkeylen)
|
|
|
|
{
|
|
|
|
return tcp_v6_md5_do_add(sk, &inet6_sk(addr_sk)->daddr,
|
|
|
|
newkey, newkeylen);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int tcp_v6_md5_do_del(struct sock *sk, struct in6_addr *peer)
|
|
|
|
{
|
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
|
|
|
int i;
|
|
|
|
|
|
|
|
for (i = 0; i < tp->md5sig_info->entries6; i++) {
|
2008-04-10 02:42:07 -04:00
|
|
|
if (ipv6_addr_equal(&tp->md5sig_info->keys6[i].addr, peer)) {
|
2006-11-14 22:07:45 -05:00
|
|
|
/* Free the key */
|
2007-09-28 18:18:35 -04:00
|
|
|
kfree(tp->md5sig_info->keys6[i].base.key);
|
2006-11-14 22:07:45 -05:00
|
|
|
tp->md5sig_info->entries6--;
|
|
|
|
|
|
|
|
if (tp->md5sig_info->entries6 == 0) {
|
|
|
|
kfree(tp->md5sig_info->keys6);
|
|
|
|
tp->md5sig_info->keys6 = NULL;
|
2007-07-24 18:27:30 -04:00
|
|
|
tp->md5sig_info->alloced6 = 0;
|
2006-11-14 22:07:45 -05:00
|
|
|
} else {
|
|
|
|
/* shrink the database */
|
|
|
|
if (tp->md5sig_info->entries6 != i)
|
|
|
|
memmove(&tp->md5sig_info->keys6[i],
|
|
|
|
&tp->md5sig_info->keys6[i+1],
|
|
|
|
(tp->md5sig_info->entries6 - i)
|
|
|
|
* sizeof (tp->md5sig_info->keys6[0]));
|
|
|
|
}
|
2007-11-20 20:31:23 -05:00
|
|
|
tcp_free_md5sig_pool();
|
|
|
|
return 0;
|
2006-11-14 22:07:45 -05:00
|
|
|
}
|
|
|
|
}
|
|
|
|
return -ENOENT;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void tcp_v6_clear_md5_list (struct sock *sk)
|
|
|
|
{
|
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
|
|
|
int i;
|
|
|
|
|
|
|
|
if (tp->md5sig_info->entries6) {
|
|
|
|
for (i = 0; i < tp->md5sig_info->entries6; i++)
|
2007-09-28 18:18:35 -04:00
|
|
|
kfree(tp->md5sig_info->keys6[i].base.key);
|
2006-11-14 22:07:45 -05:00
|
|
|
tp->md5sig_info->entries6 = 0;
|
|
|
|
tcp_free_md5sig_pool();
|
|
|
|
}
|
|
|
|
|
|
|
|
kfree(tp->md5sig_info->keys6);
|
|
|
|
tp->md5sig_info->keys6 = NULL;
|
|
|
|
tp->md5sig_info->alloced6 = 0;
|
|
|
|
|
|
|
|
if (tp->md5sig_info->entries4) {
|
|
|
|
for (i = 0; i < tp->md5sig_info->entries4; i++)
|
2007-09-28 18:18:35 -04:00
|
|
|
kfree(tp->md5sig_info->keys4[i].base.key);
|
2006-11-14 22:07:45 -05:00
|
|
|
tp->md5sig_info->entries4 = 0;
|
|
|
|
tcp_free_md5sig_pool();
|
|
|
|
}
|
|
|
|
|
|
|
|
kfree(tp->md5sig_info->keys4);
|
|
|
|
tp->md5sig_info->keys4 = NULL;
|
|
|
|
tp->md5sig_info->alloced4 = 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int tcp_v6_parse_md5_keys (struct sock *sk, char __user *optval,
|
|
|
|
int optlen)
|
|
|
|
{
|
|
|
|
struct tcp_md5sig cmd;
|
|
|
|
struct sockaddr_in6 *sin6 = (struct sockaddr_in6 *)&cmd.tcpm_addr;
|
|
|
|
u8 *newkey;
|
|
|
|
|
|
|
|
if (optlen < sizeof(cmd))
|
|
|
|
return -EINVAL;
|
|
|
|
|
|
|
|
if (copy_from_user(&cmd, optval, sizeof(cmd)))
|
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
if (sin6->sin6_family != AF_INET6)
|
|
|
|
return -EINVAL;
|
|
|
|
|
|
|
|
if (!cmd.tcpm_keylen) {
|
|
|
|
if (!tcp_sk(sk)->md5sig_info)
|
|
|
|
return -ENOENT;
|
2007-08-25 02:16:08 -04:00
|
|
|
if (ipv6_addr_v4mapped(&sin6->sin6_addr))
|
2006-11-14 22:07:45 -05:00
|
|
|
return tcp_v4_md5_do_del(sk, sin6->sin6_addr.s6_addr32[3]);
|
|
|
|
return tcp_v6_md5_do_del(sk, &sin6->sin6_addr);
|
|
|
|
}
|
|
|
|
|
|
|
|
if (cmd.tcpm_keylen > TCP_MD5SIG_MAXKEYLEN)
|
|
|
|
return -EINVAL;
|
|
|
|
|
|
|
|
if (!tcp_sk(sk)->md5sig_info) {
|
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
|
|
|
struct tcp_md5sig_info *p;
|
|
|
|
|
|
|
|
p = kzalloc(sizeof(struct tcp_md5sig_info), GFP_KERNEL);
|
|
|
|
if (!p)
|
|
|
|
return -ENOMEM;
|
|
|
|
|
|
|
|
tp->md5sig_info = p;
|
2007-06-12 17:36:42 -04:00
|
|
|
sk->sk_route_caps &= ~NETIF_F_GSO_MASK;
|
2006-11-14 22:07:45 -05:00
|
|
|
}
|
|
|
|
|
2006-11-17 09:14:37 -05:00
|
|
|
newkey = kmemdup(cmd.tcpm_key, cmd.tcpm_keylen, GFP_KERNEL);
|
2006-11-14 22:07:45 -05:00
|
|
|
if (!newkey)
|
|
|
|
return -ENOMEM;
|
2007-08-25 02:16:08 -04:00
|
|
|
if (ipv6_addr_v4mapped(&sin6->sin6_addr)) {
|
2006-11-14 22:07:45 -05:00
|
|
|
return tcp_v4_md5_do_add(sk, sin6->sin6_addr.s6_addr32[3],
|
|
|
|
newkey, cmd.tcpm_keylen);
|
|
|
|
}
|
|
|
|
return tcp_v6_md5_do_add(sk, &sin6->sin6_addr, newkey, cmd.tcpm_keylen);
|
|
|
|
}
|
|
|
|
|
2008-07-19 03:01:42 -04:00
|
|
|
static int tcp_v6_md5_hash_pseudoheader(struct tcp_md5sig_pool *hp,
|
|
|
|
struct in6_addr *daddr,
|
|
|
|
struct in6_addr *saddr, int nbytes)
|
2006-11-14 22:07:45 -05:00
|
|
|
{
|
|
|
|
struct tcp6_pseudohdr *bp;
|
2008-07-19 03:01:42 -04:00
|
|
|
struct scatterlist sg;
|
2008-04-17 00:19:16 -04:00
|
|
|
|
2006-11-14 22:07:45 -05:00
|
|
|
bp = &hp->md5_blk.ip6;
|
|
|
|
/* 1. TCP pseudo-header (RFC2460) */
|
|
|
|
ipv6_addr_copy(&bp->saddr, saddr);
|
|
|
|
ipv6_addr_copy(&bp->daddr, daddr);
|
2008-07-19 03:01:42 -04:00
|
|
|
bp->protocol = cpu_to_be32(IPPROTO_TCP);
|
2008-08-01 00:36:07 -04:00
|
|
|
bp->len = cpu_to_be32(nbytes);
|
2006-11-14 22:07:45 -05:00
|
|
|
|
2008-07-19 03:01:42 -04:00
|
|
|
sg_init_one(&sg, bp, sizeof(*bp));
|
|
|
|
return crypto_hash_update(&hp->md5_desc, &sg, sizeof(*bp));
|
|
|
|
}
|
2007-10-26 03:41:21 -04:00
|
|
|
|
2008-07-19 03:01:42 -04:00
|
|
|
static int tcp_v6_md5_hash_hdr(char *md5_hash, struct tcp_md5sig_key *key,
|
|
|
|
struct in6_addr *daddr, struct in6_addr *saddr,
|
|
|
|
struct tcphdr *th)
|
|
|
|
{
|
|
|
|
struct tcp_md5sig_pool *hp;
|
|
|
|
struct hash_desc *desc;
|
|
|
|
|
|
|
|
hp = tcp_get_md5sig_pool();
|
|
|
|
if (!hp)
|
|
|
|
goto clear_hash_noput;
|
|
|
|
desc = &hp->md5_desc;
|
|
|
|
|
|
|
|
if (crypto_hash_init(desc))
|
|
|
|
goto clear_hash;
|
|
|
|
if (tcp_v6_md5_hash_pseudoheader(hp, daddr, saddr, th->doff << 2))
|
|
|
|
goto clear_hash;
|
|
|
|
if (tcp_md5_hash_header(hp, th))
|
|
|
|
goto clear_hash;
|
|
|
|
if (tcp_md5_hash_key(hp, key))
|
|
|
|
goto clear_hash;
|
|
|
|
if (crypto_hash_final(desc, md5_hash))
|
2006-11-14 22:07:45 -05:00
|
|
|
goto clear_hash;
|
|
|
|
|
|
|
|
tcp_put_md5sig_pool();
|
|
|
|
return 0;
|
2008-07-19 03:01:42 -04:00
|
|
|
|
2006-11-14 22:07:45 -05:00
|
|
|
clear_hash:
|
|
|
|
tcp_put_md5sig_pool();
|
|
|
|
clear_hash_noput:
|
|
|
|
memset(md5_hash, 0, 16);
|
2008-07-19 03:01:42 -04:00
|
|
|
return 1;
|
2006-11-14 22:07:45 -05:00
|
|
|
}
|
|
|
|
|
2008-07-19 03:01:42 -04:00
|
|
|
static int tcp_v6_md5_hash_skb(char *md5_hash, struct tcp_md5sig_key *key,
|
|
|
|
struct sock *sk, struct request_sock *req,
|
|
|
|
struct sk_buff *skb)
|
2006-11-14 22:07:45 -05:00
|
|
|
{
|
|
|
|
struct in6_addr *saddr, *daddr;
|
2008-07-19 03:01:42 -04:00
|
|
|
struct tcp_md5sig_pool *hp;
|
|
|
|
struct hash_desc *desc;
|
|
|
|
struct tcphdr *th = tcp_hdr(skb);
|
2006-11-14 22:07:45 -05:00
|
|
|
|
|
|
|
if (sk) {
|
|
|
|
saddr = &inet6_sk(sk)->saddr;
|
|
|
|
daddr = &inet6_sk(sk)->daddr;
|
2008-07-19 03:01:42 -04:00
|
|
|
} else if (req) {
|
2006-11-14 22:07:45 -05:00
|
|
|
saddr = &inet6_rsk(req)->loc_addr;
|
|
|
|
daddr = &inet6_rsk(req)->rmt_addr;
|
2008-07-19 03:01:42 -04:00
|
|
|
} else {
|
|
|
|
struct ipv6hdr *ip6h = ipv6_hdr(skb);
|
|
|
|
saddr = &ip6h->saddr;
|
|
|
|
daddr = &ip6h->daddr;
|
2006-11-14 22:07:45 -05:00
|
|
|
}
|
2008-07-19 03:01:42 -04:00
|
|
|
|
|
|
|
hp = tcp_get_md5sig_pool();
|
|
|
|
if (!hp)
|
|
|
|
goto clear_hash_noput;
|
|
|
|
desc = &hp->md5_desc;
|
|
|
|
|
|
|
|
if (crypto_hash_init(desc))
|
|
|
|
goto clear_hash;
|
|
|
|
|
|
|
|
if (tcp_v6_md5_hash_pseudoheader(hp, daddr, saddr, skb->len))
|
|
|
|
goto clear_hash;
|
|
|
|
if (tcp_md5_hash_header(hp, th))
|
|
|
|
goto clear_hash;
|
|
|
|
if (tcp_md5_hash_skb_data(hp, skb, th->doff << 2))
|
|
|
|
goto clear_hash;
|
|
|
|
if (tcp_md5_hash_key(hp, key))
|
|
|
|
goto clear_hash;
|
|
|
|
if (crypto_hash_final(desc, md5_hash))
|
|
|
|
goto clear_hash;
|
|
|
|
|
|
|
|
tcp_put_md5sig_pool();
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
clear_hash:
|
|
|
|
tcp_put_md5sig_pool();
|
|
|
|
clear_hash_noput:
|
|
|
|
memset(md5_hash, 0, 16);
|
|
|
|
return 1;
|
2006-11-14 22:07:45 -05:00
|
|
|
}
|
|
|
|
|
|
|
|
static int tcp_v6_inbound_md5_hash (struct sock *sk, struct sk_buff *skb)
|
|
|
|
{
|
|
|
|
__u8 *hash_location = NULL;
|
|
|
|
struct tcp_md5sig_key *hash_expected;
|
2007-04-25 20:54:47 -04:00
|
|
|
struct ipv6hdr *ip6h = ipv6_hdr(skb);
|
2007-04-11 00:04:22 -04:00
|
|
|
struct tcphdr *th = tcp_hdr(skb);
|
2006-11-14 22:07:45 -05:00
|
|
|
int genhash;
|
|
|
|
u8 newhash[16];
|
|
|
|
|
|
|
|
hash_expected = tcp_v6_md5_do_lookup(sk, &ip6h->saddr);
|
2008-04-16 23:29:53 -04:00
|
|
|
hash_location = tcp_parse_md5sig_option(th);
|
2006-11-14 22:07:45 -05:00
|
|
|
|
2008-07-30 06:03:15 -04:00
|
|
|
/* We've parsed the options - do we have a hash? */
|
|
|
|
if (!hash_expected && !hash_location)
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
if (hash_expected && !hash_location) {
|
|
|
|
NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_TCPMD5NOTFOUND);
|
2006-11-14 22:07:45 -05:00
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
|
2008-07-30 06:03:15 -04:00
|
|
|
if (!hash_expected && hash_location) {
|
|
|
|
NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_TCPMD5UNEXPECTED);
|
2006-11-14 22:07:45 -05:00
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* check the signature */
|
2008-07-19 03:01:42 -04:00
|
|
|
genhash = tcp_v6_md5_hash_skb(newhash,
|
|
|
|
hash_expected,
|
|
|
|
NULL, NULL, skb);
|
|
|
|
|
2006-11-14 22:07:45 -05:00
|
|
|
if (genhash || memcmp(hash_location, newhash, 16) != 0) {
|
|
|
|
if (net_ratelimit()) {
|
2008-10-29 15:52:50 -04:00
|
|
|
printk(KERN_INFO "MD5 Hash %s for (%pI6, %u)->(%pI6, %u)\n",
|
2006-11-14 22:07:45 -05:00
|
|
|
genhash ? "failed" : "mismatch",
|
2008-10-28 19:09:23 -04:00
|
|
|
&ip6h->saddr, ntohs(th->source),
|
|
|
|
&ip6h->daddr, ntohs(th->dest));
|
2006-11-14 22:07:45 -05:00
|
|
|
}
|
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
2008-02-08 00:49:26 -05:00
|
|
|
struct request_sock_ops tcp6_request_sock_ops __read_mostly = {
|
2005-04-16 18:20:36 -04:00
|
|
|
.family = AF_INET6,
|
[NET] Generalise TCP's struct open_request minisock infrastructure
Kept this first changeset minimal, without changing existing names to
ease peer review.
Basicaly tcp_openreq_alloc now receives the or_calltable, that in turn
has two new members:
->slab, that replaces tcp_openreq_cachep
->obj_size, to inform the size of the openreq descendant for
a specific protocol
The protocol specific fields in struct open_request were moved to a
class hierarchy, with the things that are common to all connection
oriented PF_INET protocols in struct inet_request_sock, the TCP ones
in tcp_request_sock, that is an inet_request_sock, that is an
open_request.
I.e. this uses the same approach used for the struct sock class
hierarchy, with sk_prot indicating if the protocol wants to use the
open_request infrastructure by filling in sk_prot->rsk_prot with an
or_calltable.
Results? Performance is improved and TCP v4 now uses only 64 bytes per
open request minisock, down from 96 without this patch :-)
Next changeset will rename some of the structs, fields and functions
mentioned above, struct or_calltable is way unclear, better name it
struct request_sock_ops, s/struct open_request/struct request_sock/g,
etc.
