[net-next-2.6.git] / net / netfilter / xt_TCPMSS.c

/*
 * This is a module which is used for setting the MSS option in TCP packets.
 *
 * Copyright (C) 2000 Marc Boucher <marc@mbsi.ca>
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License version 2 as
 * published by the Free Software Foundation.
 */

#include <linux/module.h>
#include <linux/skbuff.h>
#include <linux/ip.h>
#include <linux/ipv6.h>
#include <linux/tcp.h>
#include <net/ipv6.h>
#include <net/tcp.h>

#include <linux/netfilter_ipv4/ip_tables.h>
#include <linux/netfilter_ipv6/ip6_tables.h>
#include <linux/netfilter/x_tables.h>
#include <linux/netfilter/xt_tcpudp.h>
#include <linux/netfilter/xt_TCPMSS.h>

MODULE_LICENSE("GPL");
MODULE_AUTHOR("Marc Boucher <marc@mbsi.ca>");
MODULE_DESCRIPTION("x_tables TCP MSS modification module");
MODULE_ALIAS("ipt_TCPMSS");
MODULE_ALIAS("ip6t_TCPMSS");

static inline unsigned int
optlen(const u_int8_t *opt, unsigned int offset)
{
	/* Beware zero-length options: make finite progress */
	if (opt[offset] <= TCPOPT_NOP || opt[offset+1] == 0)
		return 1;
	else
		return opt[offset+1];
}

static int
tcpmss_mangle_packet(struct sk_buff *skb,
		     const struct xt_tcpmss_info *info,
		     unsigned int tcphoff,
		     unsigned int minlen)
{
	struct tcphdr *tcph;
	unsigned int tcplen, i;
	__be16 oldval;
	u16 newmss;
	u8 *opt;

	if (!skb_make_writable(skb, skb->len))
		return -1;

	tcplen = skb->len - tcphoff;
	tcph = (struct tcphdr *)(skb_network_header(skb) + tcphoff);

	/* Since it passed flags test in tcp match, we know it is is
	   not a fragment, and has data >= tcp header length.  SYN
	   packets should not contain data: if they did, then we risk
	   running over MTU, sending Frag Needed and breaking things
	   badly. --RR */
	if (tcplen != tcph->doff*4) {
		if (net_ratelimit())
			printk(KERN_ERR "xt_TCPMSS: bad length (%u bytes)\n",
			       skb->len);
		return -1;
	}

	if (info->mss == XT_TCPMSS_CLAMP_PMTU) {
		if (dst_mtu(skb->dst) <= minlen) {
			if (net_ratelimit())
				printk(KERN_ERR "xt_TCPMSS: "
				       "unknown or invalid path-MTU (%u)\n",
				       dst_mtu(skb->dst));
			return -1;
		}
		newmss = dst_mtu(skb->dst) - minlen;
	} else
		newmss = info->mss;

	opt = (u_int8_t *)tcph;
	for (i = sizeof(struct tcphdr); i < tcph->doff*4; i += optlen(opt, i)) {
		if (opt[i] == TCPOPT_MSS && tcph->doff*4 - i >= TCPOLEN_MSS &&
		    opt[i+1] == TCPOLEN_MSS) {
			u_int16_t oldmss;

			oldmss = (opt[i+2] << 8) | opt[i+3];

			if (info->mss == XT_TCPMSS_CLAMP_PMTU &&
			    oldmss <= newmss)
				return 0;

			opt[i+2] = (newmss & 0xff00) >> 8;
			opt[i+3] = newmss & 0x00ff;

			inet_proto_csum_replace2(&tcph->check, skb,
						 htons(oldmss), htons(newmss),
						 0);
			return 0;
		}
	}