Signed-off-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-06-19 01:46:52 -04:00
|
|
|
.obj_size = sizeof(struct tcp6_request_sock),
|
2005-04-16 18:20:36 -04:00
|
|
|
.rtx_syn_ack = tcp_v6_send_synack,
|
2005-06-19 01:47:21 -04:00
|
|
|
.send_ack = tcp_v6_reqsk_send_ack,
|
|
|
|
.destructor = tcp_v6_reqsk_destructor,
|
2005-04-16 18:20:36 -04:00
|
|
|
.send_reset = tcp_v6_send_reset
|
|
|
|
};
|
|
|
|
|
2006-11-14 22:07:45 -05:00
|
|
|
#ifdef CONFIG_TCP_MD5SIG
|
2006-11-30 22:16:28 -05:00
|
|
|
static struct tcp_request_sock_ops tcp_request_sock_ipv6_ops = {
|
2006-11-14 22:07:45 -05:00
|
|
|
.md5_lookup = tcp_v6_reqsk_md5_lookup,
|
|
|
|
};
|
2006-11-30 22:16:28 -05:00
|
|
|
#endif
|
2006-11-14 22:07:45 -05:00
|
|
|
|
2005-12-14 02:25:19 -05:00
|
|
|
static struct timewait_sock_ops tcp6_timewait_sock_ops = {
|
|
|
|
.twsk_obj_size = sizeof(struct tcp6_timewait_sock),
|
|
|
|
.twsk_unique = tcp_twsk_unique,
|
2006-11-14 22:07:45 -05:00
|
|
|
.twsk_destructor= tcp_twsk_destructor,
|
2005-12-14 02:25:19 -05:00
|
|
|
};
|
|
|
|
|
2005-12-14 02:15:52 -05:00
|
|
|
static void tcp_v6_send_check(struct sock *sk, int len, struct sk_buff *skb)
|
2005-04-16 18:20:36 -04:00
|
|
|
{
|
|
|
|
struct ipv6_pinfo *np = inet6_sk(sk);
|
2007-04-11 00:04:22 -04:00
|
|
|
struct tcphdr *th = tcp_hdr(skb);
|
2005-04-16 18:20:36 -04:00
|
|
|
|
2006-08-29 19:44:56 -04:00
|
|
|
if (skb->ip_summed == CHECKSUM_PARTIAL) {
|
2005-04-16 18:20:36 -04:00
|
|
|
th->check = ~csum_ipv6_magic(&np->saddr, &np->daddr, len, IPPROTO_TCP, 0);
|
2007-04-09 14:59:07 -04:00
|
|
|
skb->csum_start = skb_transport_header(skb) - skb->head;
|
2006-11-20 21:07:29 -05:00
|
|
|
skb->csum_offset = offsetof(struct tcphdr, check);
|
2005-04-16 18:20:36 -04:00
|
|
|
} else {
|
2007-02-09 09:24:49 -05:00
|
|
|
th->check = csum_ipv6_magic(&np->saddr, &np->daddr, len, IPPROTO_TCP,
|
2008-11-19 18:44:53 -05:00
|
|
|
csum_partial(th, th->doff<<2,
|
2005-04-16 18:20:36 -04:00
|
|
|
skb->csum));
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2006-07-08 16:34:56 -04:00
|
|
|
static int tcp_v6_gso_send_check(struct sk_buff *skb)
|
|
|
|
{
|
|
|
|
struct ipv6hdr *ipv6h;
|
|
|
|
struct tcphdr *th;
|
|
|
|
|
|
|
|
if (!pskb_may_pull(skb, sizeof(*th)))
|
|
|
|
return -EINVAL;
|
|
|
|
|
2007-04-25 20:54:47 -04:00
|
|
|
ipv6h = ipv6_hdr(skb);
|
2007-04-11 00:04:22 -04:00
|
|
|
th = tcp_hdr(skb);
|
2006-07-08 16:34:56 -04:00
|
|
|
|
|
|
|
th->check = 0;
|
|
|
|
th->check = ~csum_ipv6_magic(&ipv6h->saddr, &ipv6h->daddr, skb->len,
|
|
|
|
IPPROTO_TCP, 0);
|
2007-04-09 14:59:07 -04:00
|
|
|
skb->csum_start = skb_transport_header(skb) - skb->head;
|
2006-11-20 21:07:29 -05:00
|
|
|
skb->csum_offset = offsetof(struct tcphdr, check);
|
2006-08-29 19:44:56 -04:00
|
|
|
skb->ip_summed = CHECKSUM_PARTIAL;
|
2006-07-08 16:34:56 -04:00
|
|
|
return 0;
|
|
|
|
}
|
2005-04-16 18:20:36 -04:00
|
|
|
|
2009-01-08 13:41:23 -05:00
|
|
|
struct sk_buff **tcp6_gro_receive(struct sk_buff **head, struct sk_buff *skb)
|
|
|
|
{
|
2009-04-27 08:44:45 -04:00
|
|
|
struct ipv6hdr *iph = skb_gro_network_header(skb);
|
2009-01-08 13:41:23 -05:00
|
|
|
|
|
|
|
switch (skb->ip_summed) {
|
|
|
|
case CHECKSUM_COMPLETE:
|
2009-01-29 09:19:50 -05:00
|
|
|
if (!tcp_v6_check(skb_gro_len(skb), &iph->saddr, &iph->daddr,
|
2009-01-08 13:41:23 -05:00
|
|
|
skb->csum)) {
|
|
|
|
skb->ip_summed = CHECKSUM_UNNECESSARY;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* fall through */
|
|
|
|
case CHECKSUM_NONE:
|
|
|
|
NAPI_GRO_CB(skb)->flush = 1;
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
return tcp_gro_receive(head, skb);
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(tcp6_gro_receive);
|
|
|
|
|
|
|
|
int tcp6_gro_complete(struct sk_buff *skb)
|
|
|
|
{
|
|
|
|
struct ipv6hdr *iph = ipv6_hdr(skb);
|
|
|
|
struct tcphdr *th = tcp_hdr(skb);
|
|
|
|
|
|
|
|
th->check = ~tcp_v6_check(skb->len - skb_transport_offset(skb),
|
|
|
|
&iph->saddr, &iph->daddr, 0);
|
|
|
|
skb_shinfo(skb)->gso_type = SKB_GSO_TCPV6;
|
|
|
|
|
|
|
|
return tcp_gro_complete(skb);
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(tcp6_gro_complete);
|
|
|
|
|
2008-10-09 17:42:40 -04:00
|
|
|
static void tcp_v6_send_response(struct sk_buff *skb, u32 seq, u32 ack, u32 win,
|
|
|
|
u32 ts, struct tcp_md5sig_key *key, int rst)
|
2005-04-16 18:20:36 -04:00
|
|
|
{
|
2007-04-11 00:04:22 -04:00
|
|
|
struct tcphdr *th = tcp_hdr(skb), *t1;
|
2005-04-16 18:20:36 -04:00
|
|
|
struct sk_buff *buff;
|
|
|
|
struct flowi fl;
|
2008-03-25 08:47:49 -04:00
|
|
|
struct net *net = dev_net(skb->dst->dev);
|
2008-03-07 14:16:26 -05:00
|
|
|
struct sock *ctl_sk = net->ipv6.tcp_sk;
|
2008-10-09 17:41:38 -04:00
|
|
|
unsigned int tot_len = sizeof(struct tcphdr);
|
tcpv6: convert opt[] -> topt in tcp_v6_send_reset
after this I get:
$ diff-funcs tcp_v6_send_reset tcp_ipv6.c tcp_ipv6.c tcp_v6_send_ack
--- tcp_ipv6.c:tcp_v6_send_reset()
+++ tcp_ipv6.c:tcp_v6_send_ack()
@@ -1,4 +1,5 @@
-static void tcp_v6_send_reset(struct sock *sk, struct sk_buff *skb)
+static void tcp_v6_send_ack(struct sk_buff *skb, u32 seq, u32 ack, u32 win,
u32 ts,
+ struct tcp_md5sig_key *key)
{
struct tcphdr *th = tcp_hdr(skb), *t1;
struct sk_buff *buff;
@@ -7,31 +8,14 @@
struct sock *ctl_sk = net->ipv6.tcp_sk;
unsigned int tot_len = sizeof(struct tcphdr);
__be32 *topt;
-#ifdef CONFIG_TCP_MD5SIG
- struct tcp_md5sig_key *key;
-#endif
-
- if (th->rst)
- return;
-
- if (!ipv6_unicast_destination(skb))
- return;
+ if (ts)
+ tot_len += TCPOLEN_TSTAMP_ALIGNED;
#ifdef CONFIG_TCP_MD5SIG
- if (sk)
- key = tcp_v6_md5_do_lookup(sk, &ipv6_hdr(skb)->daddr);
- else
- key = NULL;
-
if (key)
tot_len += TCPOLEN_MD5SIG_ALIGNED;
#endif
- /*
- * We need to grab some memory, and put together an RST,
- * and then put it into the queue to be sent.
- */
-
buff = alloc_skb(MAX_HEADER + sizeof(struct ipv6hdr) + tot_len,
GFP_ATOMIC);
if (buff == NULL)
@@ -46,18 +30,20 @@
t1->dest = th->source;
t1->source = th->dest;
t1->doff = tot_len / 4;
- t1->rst = 1;
-
- if(th->ack) {
- t1->seq = th->ack_seq;
- } else {
- t1->ack = 1;
- t1->ack_seq = htonl(ntohl(th->seq) + th->syn + th->fin
- + skb->len - (th->doff<<2));
- }
+ t1->seq = htonl(seq);
+ t1->ack_seq = htonl(ack);
+ t1->ack = 1;
+ t1->window = htons(win);
topt = (__be32 *)(t1 + 1);
+ if (ts) {
+ *topt++ = htonl((TCPOPT_NOP << 24) | (TCPOPT_NOP << 16) |
+ (TCPOPT_TIMESTAMP << 8) |
TCPOLEN_TIMESTAMP);
+ *topt++ = htonl(tcp_time_stamp);
+ *topt++ = htonl(ts);
+ }
+
#ifdef CONFIG_TCP_MD5SIG
if (key) {
*topt++ = htonl((TCPOPT_NOP << 24) | (TCPOPT_NOP << 16) |
@@ -84,15 +70,10 @@
fl.fl_ip_sport = t1->source;
security_skb_classify_flow(skb, &fl);
- /* Pass a socket to ip6_dst_lookup either it is for RST
- * Underlying function will use this to retrieve the network
- * namespace
- */
if (!ip6_dst_lookup(ctl_sk, &buff->dst, &fl)) {
if (xfrm_lookup(&buff->dst, &fl, NULL, 0) >= 0) {
ip6_xmit(ctl_sk, buff, &fl, NULL, 0);
TCP_INC_STATS_BH(net, TCP_MIB_OUTSEGS);
- TCP_INC_STATS_BH(net, TCP_MIB_OUTRSTS);
return;
}
}
...which starts to be trivial to combine.
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-10-09 17:42:01 -04:00
|
|
|
__be32 *topt;
|
2005-04-16 18:20:36 -04:00
|
|
|
|
2008-10-09 17:42:40 -04:00
|
|
|
if (ts)
|
|
|
|
tot_len += TCPOLEN_TSTAMP_ALIGNED;
|
2006-11-14 22:07:45 -05:00
|
|
|
#ifdef CONFIG_TCP_MD5SIG
|
|
|
|
if (key)
|
|
|
|
tot_len += TCPOLEN_MD5SIG_ALIGNED;
|
|
|
|
#endif
|
|
|
|
|
|
|
|
buff = alloc_skb(MAX_HEADER + sizeof(struct ipv6hdr) + tot_len,
|
2005-04-16 18:20:36 -04:00
|
|
|
GFP_ATOMIC);
|
2007-02-09 09:24:49 -05:00
|
|
|
if (buff == NULL)
|
|
|
|
return;
|
2005-04-16 18:20:36 -04:00
|
|
|
|
2006-11-14 22:07:45 -05:00
|
|
|
skb_reserve(buff, MAX_HEADER + sizeof(struct ipv6hdr) + tot_len);
|
2005-04-16 18:20:36 -04:00
|
|
|
|
2006-11-14 22:07:45 -05:00
|
|
|
t1 = (struct tcphdr *) skb_push(buff, tot_len);
|
2005-04-16 18:20:36 -04:00
|
|
|
|
|
|
|
/* Swap the send and the receive. */
|
|
|
|
memset(t1, 0, sizeof(*t1));
|
|
|
|
t1->dest = th->source;
|
|
|
|
t1->source = th->dest;
|
2006-11-14 22:07:45 -05:00
|
|
|
t1->doff = tot_len / 4;
|
2008-10-09 17:42:40 -04:00
|
|
|
t1->seq = htonl(seq);
|
|
|
|
t1->ack_seq = htonl(ack);
|
|
|
|
t1->ack = !rst || !th->ack;
|
|
|
|
t1->rst = rst;
|
|
|
|
t1->window = htons(win);
|
2005-04-16 18:20:36 -04:00
|
|
|
|
tcpv6: convert opt[] -> topt in tcp_v6_send_reset
after this I get:
$ diff-funcs tcp_v6_send_reset tcp_ipv6.c tcp_ipv6.c tcp_v6_send_ack
--- tcp_ipv6.c:tcp_v6_send_reset()
+++ tcp_ipv6.c:tcp_v6_send_ack()
@@ -1,4 +1,5 @@
-static void tcp_v6_send_reset(struct sock *sk, struct sk_buff *skb)
+static void tcp_v6_send_ack(struct sk_buff *skb, u32 seq, u32 ack, u32 win,
u32 ts,
+ struct tcp_md5sig_key *key)
{
struct tcphdr *th = tcp_hdr(skb), *t1;
struct sk_buff *buff;
@@ -7,31 +8,14 @@
struct sock *ctl_sk = net->ipv6.tcp_sk;
unsigned int tot_len = sizeof(struct tcphdr);
__be32 *topt;
-#ifdef CONFIG_TCP_MD5SIG
- struct tcp_md5sig_key *key;
-#endif
-
- if (th->rst)
- return;
-
- if (!ipv6_unicast_destination(skb))
- return;
+ if (ts)
+ tot_len += TCPOLEN_TSTAMP_ALIGNED;
#ifdef CONFIG_TCP_MD5SIG
- if (sk)
- key = tcp_v6_md5_do_lookup(sk, &ipv6_hdr(skb)->daddr);
- else
- key = NULL;
-
if (key)
tot_len += TCPOLEN_MD5SIG_ALIGNED;
#endif
- /*
- * We need to grab some memory, and put together an RST,
- * and then put it into the queue to be sent.
- */
-
buff = alloc_skb(MAX_HEADER + sizeof(struct ipv6hdr) + tot_len,
GFP_ATOMIC);
if (buff == NULL)
@@ -46,18 +30,20 @@
t1->dest = th->source;
t1->source = th->dest;
t1->doff = tot_len / 4;
- t1->rst = 1;
-
- if(th->ack) {
- t1->seq = th->ack_seq;
- } else {
- t1->ack = 1;
- t1->ack_seq = htonl(ntohl(th->seq) + th->syn + th->fin
- + skb->len - (th->doff<<2));
- }
+ t1->seq = htonl(seq);
+ t1->ack_seq = htonl(ack);
+ t1->ack = 1;
+ t1->window = htons(win);
topt = (__be32 *)(t1 + 1);
+ if (ts) {
+ *topt++ = htonl((TCPOPT_NOP << 24) | (TCPOPT_NOP << 16) |
+ (TCPOPT_TIMESTAMP << 8) |
TCPOLEN_TIMESTAMP);
+ *topt++ = htonl(tcp_time_stamp);
+ *topt++ = htonl(ts);
+ }
+
#ifdef CONFIG_TCP_MD5SIG
if (key) {
*topt++ = htonl((TCPOPT_NOP << 24) | (TCPOPT_NOP << 16) |
@@ -84,15 +70,10 @@
fl.fl_ip_sport = t1->source;
security_skb_classify_flow(skb, &fl);
- /* Pass a socket to ip6_dst_lookup either it is for RST
- * Underlying function will use this to retrieve the network
- * namespace
- */
if (!ip6_dst_lookup(ctl_sk, &buff->dst, &fl)) {
if (xfrm_lookup(&buff->dst, &fl, NULL, 0) >= 0) {
ip6_xmit(ctl_sk, buff, &fl, NULL, 0);
TCP_INC_STATS_BH(net, TCP_MIB_OUTSEGS);
- TCP_INC_STATS_BH(net, TCP_MIB_OUTRSTS);
return;
}
}
...which starts to be trivial to combine.
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-10-09 17:42:01 -04:00
|
|
|
topt = (__be32 *)(t1 + 1);
|
|
|
|
|
2008-10-09 17:42:40 -04:00
|
|
|
if (ts) {
|
|
|
|
*topt++ = htonl((TCPOPT_NOP << 24) | (TCPOPT_NOP << 16) |
|
|
|
|
(TCPOPT_TIMESTAMP << 8) | TCPOLEN_TIMESTAMP);
|
|
|
|
*topt++ = htonl(tcp_time_stamp);
|
|
|
|
*topt++ = htonl(ts);
|
|
|
|
}
|
|
|
|
|
2006-11-14 22:07:45 -05:00
|
|
|
#ifdef CONFIG_TCP_MD5SIG
|
|
|
|
if (key) {
|
tcpv6: convert opt[] -> topt in tcp_v6_send_reset
after this I get:
$ diff-funcs tcp_v6_send_reset tcp_ipv6.c tcp_ipv6.c tcp_v6_send_ack
--- tcp_ipv6.c:tcp_v6_send_reset()
+++ tcp_ipv6.c:tcp_v6_send_ack()
@@ -1,4 +1,5 @@
-static void tcp_v6_send_reset(struct sock *sk, struct sk_buff *skb)
+static void tcp_v6_send_ack(struct sk_buff *skb, u32 seq, u32 ack, u32 win,
u32 ts,
+ struct tcp_md5sig_key *key)
{
struct tcphdr *th = tcp_hdr(skb), *t1;
struct sk_buff *buff;
@@ -7,31 +8,14 @@
struct sock *ctl_sk = net->ipv6.tcp_sk;
unsigned int tot_len = sizeof(struct tcphdr);
__be32 *topt;
-#ifdef CONFIG_TCP_MD5SIG
- struct tcp_md5sig_key *key;
-#endif
-
- if (th->rst)
- return;
-
- if (!ipv6_unicast_destination(skb))
- return;
+ if (ts)
+ tot_len += TCPOLEN_TSTAMP_ALIGNED;
#ifdef CONFIG_TCP_MD5SIG
- if (sk)
- key = tcp_v6_md5_do_lookup(sk, &ipv6_hdr(skb)->daddr);
- else
- key = NULL;
-
if (key)
tot_len += TCPOLEN_MD5SIG_ALIGNED;
#endif
- /*
- * We need to grab some memory, and put together an RST,
- * and then put it into the queue to be sent.