	/*
	 * MSS Option not found ?! add it..
	 */
	if (skb_tailroom(skb) < TCPOLEN_MSS) {
		if (pskb_expand_head(skb, 0,
				     TCPOLEN_MSS - skb_tailroom(skb),
				     GFP_ATOMIC))
			return -1;
		tcph = (struct tcphdr *)(skb_network_header(skb) + tcphoff);
	}

	skb_put(skb, TCPOLEN_MSS);

	opt = (u_int8_t *)tcph + sizeof(struct tcphdr);
	memmove(opt + TCPOLEN_MSS, opt, tcplen - sizeof(struct tcphdr));

	inet_proto_csum_replace2(&tcph->check, skb,
				 htons(tcplen), htons(tcplen + TCPOLEN_MSS), 1);
	opt[0] = TCPOPT_MSS;
	opt[1] = TCPOLEN_MSS;
	opt[2] = (newmss & 0xff00) >> 8;
	opt[3] = newmss & 0x00ff;

	inet_proto_csum_replace4(&tcph->check, skb, 0, *((__be32 *)opt), 0);

	oldval = ((__be16 *)tcph)[6];
	tcph->doff += TCPOLEN_MSS/4;
	inet_proto_csum_replace2(&tcph->check, skb,
				 oldval, ((__be16 *)tcph)[6], 0);
	return TCPOLEN_MSS;
}

static unsigned int
xt_tcpmss_target4(struct sk_buff *skb,
		  const struct net_device *in,
		  const struct net_device *out,
		  unsigned int hooknum,
		  const struct xt_target *target,
		  const void *targinfo)
{
	struct iphdr *iph = ip_hdr(skb);
	__be16 newlen;
	int ret;

	ret = tcpmss_mangle_packet(skb, targinfo, iph->ihl * 4,
				   sizeof(*iph) + sizeof(struct tcphdr));
	if (ret < 0)
		return NF_DROP;
	if (ret > 0) {
		iph = ip_hdr(skb);
		newlen = htons(ntohs(iph->tot_len) + ret);
		csum_replace2(&iph->check, iph->tot_len, newlen);
		iph->tot_len = newlen;
	}
	return XT_CONTINUE;
}

#if defined(CONFIG_IP6_NF_IPTABLES) || defined(CONFIG_IP6_NF_IPTABLES_MODULE)
static unsigned int
xt_tcpmss_target6(struct sk_buff *skb,
		  const struct net_device *in,
		  const struct net_device *out,
		  unsigned int hooknum,
		  const struct xt_target *target,
		  const void *targinfo)
{
	struct ipv6hdr *ipv6h = ipv6_hdr(skb);
	u8 nexthdr;
	int tcphoff;
	int ret;

	nexthdr = ipv6h->nexthdr;
	tcphoff = ipv6_skip_exthdr(skb, sizeof(*ipv6h), &nexthdr);
	if (tcphoff < 0)
		return NF_DROP;
	ret = tcpmss_mangle_packet(skb, targinfo, tcphoff,
				   sizeof(*ipv6h) + sizeof(struct tcphdr));
	if (ret < 0)
		return NF_DROP;
	if (ret > 0) {
		ipv6h = ipv6_hdr(skb);
		ipv6h->payload_len = htons(ntohs(ipv6h->payload_len) + ret);
	}
	return XT_CONTINUE;
}
#endif

#define TH_SYN 0x02

/* Must specify -p tcp --syn */
static inline bool find_syn_match(const struct xt_entry_match *m)
{
	const struct xt_tcp *tcpinfo = (const struct xt_tcp *)m->data;

	if (strcmp(m->u.kernel.match->name, "tcp") == 0 &&
	    tcpinfo->flg_cmp & TH_SYN &&
	    !(tcpinfo->invflags & XT_TCP_INV_FLAGS))
		return true;

	return false;
}

static bool
xt_tcpmss_checkentry4(const char *tablename,
		      const void *entry,
		      const struct xt_target *target,
		      void *targinfo,
		      unsigned int hook_mask)
{
	const struct xt_tcpmss_info *info = targinfo;
	const struct ipt_entry *e = entry;

	if (info->mss == XT_TCPMSS_CLAMP_PMTU &&
	    (hook_mask & ~((1 << NF_INET_FORWARD) |
			   (1 << NF_INET_LOCAL_OUT) |
			   (1 << NF_INET_POST_ROUTING))) != 0) {
		printk("xt_TCPMSS: path-MTU clamping only supported in "
		       "FORWARD, OUTPUT and POSTROUTING hooks\n");
		return false;
	}
	if (IPT_MATCH_ITERATE(e, find_syn_match))
		return true;
	printk("xt_TCPMSS: Only works on TCP SYN packets\n");
	return false;
}