- */
-
buff = alloc_skb(MAX_HEADER + sizeof(struct ipv6hdr) + tot_len,
GFP_ATOMIC);
if (buff == NULL)
@@ -46,18 +30,20 @@
t1->dest = th->source;
t1->source = th->dest;
t1->doff = tot_len / 4;
- t1->rst = 1;
-
- if(th->ack) {
- t1->seq = th->ack_seq;
- } else {
- t1->ack = 1;
- t1->ack_seq = htonl(ntohl(th->seq) + th->syn + th->fin
- + skb->len - (th->doff<<2));
- }
+ t1->seq = htonl(seq);
+ t1->ack_seq = htonl(ack);
+ t1->ack = 1;
+ t1->window = htons(win);
topt = (__be32 *)(t1 + 1);
+ if (ts) {
+ *topt++ = htonl((TCPOPT_NOP << 24) | (TCPOPT_NOP << 16) |
+ (TCPOPT_TIMESTAMP << 8) |
TCPOLEN_TIMESTAMP);
+ *topt++ = htonl(tcp_time_stamp);
+ *topt++ = htonl(ts);
+ }
+
#ifdef CONFIG_TCP_MD5SIG
if (key) {
*topt++ = htonl((TCPOPT_NOP << 24) | (TCPOPT_NOP << 16) |
@@ -84,15 +70,10 @@
fl.fl_ip_sport = t1->source;
security_skb_classify_flow(skb, &fl);
- /* Pass a socket to ip6_dst_lookup either it is for RST
- * Underlying function will use this to retrieve the network
- * namespace
- */
if (!ip6_dst_lookup(ctl_sk, &buff->dst, &fl)) {
if (xfrm_lookup(&buff->dst, &fl, NULL, 0) >= 0) {
ip6_xmit(ctl_sk, buff, &fl, NULL, 0);
TCP_INC_STATS_BH(net, TCP_MIB_OUTSEGS);
- TCP_INC_STATS_BH(net, TCP_MIB_OUTRSTS);
return;
}
}
...which starts to be trivial to combine.
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-10-09 17:42:01 -04:00
|
|
|
*topt++ = htonl((TCPOPT_NOP << 24) | (TCPOPT_NOP << 16) |
|
|
|
|
(TCPOPT_MD5SIG << 8) | TCPOLEN_MD5SIG);
|
|
|
|
tcp_v6_md5_hash_hdr((__u8 *)topt, key,
|
2008-10-09 17:37:47 -04:00
|
|
|
&ipv6_hdr(skb)->saddr,
|
|
|
|
&ipv6_hdr(skb)->daddr, t1);
|
2006-11-14 22:07:45 -05:00
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
2008-11-19 18:44:53 -05:00
|
|
|
buff->csum = csum_partial(t1, tot_len, 0);
|
2005-04-16 18:20:36 -04:00
|
|
|
|
|
|
|
memset(&fl, 0, sizeof(fl));
|
2007-04-25 20:54:47 -04:00
|
|
|
ipv6_addr_copy(&fl.fl6_dst, &ipv6_hdr(skb)->saddr);
|
|
|
|
ipv6_addr_copy(&fl.fl6_src, &ipv6_hdr(skb)->daddr);
|
2005-04-16 18:20:36 -04:00
|
|
|
|
|
|
|
t1->check = csum_ipv6_magic(&fl.fl6_src, &fl.fl6_dst,
|
2008-10-08 14:34:06 -04:00
|
|
|
tot_len, IPPROTO_TCP,
|
2005-04-16 18:20:36 -04:00
|
|
|
buff->csum);
|
|
|
|
|
|
|
|
fl.proto = IPPROTO_TCP;
|
2005-08-12 08:19:38 -04:00
|
|
|
fl.oif = inet6_iif(skb);
|
2005-04-16 18:20:36 -04:00
|
|
|
fl.fl_ip_dport = t1->dest;
|
|
|
|
fl.fl_ip_sport = t1->source;
|
2006-08-05 02:12:42 -04:00
|
|
|
security_skb_classify_flow(skb, &fl);
|
2005-04-16 18:20:36 -04:00
|
|
|
|
2008-03-05 13:48:35 -05:00
|
|
|
/* Pass a socket to ip6_dst_lookup either it is for RST
|
|
|
|
* Underlying function will use this to retrieve the network
|
|
|
|
* namespace
|
|
|
|
*/
|
2008-03-07 14:16:26 -05:00
|
|
|
if (!ip6_dst_lookup(ctl_sk, &buff->dst, &fl)) {
|
2008-11-25 20:35:18 -05:00
|
|
|
if (xfrm_lookup(net, &buff->dst, &fl, NULL, 0) >= 0) {
|
2008-03-07 14:16:26 -05:00
|
|
|
ip6_xmit(ctl_sk, buff, &fl, NULL, 0);
|
2008-07-16 23:22:25 -04:00
|
|
|
TCP_INC_STATS_BH(net, TCP_MIB_OUTSEGS);
|
2008-10-09 17:42:40 -04:00
|
|
|
if (rst)
|
|
|
|
TCP_INC_STATS_BH(net, TCP_MIB_OUTRSTS);
|
2005-04-16 18:20:36 -04:00
|
|
|
return;
|
2005-12-12 17:38:10 -05:00
|
|
|
}
|
2005-04-16 18:20:36 -04:00
|
|
|
}
|
|
|
|
|
|
|
|
kfree_skb(buff);
|
|
|
|
}
|
|
|
|
|
2008-10-09 17:42:40 -04:00
|
|
|
static void tcp_v6_send_reset(struct sock *sk, struct sk_buff *skb)
|
2005-04-16 18:20:36 -04:00
|
|
|
{
|
2008-10-09 17:42:40 -04:00
|
|
|
struct tcphdr *th = tcp_hdr(skb);
|
|
|
|
u32 seq = 0, ack_seq = 0;
|
2008-10-10 00:11:56 -04:00
|
|
|
struct tcp_md5sig_key *key = NULL;
|
2005-04-16 18:20:36 -04:00
|
|
|
|
2008-10-09 17:42:40 -04:00
|
|
|
if (th->rst)
|
2005-04-16 18:20:36 -04:00
|
|
|
return;
|
|
|
|
|
2008-10-09 17:42:40 -04:00
|
|
|
if (!ipv6_unicast_destination(skb))
|
|
|
|
return;
|
2005-04-16 18:20:36 -04:00
|
|
|
|
2006-11-14 22:07:45 -05:00
|
|
|
#ifdef CONFIG_TCP_MD5SIG
|
2008-10-09 17:42:40 -04:00
|
|
|
if (sk)
|
|
|
|
key = tcp_v6_md5_do_lookup(sk, &ipv6_hdr(skb)->daddr);
|
2006-11-14 22:07:45 -05:00
|
|
|
#endif
|
|
|
|
|
2008-10-09 17:42:40 -04:00
|
|
|
if (th->ack)
|
|
|
|
seq = ntohl(th->ack_seq);
|
|
|
|
else
|
|
|
|
ack_seq = ntohl(th->seq) + th->syn + th->fin + skb->len -
|
|
|
|
(th->doff << 2);
|
2005-04-16 18:20:36 -04:00
|
|
|
|
2008-10-09 17:42:40 -04:00
|
|
|
tcp_v6_send_response(skb, seq, ack_seq, 0, 0, key, 1);
|
|
|
|
}
|
2005-04-16 18:20:36 -04:00
|
|
|
|
2008-10-09 17:42:40 -04:00
|
|
|
static void tcp_v6_send_ack(struct sk_buff *skb, u32 seq, u32 ack, u32 win, u32 ts,
|
|
|
|
struct tcp_md5sig_key *key)
|
|
|
|
{
|
|
|
|
tcp_v6_send_response(skb, seq, ack, win, ts, key, 0);
|
2005-04-16 18:20:36 -04:00
|
|
|
}
|
|
|
|
|
|
|
|
static void tcp_v6_timewait_ack(struct sock *sk, struct sk_buff *skb)
|
|
|
|
{
|
2005-08-09 23:09:30 -04:00
|
|
|
struct inet_timewait_sock *tw = inet_twsk(sk);
|
2006-11-14 22:07:45 -05:00
|
|
|
struct tcp_timewait_sock *tcptw = tcp_twsk(sk);
|
2005-04-16 18:20:36 -04:00
|
|
|
|
2008-04-17 23:45:16 -04:00
|
|
|
tcp_v6_send_ack(skb, tcptw->tw_snd_nxt, tcptw->tw_rcv_nxt,
|
2005-08-09 23:09:30 -04:00
|
|
|
tcptw->tw_rcv_wnd >> tw->tw_rcv_wscale,
|
2008-04-17 23:45:16 -04:00
|
|
|
tcptw->tw_ts_recent, tcp_twsk_md5_key(tcptw));
|
2005-04-16 18:20:36 -04:00
|
|
|
|
2005-08-09 23:09:30 -04:00
|
|
|
inet_twsk_put(tw);
|
2005-04-16 18:20:36 -04:00
|
|
|
}
|
|
|
|
|
2008-08-07 02:50:04 -04:00
|
|
|
static void tcp_v6_reqsk_send_ack(struct sock *sk, struct sk_buff *skb,
|
|
|
|
struct request_sock *req)
|
2005-04-16 18:20:36 -04:00
|
|
|
{
|
2008-04-17 23:45:16 -04:00
|
|
|
tcp_v6_send_ack(skb, tcp_rsk(req)->snt_isn + 1, tcp_rsk(req)->rcv_isn + 1, req->rcv_wnd, req->ts_recent,
|
2008-08-07 02:50:04 -04:00
|
|
|
tcp_v6_md5_do_lookup(sk, &ipv6_hdr(skb)->daddr));
|
2005-04-16 18:20:36 -04:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
static struct sock *tcp_v6_hnd_req(struct sock *sk,struct sk_buff *skb)
|
|
|
|
{
|
2005-06-19 01:47:21 -04:00
|
|
|
struct request_sock *req, **prev;
|
2007-04-11 00:04:22 -04:00
|
|
|
const struct tcphdr *th = tcp_hdr(skb);
|
2005-04-16 18:20:36 -04:00
|
|
|
struct sock *nsk;
|
|
|
|
|
|
|
|
/* Find possible connection requests. */
|
2005-12-14 02:15:24 -05:00
|
|
|
req = inet6_csk_search_req(sk, &prev, th->source,
|
2007-04-25 20:54:47 -04:00
|
|
|
&ipv6_hdr(skb)->saddr,
|
|
|
|
&ipv6_hdr(skb)->daddr, inet6_iif(skb));
|
2005-04-16 18:20:36 -04:00
|
|
|
if (req)
|
|
|
|
return tcp_check_req(sk, skb, req, prev);
|
|
|
|
|
2008-03-25 13:26:21 -04:00
|
|
|
nsk = __inet6_lookup_established(sock_net(sk), &tcp_hashinfo,
|
2008-01-31 08:07:21 -05:00
|
|
|
&ipv6_hdr(skb)->saddr, th->source,
|
|
|
|
&ipv6_hdr(skb)->daddr, ntohs(th->dest), inet6_iif(skb));
|
2005-04-16 18:20:36 -04:00
|
|
|
|
|
|
|
if (nsk) {
|
|
|
|
if (nsk->sk_state != TCP_TIME_WAIT) {
|
|
|
|
bh_lock_sock(nsk);
|
|
|
|
return nsk;
|
|
|
|
}
|
2006-10-10 22:41:46 -04:00
|
|
|
inet_twsk_put(inet_twsk(nsk));
|
2005-04-16 18:20:36 -04:00
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
2008-02-08 00:49:26 -05:00
|
|
|
#ifdef CONFIG_SYN_COOKIES
|
2005-04-16 18:20:36 -04:00
|
|
|
if (!th->rst && !th->syn && th->ack)
|
2008-02-08 00:49:26 -05:00
|
|
|
sk = cookie_v6_check(sk, skb);
|
2005-04-16 18:20:36 -04:00
|
|
|
#endif
|
|
|
|
return sk;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* FIXME: this is substantially similar to the ipv4 code.
|
|
|
|
* Can some kind of merge be done? -- erics
|
|
|
|
*/
|
|
|
|
static int tcp_v6_conn_request(struct sock *sk, struct sk_buff *skb)
|
|
|
|
{
|
2005-12-14 02:15:40 -05:00
|
|
|
struct inet6_request_sock *treq;
|
2005-04-16 18:20:36 -04:00
|
|
|
struct ipv6_pinfo *np = inet6_sk(sk);
|
|
|
|
struct tcp_options_received tmp_opt;
|
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
2005-06-19 01:47:21 -04:00
|
|
|
struct request_sock *req = NULL;
|
2005-04-16 18:20:36 -04:00
|
|
|
__u32 isn = TCP_SKB_CB(skb)->when;
|
2008-02-08 00:49:26 -05:00
|
|
|
#ifdef CONFIG_SYN_COOKIES
|
|
|
|
int want_cookie = 0;
|
|
|
|
#else
|
|
|
|
#define want_cookie 0
|
|
|
|
#endif
|
2005-04-16 18:20:36 -04:00
|
|
|
|
|
|
|
if (skb->protocol == htons(ETH_P_IP))
|
|
|
|
return tcp_v4_conn_request(sk, skb);
|
|
|
|
|
|
|
|
if (!ipv6_unicast_destination(skb))
|
2007-02-09 09:24:49 -05:00
|
|
|
goto drop;
|
2005-04-16 18:20:36 -04:00
|
|
|
|
2005-08-09 23:10:42 -04:00
|
|
|
if (inet_csk_reqsk_queue_is_full(sk) && !isn) {
|
2005-04-16 18:20:36 -04:00
|
|
|
if (net_ratelimit())
|
2008-02-08 00:49:26 -05:00
|
|
|
syn_flood_warning(skb);
|
|
|
|
#ifdef CONFIG_SYN_COOKIES
|
|
|
|
if (sysctl_tcp_syncookies)
|
|
|
|
want_cookie = 1;
|
|
|
|
else
|
|
|
|
#endif
|
2007-02-09 09:24:49 -05:00
|
|
|
goto drop;
|
2005-04-16 18:20:36 -04:00
|
|
|
}
|
|
|
|
|
2005-08-09 23:10:42 -04:00
|
|
|
if (sk_acceptq_is_full(sk) && inet_csk_reqsk_queue_young(sk) > 1)
|
2005-04-16 18:20:36 -04:00
|
|
|
goto drop;
|
|
|
|
|
2005-12-14 02:15:40 -05:00
|
|
|
req = inet6_reqsk_alloc(&tcp6_request_sock_ops);
|
2005-04-16 18:20:36 -04:00
|
|
|
if (req == NULL)
|
|
|
|
goto drop;
|
|
|
|
|
2006-11-14 22:07:45 -05:00
|
|
|
#ifdef CONFIG_TCP_MD5SIG
|
|
|
|
tcp_rsk(req)->af_specific = &tcp_request_sock_ipv6_ops;
|
|
|
|
#endif
|
|
|
|
|
2005-04-16 18:20:36 -04:00
|
|
|
tcp_clear_options(&tmp_opt);
|
|
|
|
tmp_opt.mss_clamp = IPV6_MIN_MTU - sizeof(struct tcphdr) - sizeof(struct ipv6hdr);
|
|
|
|
tmp_opt.user_mss = tp->rx_opt.user_mss;
|
|
|
|
|
|
|
|
tcp_parse_options(skb, &tmp_opt, 0);
|
|
|
|
|
2008-04-10 06:12:40 -04:00
|
|
|
if (want_cookie && !tmp_opt.saw_tstamp)
|
2008-02-08 00:49:26 -05:00
|
|
|
tcp_clear_options(&tmp_opt);
|
|
|
|
|
2005-04-16 18:20:36 -04:00
|
|
|
tmp_opt.tstamp_ok = tmp_opt.saw_tstamp;
|
|
|
|
tcp_openreq_init(req, &tmp_opt, skb);
|
|
|
|
|
2005-12-14 02:15:40 -05:00
|
|
|
treq = inet6_rsk(req);
|
2007-04-25 20:54:47 -04:00
|
|
|
ipv6_addr_copy(&treq->rmt_addr, &ipv6_hdr(skb)->saddr);
|
|
|
|
ipv6_addr_copy(&treq->loc_addr, &ipv6_hdr(skb)->daddr);
|
2008-02-08 00:49:26 -05:00
|
|
|
if (!want_cookie)
|
|
|
|
TCP_ECN_create_request(req, tcp_hdr(skb));
|
|
|
|
|
|
|
|
if (want_cookie) {
|
|
|
|
isn = cookie_v6_init_sequence(sk, skb, &req->mss);
|
2008-04-10 06:12:40 -04:00
|
|
|
req->cookie_ts = tmp_opt.tstamp_ok;
|
2008-02-08 00:49:26 -05:00
|
|
|
} else if (!isn) {
|
|
|
|
if (ipv6_opt_accepted(sk, skb) ||
|
|
|
|
np->rxopt.bits.rxinfo || np->rxopt.bits.rxoinfo ||
|
|
|
|
np->rxopt.bits.rxhlim || np->rxopt.bits.rxohlim) {
|
|
|
|
atomic_inc(&skb->users);
|
|
|
|
treq->pktopts = skb;
|
|
|
|
}
|
|
|
|
treq->iif = sk->sk_bound_dev_if;
|
2005-04-16 18:20:36 -04:00
|
|
|
|
2008-02-08 00:49:26 -05:00
|
|
|
/* So that link locals have meaning */
|
|
|
|
if (!sk->sk_bound_dev_if &&
|
|
|
|
ipv6_addr_type(&treq->rmt_addr) & IPV6_ADDR_LINKLOCAL)
|
|
|
|
treq->iif = inet6_iif(skb);
|
2005-04-16 18:20:36 -04:00
|
|
|
|
2006-11-10 17:06:49 -05:00
|
|
|
isn = tcp_v6_init_sequence(skb);
|
2008-02-08 00:49:26 -05:00
|
|
|
}
|
2005-04-16 18:20:36 -04:00
|
|
|
|
[NET] Generalise TCP's struct open_request minisock infrastructure
Kept this first changeset minimal, without changing existing names to
ease peer review.