#if defined(CONFIG_IP6_NF_IPTABLES) || defined(CONFIG_IP6_NF_IPTABLES_MODULE)
static bool
xt_tcpmss_checkentry6(const char *tablename,
		      const void *entry,
		      const struct xt_target *target,
		      void *targinfo,
		      unsigned int hook_mask)
{
	const struct xt_tcpmss_info *info = targinfo;
	const struct ip6t_entry *e = entry;

	if (info->mss == XT_TCPMSS_CLAMP_PMTU &&
	    (hook_mask & ~((1 << NF_INET_FORWARD) |
			   (1 << NF_INET_LOCAL_OUT) |
			   (1 << NF_INET_POST_ROUTING))) != 0) {
		printk("xt_TCPMSS: path-MTU clamping only supported in "
		       "FORWARD, OUTPUT and POSTROUTING hooks\n");
		return false;
	}
	if (IP6T_MATCH_ITERATE(e, find_syn_match))
		return true;
	printk("xt_TCPMSS: Only works on TCP SYN packets\n");
	return false;
}
#endif

static struct xt_target xt_tcpmss_reg[] __read_mostly = {
	{
		.family		= AF_INET,
		.name		= "TCPMSS",
		.checkentry	= xt_tcpmss_checkentry4,
		.target		= xt_tcpmss_target4,
		.targetsize	= sizeof(struct xt_tcpmss_info),
		.proto		= IPPROTO_TCP,
		.me		= THIS_MODULE,
	},
#if defined(CONFIG_IP6_NF_IPTABLES) || defined(CONFIG_IP6_NF_IPTABLES_MODULE)
	{
		.family		= AF_INET6,
		.name		= "TCPMSS",
		.checkentry	= xt_tcpmss_checkentry6,
		.target		= xt_tcpmss_target6,
		.targetsize	= sizeof(struct xt_tcpmss_info),
		.proto		= IPPROTO_TCP,
		.me		= THIS_MODULE,
	},
#endif
};

static int __init xt_tcpmss_init(void)
{
	return xt_register_targets(xt_tcpmss_reg, ARRAY_SIZE(xt_tcpmss_reg));
}

static void __exit xt_tcpmss_fini(void)
{
	xt_unregister_targets(xt_tcpmss_reg, ARRAY_SIZE(xt_tcpmss_reg));
}