Basicaly tcp_openreq_alloc now receives the or_calltable, that in turn
has two new members:
->slab, that replaces tcp_openreq_cachep
->obj_size, to inform the size of the openreq descendant for
a specific protocol
The protocol specific fields in struct open_request were moved to a
class hierarchy, with the things that are common to all connection
oriented PF_INET protocols in struct inet_request_sock, the TCP ones
in tcp_request_sock, that is an inet_request_sock, that is an
open_request.
I.e. this uses the same approach used for the struct sock class
hierarchy, with sk_prot indicating if the protocol wants to use the
open_request infrastructure by filling in sk_prot->rsk_prot with an
or_calltable.
Results? Performance is improved and TCP v4 now uses only 64 bytes per
open request minisock, down from 96 without this patch :-)
Next changeset will rename some of the structs, fields and functions
mentioned above, struct or_calltable is way unclear, better name it
struct request_sock_ops, s/struct open_request/struct request_sock/g,
etc.
Signed-off-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-06-19 01:46:52 -04:00
|
|
|
tcp_rsk(req)->snt_isn = isn;
|
2005-04-16 18:20:36 -04:00
|
|
|
|
2006-07-25 02:32:50 -04:00
|
|
|
security_inet_conn_request(sk, skb, req);
|
|
|
|
|
2008-02-29 14:43:03 -05:00
|
|
|
if (tcp_v6_send_synack(sk, req))
|
2005-04-16 18:20:36 -04:00
|
|
|
goto drop;
|
|
|
|
|
2008-02-08 00:49:26 -05:00
|
|
|
if (!want_cookie) {
|
|
|
|
inet6_csk_reqsk_queue_hash_add(sk, req, TCP_TIMEOUT_INIT);
|
|
|
|
return 0;
|
|
|
|
}
|
2005-04-16 18:20:36 -04:00
|
|
|
|
|
|
|
drop:
|
|
|
|
if (req)
|
2005-06-19 01:47:21 -04:00
|
|
|
reqsk_free(req);
|
2005-04-16 18:20:36 -04:00
|
|
|
|
|
|
|
return 0; /* don't send reset */
|
|
|
|
}
|
|
|
|
|
|
|
|
static struct sock * tcp_v6_syn_recv_sock(struct sock *sk, struct sk_buff *skb,
|
2005-06-19 01:47:21 -04:00
|
|
|
struct request_sock *req,
|
2005-04-16 18:20:36 -04:00
|
|
|
struct dst_entry *dst)
|
|
|
|
{
|
2008-09-12 19:17:43 -04:00
|
|
|
struct inet6_request_sock *treq;
|
2005-04-16 18:20:36 -04:00
|
|
|
struct ipv6_pinfo *newnp, *np = inet6_sk(sk);
|
|
|
|
struct tcp6_sock *newtcp6sk;
|
|
|
|
struct inet_sock *newinet;
|
|
|
|
struct tcp_sock *newtp;
|
|
|
|
struct sock *newsk;
|
|
|
|
struct ipv6_txoptions *opt;
|
2006-11-14 22:07:45 -05:00
|
|
|
#ifdef CONFIG_TCP_MD5SIG
|
|
|
|
struct tcp_md5sig_key *key;
|
|
|
|
#endif
|
2005-04-16 18:20:36 -04:00
|
|
|
|
|
|
|
if (skb->protocol == htons(ETH_P_IP)) {
|
|
|
|
/*
|
|
|
|
* v6 mapped
|
|
|
|
*/
|
|
|
|
|
|
|
|
newsk = tcp_v4_syn_recv_sock(sk, skb, req, dst);
|
|
|
|
|
2007-02-09 09:24:49 -05:00
|
|
|
if (newsk == NULL)
|
2005-04-16 18:20:36 -04:00
|
|
|
return NULL;
|
|
|
|
|
|
|
|
newtcp6sk = (struct tcp6_sock *)newsk;
|
|
|
|
inet_sk(newsk)->pinet6 = &newtcp6sk->inet6;
|
|
|
|
|
|
|
|
newinet = inet_sk(newsk);
|
|
|
|
newnp = inet6_sk(newsk);
|
|
|
|
newtp = tcp_sk(newsk);
|
|
|
|
|
|
|
|
memcpy(newnp, np, sizeof(struct ipv6_pinfo));
|
|
|
|
|
|
|
|
ipv6_addr_set(&newnp->daddr, 0, 0, htonl(0x0000FFFF),
|
|
|
|
newinet->daddr);
|
|
|
|
|
|
|
|
ipv6_addr_set(&newnp->saddr, 0, 0, htonl(0x0000FFFF),
|
|
|
|
newinet->saddr);
|
|
|
|
|
|
|
|
ipv6_addr_copy(&newnp->rcv_saddr, &newnp->saddr);
|
|
|
|
|
2005-12-14 02:15:52 -05:00
|
|
|
inet_csk(newsk)->icsk_af_ops = &ipv6_mapped;
|
2005-04-16 18:20:36 -04:00
|
|
|
newsk->sk_backlog_rcv = tcp_v4_do_rcv;
|
2006-11-14 22:07:45 -05:00
|
|
|
#ifdef CONFIG_TCP_MD5SIG
|
|
|
|
newtp->af_specific = &tcp_sock_ipv6_mapped_specific;
|
|
|
|
#endif
|
|
|
|
|
2005-04-16 18:20:36 -04:00
|
|
|
newnp->pktoptions = NULL;
|
|
|
|
newnp->opt = NULL;
|
2005-08-12 08:19:38 -04:00
|
|
|
newnp->mcast_oif = inet6_iif(skb);
|
2007-04-25 20:54:47 -04:00
|
|
|
newnp->mcast_hops = ipv6_hdr(skb)->hop_limit;
|
2005-04-16 18:20:36 -04:00
|
|
|
|
2005-08-09 22:45:38 -04:00
|
|
|
/*
|
|
|
|
* No need to charge this sock to the relevant IPv6 refcnt debug socks count
|
|
|
|
* here, tcp_create_openreq_child now does this for us, see the comment in
|
|
|
|
* that function for the gory details. -acme
|
2005-04-16 18:20:36 -04:00
|
|
|
*/
|
|
|
|
|
|
|
|
/* It is tricky place. Until this moment IPv4 tcp
|
2005-12-14 02:15:52 -05:00
|
|
|
worked with IPv6 icsk.icsk_af_ops.
|
2005-04-16 18:20:36 -04:00
|
|
|
Sync it now.
|
|
|
|
*/
|
2005-12-14 02:26:10 -05:00
|
|
|
tcp_sync_mss(newsk, inet_csk(newsk)->icsk_pmtu_cookie);
|
2005-04-16 18:20:36 -04:00
|
|
|
|
|
|
|
return newsk;
|
|
|
|
}
|
|
|
|
|
2008-09-12 19:17:43 -04:00
|
|
|
treq = inet6_rsk(req);
|
2005-04-16 18:20:36 -04:00
|
|
|
opt = np->opt;
|
|
|
|
|
|
|
|
if (sk_acceptq_is_full(sk))
|
|
|
|
goto out_overflow;
|
|
|
|
|
|
|
|
if (dst == NULL) {
|
|
|
|
struct in6_addr *final_p = NULL, final;
|
|
|
|
struct flowi fl;
|
|
|
|
|
|
|
|
memset(&fl, 0, sizeof(fl));
|
|
|
|
fl.proto = IPPROTO_TCP;
|
[NET] Generalise TCP's struct open_request minisock infrastructure
Kept this first changeset minimal, without changing existing names to
ease peer review.
Basicaly tcp_openreq_alloc now receives the or_calltable, that in turn
has two new members:
->slab, that replaces tcp_openreq_cachep
->obj_size, to inform the size of the openreq descendant for
a specific protocol
The protocol specific fields in struct open_request were moved to a
class hierarchy, with the things that are common to all connection
oriented PF_INET protocols in struct inet_request_sock, the TCP ones
in tcp_request_sock, that is an inet_request_sock, that is an
open_request.
I.e. this uses the same approach used for the struct sock class
hierarchy, with sk_prot indicating if the protocol wants to use the
open_request infrastructure by filling in sk_prot->rsk_prot with an
or_calltable.
Results? Performance is improved and TCP v4 now uses only 64 bytes per
open request minisock, down from 96 without this patch :-)
Next changeset will rename some of the structs, fields and functions
mentioned above, struct or_calltable is way unclear, better name it
struct request_sock_ops, s/struct open_request/struct request_sock/g,
etc.
Signed-off-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-06-19 01:46:52 -04:00
|
|
|
ipv6_addr_copy(&fl.fl6_dst, &treq->rmt_addr);
|
2005-04-16 18:20:36 -04:00
|
|
|
if (opt && opt->srcrt) {
|
|
|
|
struct rt0_hdr *rt0 = (struct rt0_hdr *) opt->srcrt;
|
|
|
|
ipv6_addr_copy(&final, &fl.fl6_dst);
|
|
|
|
ipv6_addr_copy(&fl.fl6_dst, rt0->addr);
|
|
|
|
final_p = &final;
|
|
|
|
}
|
[NET] Generalise TCP's struct open_request minisock infrastructure
Kept this first changeset minimal, without changing existing names to
ease peer review.
Basicaly tcp_openreq_alloc now receives the or_calltable, that in turn
has two new members:
->slab, that replaces tcp_openreq_cachep
->obj_size, to inform the size of the openreq descendant for
a specific protocol
The protocol specific fields in struct open_request were moved to a
class hierarchy, with the things that are common to all connection
oriented PF_INET protocols in struct inet_request_sock, the TCP ones
in tcp_request_sock, that is an inet_request_sock, that is an
open_request.
I.e. this uses the same approach used for the struct sock class
hierarchy, with sk_prot indicating if the protocol wants to use the
open_request infrastructure by filling in sk_prot->rsk_prot with an
or_calltable.
Results? Performance is improved and TCP v4 now uses only 64 bytes per
open request minisock, down from 96 without this patch :-)
Next changeset will rename some of the structs, fields and functions
mentioned above, struct or_calltable is way unclear, better name it
struct request_sock_ops, s/struct open_request/struct request_sock/g,
etc.
Signed-off-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-06-19 01:46:52 -04:00
|
|
|
ipv6_addr_copy(&fl.fl6_src, &treq->loc_addr);
|
2005-04-16 18:20:36 -04:00
|
|
|
fl.oif = sk->sk_bound_dev_if;
|
[NET] Generalise TCP's struct open_request minisock infrastructure
Kept this first changeset minimal, without changing existing names to
ease peer review.
Basicaly tcp_openreq_alloc now receives the or_calltable, that in turn
has two new members:
->slab, that replaces tcp_openreq_cachep
->obj_size, to inform the size of the openreq descendant for
a specific protocol
The protocol specific fields in struct open_request were moved to a
class hierarchy, with the things that are common to all connection
oriented PF_INET protocols in struct inet_request_sock, the TCP ones
in tcp_request_sock, that is an inet_request_sock, that is an
open_request.
I.e. this uses the same approach used for the struct sock class
hierarchy, with sk_prot indicating if the protocol wants to use the
open_request infrastructure by filling in sk_prot->rsk_prot with an
or_calltable.
Results? Performance is improved and TCP v4 now uses only 64 bytes per
open request minisock, down from 96 without this patch :-)
Next changeset will rename some of the structs, fields and functions
mentioned above, struct or_calltable is way unclear, better name it
struct request_sock_ops, s/struct open_request/struct request_sock/g,
etc.
Signed-off-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-06-19 01:46:52 -04:00
|
|
|
fl.fl_ip_dport = inet_rsk(req)->rmt_port;
|
2008-10-20 02:35:58 -04:00
|
|
|
fl.fl_ip_sport = inet_rsk(req)->loc_port;
|
2006-07-25 02:32:50 -04:00
|
|
|
security_req_classify_flow(req, &fl);
|
2005-04-16 18:20:36 -04:00
|
|
|
|
|
|
|
if (ip6_dst_lookup(sk, &dst, &fl))
|
|
|
|
goto out;
|
|
|
|
|
|
|
|
if (final_p)
|
|
|
|
ipv6_addr_copy(&fl.fl6_dst, final_p);
|
|
|
|
|
2008-11-25 20:35:18 -05:00
|
|
|
if ((xfrm_lookup(sock_net(sk), &dst, &fl, sk, 0)) < 0)
|
2005-04-16 18:20:36 -04:00
|
|
|
goto out;
|
2007-02-09 09:24:49 -05:00
|
|
|
}
|
2005-04-16 18:20:36 -04:00
|
|
|
|
|
|
|
newsk = tcp_create_openreq_child(sk, req, skb);
|
|
|
|
if (newsk == NULL)
|
|
|
|
goto out;
|
|
|
|
|
2005-08-09 22:45:38 -04:00
|
|
|
/*
|
|
|
|
* No need to charge this sock to the relevant IPv6 refcnt debug socks
|
|
|
|
* count here, tcp_create_openreq_child now does this for us, see the
|
|
|
|
* comment in that function for the gory details. -acme
|
|
|
|
*/
|
2005-04-16 18:20:36 -04:00
|
|
|
|
2006-08-25 18:55:43 -04:00
|
|
|
newsk->sk_gso_type = SKB_GSO_TCPV6;
|
2006-08-29 20:15:09 -04:00
|
|
|
__ip6_dst_store(newsk, dst, NULL, NULL);
|
2005-04-16 18:20:36 -04:00
|
|
|
|
|
|
|
newtcp6sk = (struct tcp6_sock *)newsk;
|
|
|
|
inet_sk(newsk)->pinet6 = &newtcp6sk->inet6;
|
|
|
|
|
|
|
|
newtp = tcp_sk(newsk);
|
|
|
|
newinet = inet_sk(newsk);
|
|
|
|
newnp = inet6_sk(newsk);
|
|
|
|
|
|
|
|
memcpy(newnp, np, sizeof(struct ipv6_pinfo));
|
|
|
|
|
[NET] Generalise TCP's struct open_request minisock infrastructure
Kept this first changeset minimal, without changing existing names to
ease peer review.
Basicaly tcp_openreq_alloc now receives the or_calltable, that in turn
has two new members:
->slab, that replaces tcp_openreq_cachep
->obj_size, to inform the size of the openreq descendant for
a specific protocol
The protocol specific fields in struct open_request were moved to a
class hierarchy, with the things that are common to all connection
oriented PF_INET protocols in struct inet_request_sock, the TCP ones
in tcp_request_sock, that is an inet_request_sock, that is an
open_request.
I.e. this uses the same approach used for the struct sock class
hierarchy, with sk_prot indicating if the protocol wants to use the
open_request infrastructure by filling in sk_prot->rsk_prot with an
or_calltable.
Results? Performance is improved and TCP v4 now uses only 64 bytes per
open request minisock, down from 96 without this patch :-)
Next changeset will rename some of the structs, fields and functions
mentioned above, struct or_calltable is way unclear, better name it
struct request_sock_ops, s/struct open_request/struct request_sock/g,
etc.
Signed-off-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-06-19 01:46:52 -04:00
|
|
|
ipv6_addr_copy(&newnp->daddr, &treq->rmt_addr);
|
|
|
|
ipv6_addr_copy(&newnp->saddr, &treq->loc_addr);
|
|
|
|
ipv6_addr_copy(&newnp->rcv_saddr, &treq->loc_addr);
|
|
|
|
newsk->sk_bound_dev_if = treq->iif;
|
2005-04-16 18:20:36 -04:00
|
|
|
|
2007-02-09 09:24:49 -05:00
|
|
|
/* Now IPv6 options...
|
2005-04-16 18:20:36 -04:00
|
|
|
|
|
|
|
First: no IPv4 options.
|
|
|
|
*/
|
|
|
|
newinet->opt = NULL;
|
2007-03-16 19:14:03 -04:00
|
|
|
newnp->ipv6_fl_list = NULL;
|
2005-04-16 18:20:36 -04:00
|
|
|
|
|
|
|
/* Clone RX bits */
|
|
|
|
newnp->rxopt.all = np->rxopt.all;
|
|
|
|
|
|
|
|
/* Clone pktoptions received with SYN */
|
|
|
|
newnp->pktoptions = NULL;
|
[NET] Generalise TCP's struct open_request minisock infrastructure
Kept this first changeset minimal, without changing existing names to
ease peer review.
Basicaly tcp_openreq_alloc now receives the or_calltable, that in turn
has two new members:
->slab, that replaces tcp_openreq_cachep
->obj_size, to inform the size of the openreq descendant for
a specific protocol
The protocol specific fields in struct open_request were moved to a
class hierarchy, with the things that are common to all connection
oriented PF_INET protocols in struct inet_request_sock, the TCP ones
in tcp_request_sock, that is an inet_request_sock, that is an
open_request.