module_init(xt_tcpmss_init);
module_exit(xt_tcpmss_fini);
Commit	Line	Data
cdd289a2 PM	1	/*
	2	* This is a module which is used for setting the MSS option in TCP packets.
	3	*
	4	* Copyright (C) 2000 Marc Boucher <marc@mbsi.ca>
	5	*
	6	* This program is free software; you can redistribute it and/or modify
	7	* it under the terms of the GNU General Public License version 2 as
	8	* published by the Free Software Foundation.
	9	*/
	10
	11	#include <linux/module.h>
	12	#include <linux/skbuff.h>
	13	#include <linux/ip.h>
	14	#include <linux/ipv6.h>
	15	#include <linux/tcp.h>
	16	#include <net/ipv6.h>
	17	#include <net/tcp.h>
	18
	19	#include <linux/netfilter_ipv4/ip_tables.h>
	20	#include <linux/netfilter_ipv6/ip6_tables.h>
	21	#include <linux/netfilter/x_tables.h>
	22	#include <linux/netfilter/xt_tcpudp.h>
	23	#include <linux/netfilter/xt_TCPMSS.h>
	24
	25	MODULE_LICENSE("GPL");
	26	MODULE_AUTHOR("Marc Boucher <marc@mbsi.ca>");
	27	MODULE_DESCRIPTION("x_tables TCP MSS modification module");
	28	MODULE_ALIAS("ipt_TCPMSS");
	29	MODULE_ALIAS("ip6t_TCPMSS");
	30
	31	static inline unsigned int
	32	optlen(const u_int8_t *opt, unsigned int offset)
	33	{
	34	/* Beware zero-length options: make finite progress */
	35	if (opt[offset] <= TCPOPT_NOP \|\| opt[offset+1] == 0)
	36	return 1;
	37	else
	38	return opt[offset+1];
	39	}
	40
	41	static int
3db05fea	42	tcpmss_mangle_packet(struct sk_buff *skb,
cdd289a2 PM	43	const struct xt_tcpmss_info *info,
	44	unsigned int tcphoff,
	45	unsigned int minlen)
	46	{
	47	struct tcphdr *tcph;
	48	unsigned int tcplen, i;
	49	__be16 oldval;
	50	u16 newmss;
	51	u8 *opt;
	52
3db05fea	53	if (!skb_make_writable(skb, skb->len))
cdd289a2 PM	54	return -1;
cdd289a2 PM	55
3db05fea HX	56	tcplen = skb->len - tcphoff;
3db05fea HX	57	tcph = (struct tcphdr *)(skb_network_header(skb) + tcphoff);
cdd289a2 PM	58
	59	/* Since it passed flags test in tcp match, we know it is is
	60	not a fragment, and has data >= tcp header length. SYN
	61	packets should not contain data: if they did, then we risk
	62	running over MTU, sending Frag Needed and breaking things
	63	badly. --RR */
	64	if (tcplen != tcph->doff*4) {
	65	if (net_ratelimit())
	66	printk(KERN_ERR "xt_TCPMSS: bad length (%u bytes)\n",
3db05fea	67	skb->len);
cdd289a2 PM	68	return -1;
	69	}
	70
	71	if (info->mss == XT_TCPMSS_CLAMP_PMTU) {
3db05fea	72	if (dst_mtu(skb->dst) <= minlen) {
cdd289a2 PM	73	if (net_ratelimit())
	74	printk(KERN_ERR "xt_TCPMSS: "
	75	"unknown or invalid path-MTU (%u)\n",
3db05fea	76	dst_mtu(skb->dst));
cdd289a2 PM	77	return -1;
cdd289a2 PM	78	}
3db05fea	79	newmss = dst_mtu(skb->dst) - minlen;
cdd289a2 PM	80	} else
	81	newmss = info->mss;
	82
	83	opt = (u_int8_t *)tcph;
	84	for (i = sizeof(struct tcphdr); i < tcph->doff*4; i += optlen(opt, i)) {
	85	if (opt[i] == TCPOPT_MSS && tcph->doff*4 - i >= TCPOLEN_MSS &&
	86	opt[i+1] == TCPOLEN_MSS) {
	87	u_int16_t oldmss;
	88
	89	oldmss = (opt[i+2] << 8) \| opt[i+3];
	90
	91	if (info->mss == XT_TCPMSS_CLAMP_PMTU &&
	92	oldmss <= newmss)
	93	return 0;
	94
	95	opt[i+2] = (newmss & 0xff00) >> 8;
7c4e36bc	96	opt[i+3] = newmss & 0x00ff;
cdd289a2	97