I.e. this uses the same approach used for the struct sock class
hierarchy, with sk_prot indicating if the protocol wants to use the
open_request infrastructure by filling in sk_prot->rsk_prot with an
or_calltable.
Results? Performance is improved and TCP v4 now uses only 64 bytes per
open request minisock, down from 96 without this patch :-)
Next changeset will rename some of the structs, fields and functions
mentioned above, struct or_calltable is way unclear, better name it
struct request_sock_ops, s/struct open_request/struct request_sock/g,
etc.
Signed-off-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-06-19 01:46:52 -04:00
|
|
|
if (treq->pktopts != NULL) {
|
|
|
|
newnp->pktoptions = skb_clone(treq->pktopts, GFP_ATOMIC);
|
|
|
|
kfree_skb(treq->pktopts);
|
|
|
|
treq->pktopts = NULL;
|
2005-04-16 18:20:36 -04:00
|
|
|
if (newnp->pktoptions)
|
|
|
|
skb_set_owner_r(newnp->pktoptions, newsk);
|
|
|
|
}
|
|
|
|
newnp->opt = NULL;
|
2005-08-12 08:19:38 -04:00
|
|
|
newnp->mcast_oif = inet6_iif(skb);
|
2007-04-25 20:54:47 -04:00
|
|
|
newnp->mcast_hops = ipv6_hdr(skb)->hop_limit;
|
2005-04-16 18:20:36 -04:00
|
|
|
|
|
|
|
/* Clone native IPv6 options from listening socket (if any)
|
|
|
|
|
|
|
|
Yes, keeping reference count would be much more clever,
|
|
|
|
but we make one more one thing there: reattach optmem
|
|
|
|
to newsk.
|
|
|
|
*/
|
|
|
|
if (opt) {
|
|
|
|
newnp->opt = ipv6_dup_options(newsk, opt);
|
|
|
|
if (opt != np->opt)
|
|
|
|
sock_kfree_s(sk, opt, opt->tot_len);
|
|
|
|
}
|
|
|
|
|
2005-12-14 02:26:10 -05:00
|
|
|
inet_csk(newsk)->icsk_ext_hdr_len = 0;
|
2005-04-16 18:20:36 -04:00
|
|
|
if (newnp->opt)
|
2005-12-14 02:26:10 -05:00
|
|
|
inet_csk(newsk)->icsk_ext_hdr_len = (newnp->opt->opt_nflen +
|
|
|
|
newnp->opt->opt_flen);
|
2005-04-16 18:20:36 -04:00
|
|
|
|
2006-03-20 20:53:41 -05:00
|
|
|
tcp_mtup_init(newsk);
|
2005-04-16 18:20:36 -04:00
|
|
|
tcp_sync_mss(newsk, dst_mtu(dst));
|
|
|
|
newtp->advmss = dst_metric(dst, RTAX_ADVMSS);
|
|
|
|
tcp_initialize_rcv_mss(newsk);
|
|
|
|
|
|
|
|
newinet->daddr = newinet->saddr = newinet->rcv_saddr = LOOPBACK4_IPV6;
|
|
|
|
|
2006-11-14 22:07:45 -05:00
|
|
|
#ifdef CONFIG_TCP_MD5SIG
|
|
|
|
/* Copy over the MD5 key from the original socket */
|
|
|
|
if ((key = tcp_v6_md5_do_lookup(sk, &newnp->daddr)) != NULL) {
|
|
|
|
/* We're using one, so create a matching key
|
|
|
|
* on the newsk structure. If we fail to get
|
|
|
|
* memory, then we end up not copying the key
|
|
|
|
* across. Shucks.
|
|
|
|
*/
|
2006-11-17 09:14:37 -05:00
|
|
|
char *newkey = kmemdup(key->key, key->keylen, GFP_ATOMIC);
|
|
|
|
if (newkey != NULL)
|
2006-11-14 22:07:45 -05:00
|
|
|
tcp_v6_md5_do_add(newsk, &inet6_sk(sk)->daddr,
|
|
|
|
newkey, key->keylen);
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
[SOCK] proto: Add hashinfo member to struct proto
This way we can remove TCP and DCCP specific versions of
sk->sk_prot->get_port: both v4 and v6 use inet_csk_get_port
sk->sk_prot->hash: inet_hash is directly used, only v6 need
a specific version to deal with mapped sockets
sk->sk_prot->unhash: both v4 and v6 use inet_hash directly
struct inet_connection_sock_af_ops also gets a new member, bind_conflict, so
that inet_csk_get_port can find the per family routine.
Now only the lookup routines receive as a parameter a struct inet_hashtable.
With this we further reuse code, reducing the difference among INET transport
protocols.
Eventually work has to be done on UDP and SCTP to make them share this
infrastructure and get as a bonus inet_diag interfaces so that iproute can be
used with these protocols.
net-2.6/net/ipv4/inet_hashtables.c:
struct proto | +8
struct inet_connection_sock_af_ops | +8
2 structs changed
__inet_hash_nolisten | +18
__inet_hash | -210
inet_put_port | +8
inet_bind_bucket_create | +1
__inet_hash_connect | -8
5 functions changed, 27 bytes added, 218 bytes removed, diff: -191
net-2.6/net/core/sock.c:
proto_seq_show | +3
1 function changed, 3 bytes added, diff: +3
net-2.6/net/ipv4/inet_connection_sock.c:
inet_csk_get_port | +15
1 function changed, 15 bytes added, diff: +15
net-2.6/net/ipv4/tcp.c:
tcp_set_state | -7
1 function changed, 7 bytes removed, diff: -7
net-2.6/net/ipv4/tcp_ipv4.c:
tcp_v4_get_port | -31
tcp_v4_hash | -48
tcp_v4_destroy_sock | -7
tcp_v4_syn_recv_sock | -2
tcp_unhash | -179
5 functions changed, 267 bytes removed, diff: -267
net-2.6/net/ipv6/inet6_hashtables.c:
__inet6_hash | +8
1 function changed, 8 bytes added, diff: +8
net-2.6/net/ipv4/inet_hashtables.c:
inet_unhash | +190
inet_hash | +242
2 functions changed, 432 bytes added, diff: +432
vmlinux:
16 functions changed, 485 bytes added, 492 bytes removed, diff: -7
/home/acme/git/net-2.6/net/ipv6/tcp_ipv6.c:
tcp_v6_get_port | -31
tcp_v6_hash | -7
tcp_v6_syn_recv_sock | -9
3 functions changed, 47 bytes removed, diff: -47
/home/acme/git/net-2.6/net/dccp/proto.c:
dccp_destroy_sock | -7
dccp_unhash | -179
dccp_hash | -49
dccp_set_state | -7
dccp_done | +1
5 functions changed, 1 bytes added, 242 bytes removed, diff: -241
/home/acme/git/net-2.6/net/dccp/ipv4.c:
dccp_v4_get_port | -31
dccp_v4_request_recv_sock | -2
2 functions changed, 33 bytes removed, diff: -33
/home/acme/git/net-2.6/net/dccp/ipv6.c:
dccp_v6_get_port | -31
dccp_v6_hash | -7
dccp_v6_request_recv_sock | +5
3 functions changed, 5 bytes added, 38 bytes removed, diff: -33
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-02-03 07:06:04 -05:00
|
|
|
__inet6_hash(newsk);
|
2008-04-18 02:17:34 -04:00
|
|
|
__inet_inherit_port(sk, newsk);
|
2005-04-16 18:20:36 -04:00
|
|
|
|
|
|
|
return newsk;
|
|
|
|
|
|
|
|
out_overflow:
|
2008-07-16 23:31:16 -04:00
|
|
|
NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_LISTENOVERFLOWS);
|
2005-04-16 18:20:36 -04:00
|
|
|
out:
|
2008-07-16 23:31:16 -04:00
|
|
|
NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_LISTENDROPS);
|
2005-04-16 18:20:36 -04:00
|
|
|
if (opt && opt != np->opt)
|
|
|
|
sock_kfree_s(sk, opt, opt->tot_len);
|
|
|
|
dst_release(dst);
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
2006-11-15 00:40:42 -05:00
|
|
|
static __sum16 tcp_v6_checksum_init(struct sk_buff *skb)
|
2005-04-16 18:20:36 -04:00
|
|
|
{
|
2006-08-29 19:44:56 -04:00
|
|
|
if (skb->ip_summed == CHECKSUM_COMPLETE) {
|
2009-01-08 13:41:23 -05:00
|
|
|
if (!tcp_v6_check(skb->len, &ipv6_hdr(skb)->saddr,
|
2007-04-25 20:54:47 -04:00
|
|
|
&ipv6_hdr(skb)->daddr, skb->csum)) {
|
2005-11-10 16:01:24 -05:00
|
|
|
skb->ip_summed = CHECKSUM_UNNECESSARY;
|
2005-04-16 18:20:36 -04:00
|
|
|
return 0;
|
2005-11-10 16:01:24 -05:00
|
|
|
}
|
2005-04-16 18:20:36 -04:00
|
|
|
}
|
2005-11-10 16:01:24 -05:00
|
|
|
|
2009-01-08 13:41:23 -05:00
|
|
|
skb->csum = ~csum_unfold(tcp_v6_check(skb->len,
|
2007-04-25 20:54:47 -04:00
|
|
|
&ipv6_hdr(skb)->saddr,
|
|
|
|
&ipv6_hdr(skb)->daddr, 0));
|
2005-11-10 16:01:24 -05:00
|
|
|
|
2005-04-16 18:20:36 -04:00
|
|
|
if (skb->len <= 76) {
|
2005-11-10 16:01:24 -05:00
|
|
|
return __skb_checksum_complete(skb);
|
2005-04-16 18:20:36 -04:00
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* The socket must have it's spinlock held when we get
|
|
|
|
* here.
|
|
|
|
*
|
|
|
|
* We have a potential double-lock case here, so even when
|
|
|
|
* doing backlog processing we use the BH locking scheme.
|
|
|
|
* This is because we cannot sleep with the original spinlock
|
|
|
|
* held.
|
|
|
|
*/
|
|
|
|
static int tcp_v6_do_rcv(struct sock *sk, struct sk_buff *skb)
|
|
|
|
{
|
|
|
|
struct ipv6_pinfo *np = inet6_sk(sk);
|
|
|
|
struct tcp_sock *tp;
|
|
|
|
struct sk_buff *opt_skb = NULL;
|
|
|
|
|
|
|
|
/* Imagine: socket is IPv6. IPv4 packet arrives,
|
|
|
|
goes to IPv4 receive handler and backlogged.
|
|
|
|
From backlog it always goes here. Kerboom...
|
|
|
|
Fortunately, tcp_rcv_established and rcv_established
|
|
|
|
handle them correctly, but it is not case with
|
|
|
|
tcp_v6_hnd_req and tcp_v6_send_reset(). --ANK
|
|
|
|
*/
|
|
|
|
|
|
|
|
if (skb->protocol == htons(ETH_P_IP))
|
|
|
|
return tcp_v4_do_rcv(sk, skb);
|
|
|
|
|
2006-11-14 22:07:45 -05:00
|
|
|
#ifdef CONFIG_TCP_MD5SIG
|
|
|
|
if (tcp_v6_inbound_md5_hash (sk, skb))
|
|
|
|
goto discard;
|
|
|
|
#endif
|
|
|
|
|
2006-08-31 18:28:39 -04:00
|
|
|
if (sk_filter(sk, skb))
|
2005-04-16 18:20:36 -04:00
|
|
|
goto discard;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* socket locking is here for SMP purposes as backlog rcv
|
|
|
|
* is currently called with bh processing disabled.
|
|
|
|
*/
|
|
|
|
|
|
|
|
/* Do Stevens' IPV6_PKTOPTIONS.
|
|
|
|
|
|
|
|
Yes, guys, it is the only place in our code, where we
|
|
|
|
may make it not affecting IPv4.
|
|
|
|
The rest of code is protocol independent,
|
|
|
|
and I do not like idea to uglify IPv4.
|
|
|
|
|
|
|
|
Actually, all the idea behind IPV6_PKTOPTIONS
|
|
|
|
looks not very well thought. For now we latch
|
|
|
|
options, received in the last packet, enqueued
|
|
|
|
by tcp. Feel free to propose better solution.
|
2007-02-09 09:24:49 -05:00
|
|
|
--ANK (980728)
|
2005-04-16 18:20:36 -04:00
|
|
|
*/
|
|
|
|
if (np->rxopt.all)
|
|
|
|
opt_skb = skb_clone(skb, GFP_ATOMIC);
|
|
|
|
|
|
|
|
if (sk->sk_state == TCP_ESTABLISHED) { /* Fast path */
|
|
|
|
TCP_CHECK_TIMER(sk);
|
2007-04-11 00:04:22 -04:00
|
|
|
if (tcp_rcv_established(sk, skb, tcp_hdr(skb), skb->len))
|
2005-04-16 18:20:36 -04:00
|
|
|
goto reset;
|
|
|
|
TCP_CHECK_TIMER(sk);
|
|
|
|
if (opt_skb)
|
|
|
|
goto ipv6_pktoptions;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2007-03-18 20:43:48 -04:00
|
|
|
if (skb->len < tcp_hdrlen(skb) || tcp_checksum_complete(skb))
|
2005-04-16 18:20:36 -04:00
|
|
|
goto csum_err;
|
|
|
|
|
2007-02-09 09:24:49 -05:00
|
|
|
if (sk->sk_state == TCP_LISTEN) {
|
2005-04-16 18:20:36 -04:00
|
|
|
struct sock *nsk = tcp_v6_hnd_req(sk, skb);
|
|
|
|
if (!nsk)
|
|
|
|
goto discard;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Queue it on the new socket if the new socket is active,
|
|
|
|
* otherwise we just shortcircuit this and continue with
|
|
|
|
* the new socket..
|
|
|
|
*/
|
2007-02-09 09:24:49 -05:00
|
|
|
if(nsk != sk) {
|
2005-04-16 18:20:36 -04:00
|
|
|
if (tcp_child_process(sk, nsk, skb))
|
|
|
|
goto reset;
|
|
|
|
if (opt_skb)
|
|
|
|
__kfree_skb(opt_skb);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
TCP_CHECK_TIMER(sk);
|
2007-04-11 00:04:22 -04:00
|
|
|
if (tcp_rcv_state_process(sk, skb, tcp_hdr(skb), skb->len))
|
2005-04-16 18:20:36 -04:00
|
|
|
goto reset;
|
|
|
|
TCP_CHECK_TIMER(sk);
|
|
|
|
if (opt_skb)
|
|
|
|
goto ipv6_pktoptions;
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
reset:
|
2006-11-14 22:07:45 -05:00
|
|
|
tcp_v6_send_reset(sk, skb);
|
2005-04-16 18:20:36 -04:00
|
|
|
discard:
|
|
|
|
if (opt_skb)
|
|
|
|
__kfree_skb(opt_skb);
|
|
|
|
kfree_skb(skb);
|
|
|
|
return 0;
|
|
|
|
csum_err:
|
2008-07-16 23:22:25 -04:00
|
|
|
TCP_INC_STATS_BH(sock_net(sk), TCP_MIB_INERRS);
|
2005-04-16 18:20:36 -04:00
|
|
|
goto discard;
|
|
|
|
|
|
|
|
|
|
|
|
ipv6_pktoptions:
|
|
|
|
/* Do you ask, what is it?
|
|
|
|
|
|
|
|
1. skb was enqueued by tcp.
|
|
|
|
2. skb is added to tail of read queue, rather than out of order.
|
|
|
|
3. socket is not in passive state.
|
|
|
|
4. Finally, it really contains options, which user wants to receive.
|
|
|
|
*/
|
|
|
|
tp = tcp_sk(sk);
|
|
|
|
if (TCP_SKB_CB(opt_skb)->end_seq == tp->rcv_nxt &&
|
|
|
|
!((1 << sk->sk_state) & (TCPF_CLOSE | TCPF_LISTEN))) {
|
[IPV6]: Support several new sockopt / ancillary data in Advanced API (RFC3542).
Support several new socket options / ancillary data:
IPV6_RECVPKTINFO, IPV6_PKTINFO,
IPV6_RECVHOPOPTS, IPV6_HOPOPTS,
IPV6_RECVDSTOPTS, IPV6_DSTOPTS, IPV6_RTHDRDSTOPTS,
IPV6_RECVRTHDR, IPV6_RTHDR,
IPV6_RECVHOPOPTS, IPV6_HOPOPTS
Old semantics are preserved as IPV6_2292xxxx so that
we can maintain backward compatibility.
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
2005-09-07 20:59:17 -04:00
|
|
|
if (np->rxopt.bits.rxinfo || np->rxopt.bits.rxoinfo)
|
2005-08-12 08:19:38 -04:00
|
|
|
np->mcast_oif = inet6_iif(opt_skb);
|
[IPV6]: Support several new sockopt / ancillary data in Advanced API (RFC3542).
Support several new socket options / ancillary data:
IPV6_RECVPKTINFO, IPV6_PKTINFO,
IPV6_RECVHOPOPTS, IPV6_HOPOPTS,
IPV6_RECVDSTOPTS, IPV6_DSTOPTS, IPV6_RTHDRDSTOPTS,
IPV6_RECVRTHDR, IPV6_RTHDR,
IPV6_RECVHOPOPTS, IPV6_HOPOPTS
Old semantics are preserved as IPV6_2292xxxx so that
we can maintain backward compatibility.