be0ea7d5 PM	98	inet_proto_csum_replace2(&tcph->check, skb,
	99	htons(oldmss), htons(newmss),
	100	0);
cdd289a2 PM	101	return 0;
	102	}
	103	}
	104
	105	/*
	106	* MSS Option not found ?! add it..
	107	*/
3db05fea HX	108	if (skb_tailroom(skb) < TCPOLEN_MSS) {
	109	if (pskb_expand_head(skb, 0,
	110	TCPOLEN_MSS - skb_tailroom(skb),
2ca7b0ac	111	GFP_ATOMIC))
cdd289a2	112	return -1;
3db05fea	113	tcph = (struct tcphdr *)(skb_network_header(skb) + tcphoff);
cdd289a2 PM	114	}
cdd289a2 PM	115
3db05fea	116	skb_put(skb, TCPOLEN_MSS);
cdd289a2 PM	117
	118	opt = (u_int8_t *)tcph + sizeof(struct tcphdr);
	119	memmove(opt + TCPOLEN_MSS, opt, tcplen - sizeof(struct tcphdr));
	120
be0ea7d5 PM	121	inet_proto_csum_replace2(&tcph->check, skb,
be0ea7d5 PM	122	htons(tcplen), htons(tcplen + TCPOLEN_MSS), 1);
cdd289a2 PM	123	opt[0] = TCPOPT_MSS;
	124	opt[1] = TCPOLEN_MSS;
	125	opt[2] = (newmss & 0xff00) >> 8;
7c4e36bc	126	opt[3] = newmss & 0x00ff;
cdd289a2	127
be0ea7d5	128	inet_proto_csum_replace4(&tcph->check, skb, 0, ((__be32 )opt), 0);
cdd289a2 PM	129
	130	oldval = ((__be16 *)tcph)[6];
	131	tcph->doff += TCPOLEN_MSS/4;
be0ea7d5 PM	132	inet_proto_csum_replace2(&tcph->check, skb,
be0ea7d5 PM	133	oldval, ((__be16 *)tcph)[6], 0);
cdd289a2 PM	134	return TCPOLEN_MSS;
	135	}
	136
	137	static unsigned int
3db05fea	138	xt_tcpmss_target4(struct sk_buff *skb,
cdd289a2 PM	139	const struct net_device *in,
	140	const struct net_device *out,
	141	unsigned int hooknum,
	142	const struct xt_target *target,
	143	const void *targinfo)
	144	{
3db05fea	145	struct iphdr *iph = ip_hdr(skb);
cdd289a2 PM	146	__be16 newlen;
	147	int ret;
	148
3db05fea	149	ret = tcpmss_mangle_packet(skb, targinfo, iph->ihl * 4,
cdd289a2 PM	150	sizeof(*iph) + sizeof(struct tcphdr));
	151	if (ret < 0)
	152	return NF_DROP;
	153	if (ret > 0) {
3db05fea	154	iph = ip_hdr(skb);
cdd289a2	155	newlen = htons(ntohs(iph->tot_len) + ret);
be0ea7d5	156	csum_replace2(&iph->check, iph->tot_len, newlen);
cdd289a2 PM	157	iph->tot_len = newlen;
	158	}
	159	return XT_CONTINUE;
	160	}
	161
	162	#if defined(CONFIG_IP6_NF_IPTABLES) \|\| defined(CONFIG_IP6_NF_IPTABLES_MODULE)
	163	static unsigned int
3db05fea	164	xt_tcpmss_target6(struct sk_buff *skb,
cdd289a2 PM	165	const struct net_device *in,
	166	const struct net_device *out,
	167	unsigned int hooknum,
	168	const struct xt_target *target,
	169	const void *targinfo)
	170	{
3db05fea	171	struct ipv6hdr *ipv6h = ipv6_hdr(skb);
cdd289a2 PM	172	u8 nexthdr;
	173	int tcphoff;
	174	int ret;
	175
	176	nexthdr = ipv6h->nexthdr;
3db05fea	177	tcphoff = ipv6_skip_exthdr(skb, sizeof(*ipv6h), &nexthdr);
9dc0564e	178	if (tcphoff < 0)
cdd289a2	179	return NF_DROP;
3db05fea	180	ret = tcpmss_mangle_packet(skb, targinfo, tcphoff,
cdd289a2 PM	181	sizeof(*ipv6h) + sizeof(struct tcphdr));
	182	if (ret < 0)
	183	return NF_DROP;
	184	if (ret > 0) {
3db05fea	185	ipv6h = ipv6_hdr(skb);
cdd289a2 PM	186	ipv6h->payload_len = htons(ntohs(ipv6h->payload_len) + ret);
	187	}
	188	return XT_CONTINUE;
	189	}
	190	#endif
	191
	192	#define TH_SYN 0x02
	193
	194	/* Must specify -p tcp --syn */
e1931b78	195	static inline bool find_syn_match(const struct xt_entry_match *m)