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
2005-09-07 20:59:17 -04:00
|
|
|
if (np->rxopt.bits.rxhlim || np->rxopt.bits.rxohlim)
|
2007-04-25 20:54:47 -04:00
|
|
|
np->mcast_hops = ipv6_hdr(opt_skb)->hop_limit;
|
2005-04-16 18:20:36 -04:00
|
|
|
if (ipv6_opt_accepted(sk, opt_skb)) {
|
|
|
|
skb_set_owner_r(opt_skb, sk);
|
|
|
|
opt_skb = xchg(&np->pktoptions, opt_skb);
|
|
|
|
} else {
|
|
|
|
__kfree_skb(opt_skb);
|
|
|
|
opt_skb = xchg(&np->pktoptions, NULL);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2009-02-23 16:45:33 -05:00
|
|
|
kfree_skb(opt_skb);
|
2005-04-16 18:20:36 -04:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2007-10-15 15:50:28 -04:00
|
|
|
static int tcp_v6_rcv(struct sk_buff *skb)
|
2005-04-16 18:20:36 -04:00
|
|
|
{
|
2007-02-09 09:24:49 -05:00
|
|
|
struct tcphdr *th;
|
2005-04-16 18:20:36 -04:00
|
|
|
struct sock *sk;
|
|
|
|
int ret;
|
2008-07-16 23:20:58 -04:00
|
|
|
struct net *net = dev_net(skb->dev);
|
2005-04-16 18:20:36 -04:00
|
|
|
|
|
|
|
if (skb->pkt_type != PACKET_HOST)
|
|
|
|
goto discard_it;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Count it even if it's bad.
|
|
|
|
*/
|
2008-07-16 23:22:25 -04:00
|
|
|
TCP_INC_STATS_BH(net, TCP_MIB_INSEGS);
|
2005-04-16 18:20:36 -04:00
|
|
|
|
|
|
|
if (!pskb_may_pull(skb, sizeof(struct tcphdr)))
|
|
|
|
goto discard_it;
|
|
|
|
|
2007-04-11 00:04:22 -04:00
|
|
|
th = tcp_hdr(skb);
|
2005-04-16 18:20:36 -04:00
|
|
|
|
|
|
|
if (th->doff < sizeof(struct tcphdr)/4)
|
|
|
|
goto bad_packet;
|
|
|
|
if (!pskb_may_pull(skb, th->doff*4))
|
|
|
|
goto discard_it;
|
|
|
|
|
2007-04-09 14:59:39 -04:00
|
|
|
if (!skb_csum_unnecessary(skb) && tcp_v6_checksum_init(skb))
|
2005-04-16 18:20:36 -04:00
|
|
|
goto bad_packet;
|
|
|
|
|
2007-04-11 00:04:22 -04:00
|
|
|
th = tcp_hdr(skb);
|
2005-04-16 18:20:36 -04:00
|
|
|
TCP_SKB_CB(skb)->seq = ntohl(th->seq);
|
|
|
|
TCP_SKB_CB(skb)->end_seq = (TCP_SKB_CB(skb)->seq + th->syn + th->fin +
|
|
|
|
skb->len - th->doff*4);
|
|
|
|
TCP_SKB_CB(skb)->ack_seq = ntohl(th->ack_seq);
|
|
|
|
TCP_SKB_CB(skb)->when = 0;
|
2007-04-25 20:54:47 -04:00
|
|
|
TCP_SKB_CB(skb)->flags = ipv6_get_dsfield(ipv6_hdr(skb));
|
2005-04-16 18:20:36 -04:00
|
|
|
TCP_SKB_CB(skb)->sacked = 0;
|
|
|
|
|
2008-10-07 14:41:57 -04:00
|
|
|
sk = __inet6_lookup_skb(&tcp_hashinfo, skb, th->source, th->dest);
|
2005-04-16 18:20:36 -04:00
|
|
|
if (!sk)
|
|
|
|
goto no_tcp_socket;
|
|
|
|
|
|
|
|
process:
|
|
|
|
if (sk->sk_state == TCP_TIME_WAIT)
|
|
|
|
goto do_time_wait;
|
|
|
|
|
|
|
|
if (!xfrm6_policy_check(sk, XFRM_POLICY_IN, skb))
|
|
|
|
goto discard_and_relse;
|
|
|
|
|
2006-08-31 18:28:39 -04:00
|
|
|
if (sk_filter(sk, skb))
|
2005-04-16 18:20:36 -04:00
|
|
|
goto discard_and_relse;
|
|
|
|
|
|
|
|
skb->dev = NULL;
|
|
|
|
|
2006-09-26 01:28:47 -04:00
|
|
|
bh_lock_sock_nested(sk);
|
2005-04-16 18:20:36 -04:00
|
|
|
ret = 0;
|
|
|
|
if (!sock_owned_by_user(sk)) {
|
2006-05-23 21:05:53 -04:00
|
|
|
#ifdef CONFIG_NET_DMA
|
2007-02-09 09:24:49 -05:00
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
2007-10-26 07:20:13 -04:00
|
|
|
if (!tp->ucopy.dma_chan && tp->ucopy.pinned_list)
|
2009-01-06 13:38:15 -05:00
|
|
|
tp->ucopy.dma_chan = dma_find_channel(DMA_MEMCPY);
|
2007-02-09 09:24:49 -05:00
|
|
|
if (tp->ucopy.dma_chan)
|
|
|
|
ret = tcp_v6_do_rcv(sk, skb);
|
|
|
|
else
|
2006-05-23 21:05:53 -04:00
|
|
|
#endif
|
|
|
|
{
|
|
|
|
if (!tcp_prequeue(sk, skb))
|
|
|
|
ret = tcp_v6_do_rcv(sk, skb);
|
|
|
|
}
|
2005-04-16 18:20:36 -04:00
|
|
|
} else
|
|
|
|
sk_add_backlog(sk, skb);
|
|
|
|
bh_unlock_sock(sk);
|
|
|
|
|
|
|
|
sock_put(sk);
|
|
|
|
return ret ? -1 : 0;
|
|
|
|
|
|
|
|
no_tcp_socket:
|
|
|
|
if (!xfrm6_policy_check(NULL, XFRM_POLICY_IN, skb))
|
|
|
|
goto discard_it;
|
|
|
|
|
|
|
|
if (skb->len < (th->doff<<2) || tcp_checksum_complete(skb)) {
|
|
|
|
bad_packet:
|
2008-07-16 23:22:25 -04:00
|
|
|
TCP_INC_STATS_BH(net, TCP_MIB_INERRS);
|
2005-04-16 18:20:36 -04:00
|
|
|
} else {
|
2006-11-14 22:07:45 -05:00
|
|
|
tcp_v6_send_reset(NULL, skb);
|
2005-04-16 18:20:36 -04:00
|
|
|
}
|
|
|
|
|
|
|
|
discard_it:
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Discard frame
|
|
|
|
*/
|
|
|
|
|
|
|
|
kfree_skb(skb);
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
discard_and_relse:
|
|
|
|
sock_put(sk);
|
|
|
|
goto discard_it;
|
|
|
|
|
|
|
|
do_time_wait:
|
|
|
|
if (!xfrm6_policy_check(NULL, XFRM_POLICY_IN, skb)) {
|
2006-10-10 22:41:46 -04:00
|
|
|
inet_twsk_put(inet_twsk(sk));
|
2005-04-16 18:20:36 -04:00
|
|
|
goto discard_it;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (skb->len < (th->doff<<2) || tcp_checksum_complete(skb)) {
|
2008-07-16 23:22:25 -04:00
|
|
|
TCP_INC_STATS_BH(net, TCP_MIB_INERRS);
|
2006-10-10 22:41:46 -04:00
|
|
|
inet_twsk_put(inet_twsk(sk));
|
2005-04-16 18:20:36 -04:00
|
|
|
goto discard_it;
|
|
|
|
}
|
|
|
|
|
2006-10-10 22:41:46 -04:00
|
|
|
switch (tcp_timewait_state_process(inet_twsk(sk), skb, th)) {
|
2005-04-16 18:20:36 -04:00
|
|
|
case TCP_TW_SYN:
|
|
|
|
{
|
|
|
|
struct sock *sk2;
|
|
|
|
|
2008-03-25 08:47:49 -04:00
|
|
|
sk2 = inet6_lookup_listener(dev_net(skb->dev), &tcp_hashinfo,
|
2007-04-25 20:54:47 -04:00
|
|
|
&ipv6_hdr(skb)->daddr,
|
2005-08-12 08:19:38 -04:00
|
|
|
ntohs(th->dest), inet6_iif(skb));
|
2005-04-16 18:20:36 -04:00
|
|
|
if (sk2 != NULL) {
|
2005-08-09 23:44:40 -04:00
|
|
|
struct inet_timewait_sock *tw = inet_twsk(sk);
|
|
|
|
inet_twsk_deschedule(tw, &tcp_death_row);
|
|
|
|
inet_twsk_put(tw);
|
2005-04-16 18:20:36 -04:00
|
|
|
sk = sk2;
|
|
|
|
goto process;
|
|
|
|
}
|
|
|
|
/* Fall through to ACK */
|
|
|
|
}
|
|
|
|
case TCP_TW_ACK:
|
|
|
|
tcp_v6_timewait_ack(sk, skb);
|
|
|
|
break;
|
|
|
|
case TCP_TW_RST:
|
|
|
|
goto no_tcp_socket;
|
|
|
|
case TCP_TW_SUCCESS:;
|
|
|
|
}
|
|
|
|
goto discard_it;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int tcp_v6_remember_stamp(struct sock *sk)
|
|
|
|
{
|
|
|
|
/* Alas, not yet... */
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2005-12-14 02:15:52 -05:00
|
|
|
static struct inet_connection_sock_af_ops ipv6_specific = {
|
2006-03-21 01:48:35 -05:00
|
|
|
.queue_xmit = inet6_csk_xmit,
|
|
|
|
.send_check = tcp_v6_send_check,
|
|
|
|
.rebuild_header = inet6_sk_rebuild_header,
|
|
|
|
.conn_request = tcp_v6_conn_request,
|
|
|
|
.syn_recv_sock = tcp_v6_syn_recv_sock,
|
|
|
|
.remember_stamp = tcp_v6_remember_stamp,
|
|
|
|
.net_header_len = sizeof(struct ipv6hdr),
|
|
|
|
.setsockopt = ipv6_setsockopt,
|
|
|
|
.getsockopt = ipv6_getsockopt,
|
|
|
|
.addr2sockaddr = inet6_csk_addr2sockaddr,
|
|
|
|
.sockaddr_len = sizeof(struct sockaddr_in6),
|
[SOCK] proto: Add hashinfo member to struct proto
This way we can remove TCP and DCCP specific versions of
sk->sk_prot->get_port: both v4 and v6 use inet_csk_get_port
sk->sk_prot->hash: inet_hash is directly used, only v6 need
a specific version to deal with mapped sockets
sk->sk_prot->unhash: both v4 and v6 use inet_hash directly
struct inet_connection_sock_af_ops also gets a new member, bind_conflict, so
that inet_csk_get_port can find the per family routine.
Now only the lookup routines receive as a parameter a struct inet_hashtable.
With this we further reuse code, reducing the difference among INET transport
protocols.
Eventually work has to be done on UDP and SCTP to make them share this
infrastructure and get as a bonus inet_diag interfaces so that iproute can be
used with these protocols.
net-2.6/net/ipv4/inet_hashtables.c:
struct proto | +8
struct inet_connection_sock_af_ops | +8
2 structs changed
__inet_hash_nolisten | +18
__inet_hash | -210
inet_put_port | +8
inet_bind_bucket_create | +1
__inet_hash_connect | -8
5 functions changed, 27 bytes added, 218 bytes removed, diff: -191
net-2.6/net/core/sock.c:
proto_seq_show | +3
1 function changed, 3 bytes added, diff: +3
net-2.6/net/ipv4/inet_connection_sock.c:
inet_csk_get_port | +15
1 function changed, 15 bytes added, diff: +15
net-2.6/net/ipv4/tcp.c:
tcp_set_state | -7
1 function changed, 7 bytes removed, diff: -7
net-2.6/net/ipv4/tcp_ipv4.c:
tcp_v4_get_port | -31
tcp_v4_hash | -48
tcp_v4_destroy_sock | -7
tcp_v4_syn_recv_sock | -2
tcp_unhash | -179
5 functions changed, 267 bytes removed, diff: -267
net-2.6/net/ipv6/inet6_hashtables.c:
__inet6_hash | +8
1 function changed, 8 bytes added, diff: +8
net-2.6/net/ipv4/inet_hashtables.c:
inet_unhash | +190
inet_hash | +242
2 functions changed, 432 bytes added, diff: +432
vmlinux:
16 functions changed, 485 bytes added, 492 bytes removed, diff: -7
/home/acme/git/net-2.6/net/ipv6/tcp_ipv6.c:
tcp_v6_get_port | -31
tcp_v6_hash | -7
tcp_v6_syn_recv_sock | -9
3 functions changed, 47 bytes removed, diff: -47
/home/acme/git/net-2.6/net/dccp/proto.c:
dccp_destroy_sock | -7
dccp_unhash | -179
dccp_hash | -49
dccp_set_state | -7
dccp_done | +1
5 functions changed, 1 bytes added, 242 bytes removed, diff: -241
/home/acme/git/net-2.6/net/dccp/ipv4.c:
dccp_v4_get_port | -31
dccp_v4_request_recv_sock | -2
2 functions changed, 33 bytes removed, diff: -33
/home/acme/git/net-2.6/net/dccp/ipv6.c:
dccp_v6_get_port | -31
dccp_v6_hash | -7
dccp_v6_request_recv_sock | +5
3 functions changed, 5 bytes added, 38 bytes removed, diff: -33
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-02-03 07:06:04 -05:00
|
|
|
.bind_conflict = inet6_csk_bind_conflict,
|
2006-03-21 01:45:21 -05:00
|
|
|
#ifdef CONFIG_COMPAT
|
2006-03-21 01:48:35 -05:00
|
|
|
.compat_setsockopt = compat_ipv6_setsockopt,
|
|
|
|
.compat_getsockopt = compat_ipv6_getsockopt,
|
2006-03-21 01:45:21 -05:00
|
|
|
#endif
|
2005-04-16 18:20:36 -04:00
|
|
|
};
|
|
|
|
|
2006-11-14 22:07:45 -05:00
|
|
|
#ifdef CONFIG_TCP_MD5SIG
|
2006-11-14 22:53:22 -05:00
|
|
|
static struct tcp_sock_af_ops tcp_sock_ipv6_specific = {
|
2006-11-14 22:07:45 -05:00
|
|
|
.md5_lookup = tcp_v6_md5_lookup,
|
2008-07-19 03:01:42 -04:00
|
|
|
.calc_md5_hash = tcp_v6_md5_hash_skb,
|
2006-11-14 22:07:45 -05:00
|
|
|
.md5_add = tcp_v6_md5_add_func,
|
|
|
|
.md5_parse = tcp_v6_parse_md5_keys,
|
|
|
|
};
|
2006-11-14 22:53:22 -05:00
|
|
|
#endif
|
2006-11-14 22:07:45 -05:00
|
|
|
|
2005-04-16 18:20:36 -04:00
|
|
|
/*
|
|
|
|
* TCP over IPv4 via INET6 API
|
|
|
|
*/
|
|
|
|
|
2005-12-14 02:15:52 -05:00
|
|
|
static struct inet_connection_sock_af_ops ipv6_mapped = {
|
2006-03-21 01:48:35 -05:00
|
|
|
.queue_xmit = ip_queue_xmit,
|
|
|
|
.send_check = tcp_v4_send_check,
|
|
|
|
.rebuild_header = inet_sk_rebuild_header,
|
|
|
|
.conn_request = tcp_v6_conn_request,
|
|
|
|
.syn_recv_sock = tcp_v6_syn_recv_sock,
|
|
|
|
.remember_stamp = tcp_v4_remember_stamp,
|
|
|
|
.net_header_len = sizeof(struct iphdr),
|
|
|
|
.setsockopt = ipv6_setsockopt,
|
|
|
|
.getsockopt = ipv6_getsockopt,
|
|
|
|
.addr2sockaddr = inet6_csk_addr2sockaddr,
|
|
|
|
.sockaddr_len = sizeof(struct sockaddr_in6),
|
[SOCK] proto: Add hashinfo member to struct proto
This way we can remove TCP and DCCP specific versions of
sk->sk_prot->get_port: both v4 and v6 use inet_csk_get_port
sk->sk_prot->hash: inet_hash is directly used, only v6 need
a specific version to deal with mapped sockets
sk->sk_prot->unhash: both v4 and v6 use inet_hash directly
struct inet_connection_sock_af_ops also gets a new member, bind_conflict, so
that inet_csk_get_port can find the per family routine.
Now only the lookup routines receive as a parameter a struct inet_hashtable.
With this we further reuse code, reducing the difference among INET transport
protocols.
Eventually work has to be done on UDP and SCTP to make them share this
infrastructure and get as a bonus inet_diag interfaces so that iproute can be
used with these protocols.
net-2.6/net/ipv4/inet_hashtables.c:
struct proto | +8
struct inet_connection_sock_af_ops | +8
2 structs changed
__inet_hash_nolisten | +18
__inet_hash | -210
inet_put_port | +8
inet_bind_bucket_create | +1
__inet_hash_connect | -8
5 functions changed, 27 bytes added, 218 bytes removed, diff: -191
net-2.6/net/core/sock.c:
proto_seq_show | +3
1 function changed, 3 bytes added, diff: +3
net-2.6/net/ipv4/inet_connection_sock.c:
inet_csk_get_port | +15
1 function changed, 15 bytes added, diff: +15
net-2.6/net/ipv4/tcp.c:
tcp_set_state | -7
1 function changed, 7 bytes removed, diff: -7
net-2.6/net/ipv4/tcp_ipv4.c:
tcp_v4_get_port | -31
tcp_v4_hash | -48
tcp_v4_destroy_sock | -7
tcp_v4_syn_recv_sock | -2
tcp_unhash | -179
5 functions changed, 267 bytes removed, diff: -267
net-2.6/net/ipv6/inet6_hashtables.c:
__inet6_hash | +8
1 function changed, 8 bytes added, diff: +8
net-2.6/net/ipv4/inet_hashtables.c:
inet_unhash | +190
inet_hash | +242
2 functions changed, 432 bytes added, diff: +432
vmlinux:
16 functions changed, 485 bytes added, 492 bytes removed, diff: -7
/home/acme/git/net-2.6/net/ipv6/tcp_ipv6.c:
tcp_v6_get_port | -31
tcp_v6_hash | -7
tcp_v6_syn_recv_sock | -9
3 functions changed, 47 bytes removed, diff: -47
/home/acme/git/net-2.6/net/dccp/proto.c:
dccp_destroy_sock | -7
dccp_unhash | -179
dccp_hash | -49
dccp_set_state | -7
dccp_done | +1
5 functions changed, 1 bytes added, 242 bytes removed, diff: -241
/home/acme/git/net-2.6/net/dccp/ipv4.c:
dccp_v4_get_port | -31
dccp_v4_request_recv_sock | -2
2 functions changed, 33 bytes removed, diff: -33
/home/acme/git/net-2.6/net/dccp/ipv6.c:
dccp_v6_get_port | -31
dccp_v6_hash | -7
dccp_v6_request_recv_sock | +5
3 functions changed, 5 bytes added, 38 bytes removed, diff: -33
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-02-03 07:06:04 -05:00
|
|
|
.bind_conflict = inet6_csk_bind_conflict,
|
2006-03-21 01:45:21 -05:00
|
|
|
#ifdef CONFIG_COMPAT
|
2006-03-21 01:48:35 -05:00
|
|
|
.compat_setsockopt = compat_ipv6_setsockopt,
|
|
|
|
.compat_getsockopt = compat_ipv6_getsockopt,
|
2006-03-21 01:45:21 -05:00
|
|
|
#endif
|
2005-04-16 18:20:36 -04:00
|
|
|
};
|
|
|
|
|
2006-11-14 22:07:45 -05:00
|
|
|
#ifdef CONFIG_TCP_MD5SIG
|
2006-11-14 22:53:22 -05:00
|
|
|
static struct tcp_sock_af_ops tcp_sock_ipv6_mapped_specific = {
|
2006-11-14 22:07:45 -05:00
|
|
|
.md5_lookup = tcp_v4_md5_lookup,
|
2008-07-19 03:01:42 -04:00
|
|
|
.calc_md5_hash = tcp_v4_md5_hash_skb,
|
2006-11-14 22:07:45 -05:00
|
|
|
.md5_add = tcp_v6_md5_add_func,
|
|
|
|
.md5_parse = tcp_v6_parse_md5_keys,
|
|
|
|
};
|
2006-11-14 22:53:22 -05:00
|
|
|
#endif
|
2006-11-14 22:07:45 -05:00
|
|
|
|
2005-04-16 18:20:36 -04:00
|
|
|
/* NOTE: A lot of things set to zero explicitly by call to
|
|
|
|
* sk_alloc() so need not be done here.
|
|
|
|
*/
|
|
|
|
static int tcp_v6_init_sock(struct sock *sk)
|
|
|
|
{
|
2005-08-10 03:03:31 -04:00
|
|
|
struct inet_connection_sock *icsk = inet_csk(sk);
|
2005-04-16 18:20:36 -04:00
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
|
|
|
|
|
|
|
skb_queue_head_init(&tp->out_of_order_queue);
|
|
|
|
tcp_init_xmit_timers(sk);
|
|
|
|
tcp_prequeue_init(tp);
|
|
|
|
|
2005-08-10 03:03:31 -04:00
|
|
|
icsk->icsk_rto = TCP_TIMEOUT_INIT;
|
2005-04-16 18:20:36 -04:00
|
|
|
tp->mdev = TCP_TIMEOUT_INIT;
|
|
|
|
|
|
|
|
/* So many TCP implementations out there (incorrectly) count the
|
|
|
|
* initial SYN frame in their delayed-ACK and congestion control
|
|
|
|
* algorithms that we must have the following bandaid to talk
|
|
|
|
* efficiently to them. -DaveM
|
|
|
|
*/
|
|
|
|
tp->snd_cwnd = 2;
|
|
|
|
|
|
|
|
/* See draft-stevens-tcpca-spec-01 for discussion of the
|
|
|
|
* initialization of these values.
|
|
|
|
*/
|
|
|
|
tp->snd_ssthresh = 0x7fffffff;
|
|
|
|
tp->snd_cwnd_clamp = ~0;
|
2005-07-05 18:24:38 -04:00
|
|
|
tp->mss_cache = 536;
|
2005-04-16 18:20:36 -04:00
|
|
|
|
|
|
|
tp->reordering = sysctl_tcp_reordering;
|
|
|
|
|
|
|
|
sk->sk_state = TCP_CLOSE;
|
|
|
|
|
2005-12-14 02:15:52 -05:00
|
|
|
icsk->icsk_af_ops = &ipv6_specific;
|
2005-08-10 03:03:31 -04:00
|
|
|
icsk->icsk_ca_ops = &tcp_init_congestion_ops;
|
2005-12-14 02:26:10 -05:00
|
|
|
icsk->icsk_sync_mss = tcp_sync_mss;
|
2005-04-16 18:20:36 -04:00
|
|
|
sk->sk_write_space = sk_stream_write_space;
|
|
|
|
sock_set_flag(sk, SOCK_USE_WRITE_QUEUE);
|
|
|
|
|
2006-11-14 22:07:45 -05:00
|
|
|
#ifdef CONFIG_TCP_MD5SIG
|
|
|
|
tp->af_specific = &tcp_sock_ipv6_specific;
|
|
|
|
#endif
|
|
|
|
|
2005-04-16 18:20:36 -04:00
|
|
|
sk->sk_sndbuf = sysctl_tcp_wmem[1];
|
|
|
|
sk->sk_rcvbuf = sysctl_tcp_rmem[1];
|
|
|
|
|
2008-12-30 02:04:08 -05:00
|
|
|
local_bh_disable();
|
2008-11-26 00:16:35 -05:00
|
|
|
percpu_counter_inc(&tcp_sockets_allocated);
|
2008-12-30 02:04:08 -05:00
|
|
|
local_bh_enable();
|
2005-04-16 18:20:36 -04:00
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2008-06-14 20:04:49 -04:00
|
|
|
static void tcp_v6_destroy_sock(struct sock *sk)
|
2005-04-16 18:20:36 -04:00
|
|
|
{
|
2006-11-14 22:07:45 -05:00
|
|
|
#ifdef CONFIG_TCP_MD5SIG
|
|
|
|
/* Clean up the MD5 key list */
|
|
|
|
if (tcp_sk(sk)->md5sig_info)
|
|
|
|
tcp_v6_clear_md5_list(sk);
|
|
|
|
#endif
|
2005-04-16 18:20:36 -04:00
|
|
|
tcp_v4_destroy_sock(sk);
|
2008-06-14 20:04:49 -04:00
|
|
|
inet6_destroy_sock(sk);
|
2005-04-16 18:20:36 -04:00
|
|
|
}
|
|
|
|
|
2007-04-21 07:13:44 -04:00
|
|
|
#ifdef CONFIG_PROC_FS
|
2005-04-16 18:20:36 -04:00
|
|
|
/* Proc filesystem TCPv6 sock list dumping. */
|
2007-02-09 09:24:49 -05:00
|
|
|
static void get_openreq6(struct seq_file *seq,
|
2005-06-19 01:47:21 -04:00
|
|
|
struct sock *sk, struct request_sock *req, int i, int uid)
|
2005-04-16 18:20:36 -04:00
|
|
|
{
|
|
|
|
int ttd = req->expires - jiffies;
|
2005-12-14 02:15:40 -05:00
|
|
|
struct in6_addr *src = &inet6_rsk(req)->loc_addr;
|
|
|
|
struct in6_addr *dest = &inet6_rsk(req)->rmt_addr;
|
2005-04-16 18:20:36 -04:00
|
|
|
|
|
|
|
if (ttd < 0)
|
|
|
|
ttd = 0;
|
|
|
|
|
|
|
|
seq_printf(seq,
|
|
|
|
"%4d: %08X%08X%08X%08X:%04X %08X%08X%08X%08X:%04X "
|
|
|
|
"%02X %08X:%08X %02X:%08lX %08X %5d %8d %d %d %p\n",
|
|
|
|
i,
|
|
|
|
src->s6_addr32[0], src->s6_addr32[1],
|
|
|
|
src->s6_addr32[2], src->s6_addr32[3],
|
2008-10-20 02:35:58 -04:00
|
|
|
ntohs(inet_rsk(req)->loc_port),
|
2005-04-16 18:20:36 -04:00
|
|
|
dest->s6_addr32[0], dest->s6_addr32[1],
|
|
|
|
dest->s6_addr32[2], dest->s6_addr32[3],
|
[NET] Generalise TCP's struct open_request minisock infrastructure
Kept this first changeset minimal, without changing existing names to
ease peer review.
Basicaly tcp_openreq_alloc now receives the or_calltable, that in turn
has two new members:
->slab, that replaces tcp_openreq_cachep
->obj_size, to inform the size of the openreq descendant for
a specific protocol
The protocol specific fields in struct open_request were moved to a
class hierarchy, with the things that are common to all connection
oriented PF_INET protocols in struct inet_request_sock, the TCP ones
in tcp_request_sock, that is an inet_request_sock, that is an
open_request.
I.e. this uses the same approach used for the struct sock class
hierarchy, with sk_prot indicating if the protocol wants to use the
open_request infrastructure by filling in sk_prot->rsk_prot with an
or_calltable.
Results? Performance is improved and TCP v4 now uses only 64 bytes per
open request minisock, down from 96 without this patch :-)
Next changeset will rename some of the structs, fields and functions
mentioned above, struct or_calltable is way unclear, better name it
struct request_sock_ops, s/struct open_request/struct request_sock/g,
etc.
Signed-off-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-06-19 01:46:52 -04:00
|
|
|
ntohs(inet_rsk(req)->rmt_port),
|
2005-04-16 18:20:36 -04:00
|
|
|
TCP_SYN_RECV,
|
|
|
|
0,0, /* could print option size, but that is af dependent. */
|
2007-02-09 09:24:49 -05:00
|
|
|
1, /* timers active (only the expire timer) */
|
|
|
|
jiffies_to_clock_t(ttd),
|
2005-04-16 18:20:36 -04:00
|
|
|
req->retrans,
|
|
|
|
uid,
|
2007-02-09 09:24:49 -05:00
|
|
|
0, /* non standard timer */
|
2005-04-16 18:20:36 -04:00
|
|
|
0, /* open_requests have no inode */
|
|
|
|
0, req);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void get_tcp6_sock(struct seq_file *seq, struct sock *sp, int i)
|
|
|
|
{
|
|
|
|
struct in6_addr *dest, *src;
|
|
|
|
__u16 destp, srcp;
|
|
|
|
int timer_active;
|
|
|
|
unsigned long timer_expires;
|
|
|
|
struct inet_sock *inet = inet_sk(sp);
|
|
|
|
struct tcp_sock *tp = tcp_sk(sp);
|
2005-08-09 23:10:42 -04:00
|
|
|
const struct inet_connection_sock *icsk = inet_csk(sp);
|
2005-04-16 18:20:36 -04:00
|
|
|
struct ipv6_pinfo *np = inet6_sk(sp);
|
|
|
|
|
|
|
|
dest = &np->daddr;
|
|
|
|
src = &np->rcv_saddr;
|
|
|
|
destp = ntohs(inet->dport);
|
|
|
|
srcp = ntohs(inet->sport);
|
2005-08-09 23:10:42 -04:00
|
|
|
|
|
|
|
if (icsk->icsk_pending == ICSK_TIME_RETRANS) {
|
2005-04-16 18:20:36 -04:00
|
|
|
timer_active = 1;
|
2005-08-09 23:10:42 -04:00
|
|
|
timer_expires = icsk->icsk_timeout;
|
|
|
|
} else if (icsk->icsk_pending == ICSK_TIME_PROBE0) {
|
2005-04-16 18:20:36 -04:00
|
|
|
timer_active = 4;
|
2005-08-09 23:10:42 -04:00
|
|
|
timer_expires = icsk->icsk_timeout;
|
2005-04-16 18:20:36 -04:00
|
|
|
} else if (timer_pending(&sp->sk_timer)) {
|
|
|
|
timer_active = 2;
|
|
|
|
timer_expires = sp->sk_timer.expires;
|
|
|
|
} else {
|
|
|
|
timer_active = 0;
|
|
|
|
timer_expires = jiffies;
|
|
|
|
}
|
|
|
|
|
|
|
|
seq_printf(seq,
|
|
|
|
"%4d: %08X%08X%08X%08X:%04X %08X%08X%08X%08X:%04X "
|
2008-06-27 23:00:19 -04:00
|
|
|
"%02X %08X:%08X %02X:%08lX %08X %5d %8d %lu %d %p %lu %lu %u %u %d\n",
|
2005-04-16 18:20:36 -04:00
|
|
|
i,
|
|
|
|
src->s6_addr32[0], src->s6_addr32[1],
|
|
|
|
src->s6_addr32[2], src->s6_addr32[3], srcp,
|
|
|
|
dest->s6_addr32[0], dest->s6_addr32[1],
|
|
|
|
dest->s6_addr32[2], dest->s6_addr32[3], destp,
|
2007-02-09 09:24:49 -05:00
|
|
|
sp->sk_state,
|
2006-06-27 16:29:00 -04:00
|
|
|
tp->write_seq-tp->snd_una,
|
|
|
|
(sp->sk_state == TCP_LISTEN) ? sp->sk_ack_backlog : (tp->rcv_nxt - tp->copied_seq),
|
2005-04-16 18:20:36 -04:00
|
|
|
timer_active,
|
|
|
|
jiffies_to_clock_t(timer_expires - jiffies),
|
2005-08-09 23:10:42 -04:00
|
|
|
icsk->icsk_retransmits,
|
2005-04-16 18:20:36 -04:00
|
|
|
sock_i_uid(sp),
|
2005-08-10 03:03:31 -04:00
|
|
|
icsk->icsk_probes_out,
|
2005-04-16 18:20:36 -04:00
|
|
|
sock_i_ino(sp),
|
|
|
|
atomic_read(&sp->sk_refcnt), sp,
|
2008-06-27 23:00:19 -04:00
|
|
|
jiffies_to_clock_t(icsk->icsk_rto),
|
|
|
|
jiffies_to_clock_t(icsk->icsk_ack.ato),
|
2005-08-09 23:10:42 -04:00
|
|
|
(icsk->icsk_ack.quick << 1 ) | icsk->icsk_ack.pingpong,
|
2005-04-16 18:20:36 -04:00
|
|
|
tp->snd_cwnd, tp->snd_ssthresh>=0xFFFF?-1:tp->snd_ssthresh
|
|
|
|
);
|
|
|
|
}
|
|
|
|
|
2007-02-09 09:24:49 -05:00
|
|
|
static void get_timewait6_sock(struct seq_file *seq,
|
2005-08-09 23:09:30 -04:00
|
|
|
struct inet_timewait_sock *tw, int i)
|
2005-04-16 18:20:36 -04:00
|
|
|
{
|
|
|
|
struct in6_addr *dest, *src;
|
|
|
|
__u16 destp, srcp;
|
2005-12-14 02:23:09 -05:00
|
|
|
struct inet6_timewait_sock *tw6 = inet6_twsk((struct sock *)tw);
|
2005-04-16 18:20:36 -04:00
|
|
|
int ttd = tw->tw_ttd - jiffies;
|
|
|
|
|
|
|
|
if (ttd < 0)
|
|
|
|
ttd = 0;
|
|
|
|
|
2005-12-14 02:23:09 -05:00
|
|
|
dest = &tw6->tw_v6_daddr;
|
|
|
|
src = &tw6->tw_v6_rcv_saddr;
|
2005-04-16 18:20:36 -04:00
|
|
|
destp = ntohs(tw->tw_dport);
|
|
|
|
srcp = ntohs(tw->tw_sport);
|
|
|
|
|
|
|
|
seq_printf(seq,
|
|
|
|
"%4d: %08X%08X%08X%08X:%04X %08X%08X%08X%08X:%04X "
|
|
|
|
"%02X %08X:%08X %02X:%08lX %08X %5d %8d %d %d %p\n",
|
|
|
|
i,
|
|
|
|
src->s6_addr32[0], src->s6_addr32[1],
|
|
|
|
src->s6_addr32[2], src->s6_addr32[3], srcp,
|
|
|
|
dest->s6_addr32[0], dest->s6_addr32[1],
|
|
|
|
dest->s6_addr32[2], dest->s6_addr32[3], destp,
|
|
|
|
tw->tw_substate, 0, 0,
|
|
|
|
3, jiffies_to_clock_t(ttd), 0, 0, 0, 0,
|
|
|
|
atomic_read(&tw->tw_refcnt), tw);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int tcp6_seq_show(struct seq_file *seq, void *v)
|
|
|
|
{
|
|
|
|
struct tcp_iter_state *st;
|
|
|
|
|
|
|
|
if (v == SEQ_START_TOKEN) {
|
|
|
|
seq_puts(seq,
|
|
|
|
" sl "
|
|
|
|
"local_address "
|
|
|
|
"remote_address "
|
|
|
|
"st tx_queue rx_queue tr tm->when retrnsmt"
|
|
|
|
" uid timeout inode\n");
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
st = seq->private;
|
|
|
|
|
|
|
|
switch (st->state) {
|
|
|
|
case TCP_SEQ_STATE_LISTENING:
|
|
|
|
case TCP_SEQ_STATE_ESTABLISHED:
|
|
|
|
get_tcp6_sock(seq, v, st->num);
|
|
|
|
break;
|
|
|
|
case TCP_SEQ_STATE_OPENREQ:
|
|
|
|
get_openreq6(seq, st->syn_wait_sk, v, st->num, st->uid);
|
|
|
|
break;
|
|
|
|
case TCP_SEQ_STATE_TIME_WAIT:
|
|
|
|
get_timewait6_sock(seq, v, st->num);
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
out:
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static struct tcp_seq_afinfo tcp6_seq_afinfo = {
|
|
|
|
.name = "tcp6",
|
|
|
|
.family = AF_INET6,
|
2008-04-14 01:13:53 -04:00
|
|
|
.seq_fops = {
|
|
|
|
.owner = THIS_MODULE,
|
|
|
|
},
|
2008-04-14 01:12:13 -04:00
|
|
|
.seq_ops = {
|
|
|
|
.show = tcp6_seq_show,
|
|
|
|
},
|
2005-04-16 18:20:36 -04:00
|
|
|
};
|
|
|
|
|
2008-03-21 07:14:45 -04:00
|
|
|
int tcp6_proc_init(struct net *net)
|
2005-04-16 18:20:36 -04:00
|
|
|
{
|
2008-03-21 07:14:45 -04:00
|
|
|
return tcp_proc_register(net, &tcp6_seq_afinfo);
|
2005-04-16 18:20:36 -04:00
|
|
|
}
|
|
|
|
|
2008-03-21 07:14:45 -04:00
|
|
|
void tcp6_proc_exit(struct net *net)
|
2005-04-16 18:20:36 -04:00
|
|
|
{
|
2008-03-21 07:14:45 -04:00
|
|
|
tcp_proc_unregister(net, &tcp6_seq_afinfo);
|
2005-04-16 18:20:36 -04:00
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
|
|
|
struct proto tcpv6_prot = {
|
|
|
|
.name = "TCPv6",
|
|
|
|
.owner = THIS_MODULE,
|
|
|
|
.close = tcp_close,
|
|
|
|
.connect = tcp_v6_connect,
|
|
|
|
.disconnect = tcp_disconnect,
|
2005-08-09 23:10:42 -04:00
|
|
|
.accept = inet_csk_accept,
|
2005-04-16 18:20:36 -04:00
|
|
|
.ioctl = tcp_ioctl,
|
|
|
|
.init = tcp_v6_init_sock,
|
|
|
|
.destroy = tcp_v6_destroy_sock,
|
|
|
|
.shutdown = tcp_shutdown,
|
|
|
|
.setsockopt = tcp_setsockopt,
|
|
|
|
.getsockopt = tcp_getsockopt,
|
|
|
|
.recvmsg = tcp_recvmsg,
|
|
|
|
.backlog_rcv = tcp_v6_do_rcv,
|
|
|
|
.hash = tcp_v6_hash,
|
[SOCK] proto: Add hashinfo member to struct proto
This way we can remove TCP and DCCP specific versions of
sk->sk_prot->get_port: both v4 and v6 use inet_csk_get_port
sk->sk_prot->hash: inet_hash is directly used, only v6 need
a specific version to deal with mapped sockets
sk->sk_prot->unhash: both v4 and v6 use inet_hash directly
struct inet_connection_sock_af_ops also gets a new member, bind_conflict, so
that inet_csk_get_port can find the per family routine.