cdd289a2 PM	196	{
	197	const struct xt_tcp tcpinfo = (const struct xt_tcp )m->data;
	198
	199	if (strcmp(m->u.kernel.match->name, "tcp") == 0 &&
	200	tcpinfo->flg_cmp & TH_SYN &&
	201	!(tcpinfo->invflags & XT_TCP_INV_FLAGS))
e1931b78	202	return true;
cdd289a2	203
e1931b78	204	return false;
cdd289a2 PM	205	}
cdd289a2 PM	206
e1931b78	207	static bool
cdd289a2 PM	208	xt_tcpmss_checkentry4(const char *tablename,
	209	const void *entry,
	210	const struct xt_target *target,
	211	void *targinfo,
	212	unsigned int hook_mask)
	213	{
	214	const struct xt_tcpmss_info *info = targinfo;
	215	const struct ipt_entry *e = entry;
	216
	217	if (info->mss == XT_TCPMSS_CLAMP_PMTU &&
6e23ae2a PM	218	(hook_mask & ~((1 << NF_INET_FORWARD) \|
	219	(1 << NF_INET_LOCAL_OUT) \|
	220	(1 << NF_INET_POST_ROUTING))) != 0) {
cdd289a2 PM	221	printk("xt_TCPMSS: path-MTU clamping only supported in "
cdd289a2 PM	222	"FORWARD, OUTPUT and POSTROUTING hooks\n");
e1931b78	223	return false;
cdd289a2 PM	224	}
cdd289a2 PM	225	if (IPT_MATCH_ITERATE(e, find_syn_match))
e1931b78	226	return true;
cdd289a2	227	printk("xt_TCPMSS: Only works on TCP SYN packets\n");
e1931b78	228	return false;
cdd289a2 PM	229	}
	230
	231	#if defined(CONFIG_IP6_NF_IPTABLES) \|\| defined(CONFIG_IP6_NF_IPTABLES_MODULE)
e1931b78	232	static bool
cdd289a2 PM	233	xt_tcpmss_checkentry6(const char *tablename,
	234	const void *entry,
	235	const struct xt_target *target,
	236	void *targinfo,
	237	unsigned int hook_mask)
	238	{
	239	const struct xt_tcpmss_info *info = targinfo;
	240	const struct ip6t_entry *e = entry;
	241
	242	if (info->mss == XT_TCPMSS_CLAMP_PMTU &&
6e23ae2a PM	243	(hook_mask & ~((1 << NF_INET_FORWARD) \|
	244	(1 << NF_INET_LOCAL_OUT) \|
	245	(1 << NF_INET_POST_ROUTING))) != 0) {
cdd289a2 PM	246	printk("xt_TCPMSS: path-MTU clamping only supported in "
cdd289a2 PM	247	"FORWARD, OUTPUT and POSTROUTING hooks\n");
e1931b78	248	return false;
cdd289a2 PM	249	}
cdd289a2 PM	250	if (IP6T_MATCH_ITERATE(e, find_syn_match))
e1931b78	251	return true;
cdd289a2	252	printk("xt_TCPMSS: Only works on TCP SYN packets\n");
e1931b78	253	return false;
cdd289a2 PM	254	}
	255	#endif
	256
9f15c530	257	static struct xt_target xt_tcpmss_reg[] __read_mostly = {
cdd289a2 PM	258	{
	259	.family = AF_INET,
	260	.name = "TCPMSS",
	261	.checkentry = xt_tcpmss_checkentry4,
	262	.target = xt_tcpmss_target4,
	263	.targetsize = sizeof(struct xt_tcpmss_info),
	264	.proto = IPPROTO_TCP,
	265	.me = THIS_MODULE,
	266	},
	267	#if defined(CONFIG_IP6_NF_IPTABLES) \|\| defined(CONFIG_IP6_NF_IPTABLES_MODULE)
	268	{
	269	.family = AF_INET6,
	270	.name = "TCPMSS",
	271	.checkentry = xt_tcpmss_checkentry6,
	272	.target = xt_tcpmss_target6,
	273	.targetsize = sizeof(struct xt_tcpmss_info),
	274	.proto = IPPROTO_TCP,
	275	.me = THIS_MODULE,
	276	},
	277	#endif
	278	};
	279
	280	static int __init xt_tcpmss_init(void)
	281	{
	282	return xt_register_targets(xt_tcpmss_reg, ARRAY_SIZE(xt_tcpmss_reg));
	283	}
	284
	285	static void __exit xt_tcpmss_fini(void)
	286	{
	287	xt_unregister_targets(xt_tcpmss_reg, ARRAY_SIZE(xt_tcpmss_reg));
	288	}
	289
	290	module_init(xt_tcpmss_init);
	291	module_exit(xt_tcpmss_fini);