Now only the lookup routines receive as a parameter a struct inet_hashtable.
With this we further reuse code, reducing the difference among INET transport
protocols.
Eventually work has to be done on UDP and SCTP to make them share this
infrastructure and get as a bonus inet_diag interfaces so that iproute can be
used with these protocols.
net-2.6/net/ipv4/inet_hashtables.c:
struct proto | +8
struct inet_connection_sock_af_ops | +8
2 structs changed
__inet_hash_nolisten | +18
__inet_hash | -210
inet_put_port | +8
inet_bind_bucket_create | +1
__inet_hash_connect | -8
5 functions changed, 27 bytes added, 218 bytes removed, diff: -191
net-2.6/net/core/sock.c:
proto_seq_show | +3
1 function changed, 3 bytes added, diff: +3
net-2.6/net/ipv4/inet_connection_sock.c:
inet_csk_get_port | +15
1 function changed, 15 bytes added, diff: +15
net-2.6/net/ipv4/tcp.c:
tcp_set_state | -7
1 function changed, 7 bytes removed, diff: -7
net-2.6/net/ipv4/tcp_ipv4.c:
tcp_v4_get_port | -31
tcp_v4_hash | -48
tcp_v4_destroy_sock | -7
tcp_v4_syn_recv_sock | -2
tcp_unhash | -179
5 functions changed, 267 bytes removed, diff: -267
net-2.6/net/ipv6/inet6_hashtables.c:
__inet6_hash | +8
1 function changed, 8 bytes added, diff: +8
net-2.6/net/ipv4/inet_hashtables.c:
inet_unhash | +190
inet_hash | +242
2 functions changed, 432 bytes added, diff: +432
vmlinux:
16 functions changed, 485 bytes added, 492 bytes removed, diff: -7
/home/acme/git/net-2.6/net/ipv6/tcp_ipv6.c:
tcp_v6_get_port | -31
tcp_v6_hash | -7
tcp_v6_syn_recv_sock | -9
3 functions changed, 47 bytes removed, diff: -47
/home/acme/git/net-2.6/net/dccp/proto.c:
dccp_destroy_sock | -7
dccp_unhash | -179
dccp_hash | -49
dccp_set_state | -7
dccp_done | +1
5 functions changed, 1 bytes added, 242 bytes removed, diff: -241
/home/acme/git/net-2.6/net/dccp/ipv4.c:
dccp_v4_get_port | -31
dccp_v4_request_recv_sock | -2
2 functions changed, 33 bytes removed, diff: -33
/home/acme/git/net-2.6/net/dccp/ipv6.c:
dccp_v6_get_port | -31
dccp_v6_hash | -7
dccp_v6_request_recv_sock | +5
3 functions changed, 5 bytes added, 38 bytes removed, diff: -33
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-02-03 07:06:04 -05:00
|
|
|
.unhash = inet_unhash,
|
|
|
|
.get_port = inet_csk_get_port,
|
2005-04-16 18:20:36 -04:00
|
|
|
.enter_memory_pressure = tcp_enter_memory_pressure,
|
|
|
|
.sockets_allocated = &tcp_sockets_allocated,
|
|
|
|
.memory_allocated = &tcp_memory_allocated,
|
|
|
|
.memory_pressure = &tcp_memory_pressure,
|
2005-08-09 23:11:41 -04:00
|
|
|
.orphan_count = &tcp_orphan_count,
|
2005-04-16 18:20:36 -04:00
|
|
|
.sysctl_mem = sysctl_tcp_mem,
|
|
|
|
.sysctl_wmem = sysctl_tcp_wmem,
|
|
|
|
.sysctl_rmem = sysctl_tcp_rmem,
|
|
|
|
.max_header = MAX_TCP_HEADER,
|
|
|
|
.obj_size = sizeof(struct tcp6_sock),
|
2008-11-16 22:40:17 -05:00
|
|
|
.slab_flags = SLAB_DESTROY_BY_RCU,
|
2005-12-14 02:25:19 -05:00
|
|
|
.twsk_prot = &tcp6_timewait_sock_ops,
|
2005-06-19 01:47:21 -04:00
|
|
|
.rsk_prot = &tcp6_request_sock_ops,
|
2008-03-22 19:50:58 -04:00
|
|
|
.h.hashinfo = &tcp_hashinfo,
|
2006-03-21 01:48:35 -05:00
|
|
|
#ifdef CONFIG_COMPAT
|
|
|
|
.compat_setsockopt = compat_tcp_setsockopt,
|
|
|
|
.compat_getsockopt = compat_tcp_getsockopt,
|
|
|
|
#endif
|
2005-04-16 18:20:36 -04:00
|
|
|
};
|
|
|
|
|
|
|
|
static struct inet6_protocol tcpv6_protocol = {
|
|
|
|
.handler = tcp_v6_rcv,
|
|
|
|
.err_handler = tcp_v6_err,
|
2006-07-08 16:34:56 -04:00
|
|
|
.gso_send_check = tcp_v6_gso_send_check,
|
2006-06-30 16:36:15 -04:00
|
|
|
.gso_segment = tcp_tso_segment,
|
2009-01-08 13:41:23 -05:00
|
|
|
.gro_receive = tcp6_gro_receive,
|
|
|
|
.gro_complete = tcp6_gro_complete,
|
2005-04-16 18:20:36 -04:00
|
|
|
.flags = INET6_PROTO_NOPOLICY|INET6_PROTO_FINAL,
|
|
|
|
};
|
|
|
|
|
|
|
|
static struct inet_protosw tcpv6_protosw = {
|
|
|
|
.type = SOCK_STREAM,
|
|
|
|
.protocol = IPPROTO_TCP,
|
|
|
|
.prot = &tcpv6_prot,
|
|
|
|
.ops = &inet6_stream_ops,
|
|
|
|
.capability = -1,
|
|
|
|
.no_check = 0,
|
2005-12-14 02:26:10 -05:00
|
|
|
.flags = INET_PROTOSW_PERMANENT |
|
|
|
|
INET_PROTOSW_ICSK,
|
2005-04-16 18:20:36 -04:00
|
|
|
};
|
|
|
|
|
2008-03-07 14:16:02 -05:00
|
|
|
static int tcpv6_net_init(struct net *net)
|
|
|
|
{
|
2008-04-03 17:28:30 -04:00
|
|
|
return inet_ctl_sock_create(&net->ipv6.tcp_sk, PF_INET6,
|
|
|
|
SOCK_RAW, IPPROTO_TCP, net);
|
2008-03-07 14:16:02 -05:00
|
|
|
}
|
|
|
|
|
|
|
|
static void tcpv6_net_exit(struct net *net)
|
|
|
|
{
|
2008-04-03 17:28:30 -04:00
|
|
|
inet_ctl_sock_destroy(net->ipv6.tcp_sk);
|
netns : fix kernel panic in timewait socket destruction
How to reproduce ?
- create a network namespace
- use tcp protocol and get timewait socket
- exit the network namespace
- after a moment (when the timewait socket is destroyed), the kernel
panics.
# BUG: unable to handle kernel NULL pointer dereference at
0000000000000007
IP: [<ffffffff821e394d>] inet_twdr_do_twkill_work+0x6e/0xb8
PGD 119985067 PUD 11c5c0067 PMD 0
Oops: 0000 [1] SMP
CPU 1
Modules linked in: ipv6 button battery ac loop dm_mod tg3 libphy ext3 jbd
edd fan thermal processor thermal_sys sg sata_svw libata dock serverworks
sd_mod scsi_mod ide_disk ide_core [last unloaded: freq_table]
Pid: 0, comm: swapper Not tainted 2.6.27-rc2 #3
RIP: 0010:[<ffffffff821e394d>] [<ffffffff821e394d>]
inet_twdr_do_twkill_work+0x6e/0xb8
RSP: 0018:ffff88011ff7fed0 EFLAGS: 00010246
RAX: ffffffffffffffff RBX: ffffffff82339420 RCX: ffff88011ff7ff30
RDX: 0000000000000001 RSI: ffff88011a4d03c0 RDI: ffff88011ac2fc00
RBP: ffffffff823392e0 R08: 0000000000000000 R09: ffff88002802a200
R10: ffff8800a5c4b000 R11: ffffffff823e4080 R12: ffff88011ac2fc00
R13: 0000000000000001 R14: 0000000000000001 R15: 0000000000000000
FS: 0000000041cbd940(0000) GS:ffff8800bff839c0(0000)
knlGS:0000000000000000
CS: 0010 DS: 0018 ES: 0018 CR0: 000000008005003b
CR2: 0000000000000007 CR3: 00000000bd87c000 CR4: 00000000000006e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
Process swapper (pid: 0, threadinfo ffff8800bff9e000, task
ffff88011ff76690)
Stack: ffffffff823392e0 0000000000000100 ffffffff821e3a3a
0000000000000008
0000000000000000 ffffffff821e3a61 ffff8800bff7c000 ffffffff8203c7e7
ffff88011ff7ff10 ffff88011ff7ff10 0000000000000021 ffffffff82351108
Call Trace:
<IRQ> [<ffffffff821e3a3a>] ? inet_twdr_hangman+0x0/0x9e
[<ffffffff821e3a61>] ? inet_twdr_hangman+0x27/0x9e
[<ffffffff8203c7e7>] ? run_timer_softirq+0x12c/0x193
[<ffffffff820390d1>] ? __do_softirq+0x5e/0xcd
[<ffffffff8200d08c>] ? call_softirq+0x1c/0x28
[<ffffffff8200e611>] ? do_softirq+0x2c/0x68
[<ffffffff8201a055>] ? smp_apic_timer_interrupt+0x8e/0xa9
[<ffffffff8200cad6>] ? apic_timer_interrupt+0x66/0x70
<EOI> [<ffffffff82011f4c>] ? default_idle+0x27/0x3b
[<ffffffff8200abbd>] ? cpu_idle+0x5f/0x7d
Code: e8 01 00 00 4c 89 e7 41 ff c5 e8 8d fd ff ff 49 8b 44 24 38 4c 89 e7
65 8b 14 25 24 00 00 00 89 d2 48 8b 80 e8 00 00 00 48 f7 d0 <48> 8b 04 d0
48 ff 40 58 e8 fc fc ff ff 48 89 df e8 c0 5f 04 00
RIP [<ffffffff821e394d>] inet_twdr_do_twkill_work+0x6e/0xb8
RSP <ffff88011ff7fed0>
CR2: 0000000000000007
This patch provides a function to purge all timewait sockets related
to a network namespace. The timewait sockets life cycle is not tied with
the network namespace, that means the timewait sockets stay alive while
the network namespace dies. The timewait sockets are for avoiding to
receive a duplicate packet from the network, if the network namespace is
freed, the network stack is removed, so no chance to receive any packets
from the outside world. Furthermore, having a pending destruction timer
on these sockets with a network namespace freed is not safe and will lead
to an oops if the timer callback which try to access data belonging to
the namespace like for example in:
inet_twdr_do_twkill_work
-> NET_INC_STATS_BH(twsk_net(tw), LINUX_MIB_TIMEWAITED);
Purging the timewait sockets at the network namespace destruction will:
1) speed up memory freeing for the namespace
2) fix kernel panic on asynchronous timewait destruction
Signed-off-by: Daniel Lezcano <dlezcano@fr.ibm.com>
Acked-by: Denis V. Lunev <den@openvz.org>
Acked-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-09-08 16:17:27 -04:00
|
|
|
inet_twsk_purge(net, &tcp_hashinfo, &tcp_death_row, AF_INET6);
|
2008-03-07 14:16:02 -05:00
|
|
|
}
|
|
|
|
|
|
|
|
static struct pernet_operations tcpv6_net_ops = {
|
|
|
|
.init = tcpv6_net_init,
|
|
|
|
.exit = tcpv6_net_exit,
|
|
|
|
};
|
|
|
|
|
2007-12-11 05:25:35 -05:00
|
|
|
int __init tcpv6_init(void)
|
2005-04-16 18:20:36 -04:00
|
|
|
{
|
2007-12-11 05:25:35 -05:00
|
|
|
int ret;
|
|
|
|
|
|
|
|
ret = inet6_add_protocol(&tcpv6_protocol, IPPROTO_TCP);
|
|
|
|
if (ret)
|
|
|
|
goto out;
|
|
|
|
|
2005-04-16 18:20:36 -04:00
|
|
|
/* register inet6 protocol */
|
2007-12-11 05:25:35 -05:00
|
|
|
ret = inet6_register_protosw(&tcpv6_protosw);
|
|
|
|
if (ret)
|
|
|
|
goto out_tcpv6_protocol;
|
|
|
|
|
2008-03-07 14:16:02 -05:00
|
|
|
ret = register_pernet_subsys(&tcpv6_net_ops);
|
2007-12-11 05:25:35 -05:00
|
|
|
if (ret)
|
|
|
|
goto out_tcpv6_protosw;
|
|
|
|
out:
|
|
|
|
return ret;
|
2006-01-11 18:53:04 -05:00
|
|
|
|
2007-12-11 05:25:35 -05:00
|
|
|
out_tcpv6_protocol:
|
|
|
|
inet6_del_protocol(&tcpv6_protocol, IPPROTO_TCP);
|
|
|
|
out_tcpv6_protosw:
|
|
|
|
inet6_unregister_protosw(&tcpv6_protosw);
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
2007-12-13 08:34:58 -05:00
|
|
|
void tcpv6_exit(void)
|
2007-12-11 05:25:35 -05:00
|
|
|
{
|
2008-03-07 14:16:02 -05:00
|
|
|
unregister_pernet_subsys(&tcpv6_net_ops);
|
2007-12-11 05:25:35 -05:00
|
|
|
inet6_unregister_protosw(&tcpv6_protosw);
|
|
|
|
inet6_del_protocol(&tcpv6_protocol, IPPROTO_TCP);
|
2005-04-16 18:20:36 -04:00
|
|
|
}
|