[linux-2.6-block.git] / net / sched / act_csum.c

/*
 * Checksum updating actions
 *
 * Copyright (c) 2010 Gregoire Baron <baronchon@n7mm.org>
 *
 * 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.
 *
 */

#include <linux/types.h>
#include <linux/init.h>
#include <linux/kernel.h>
#include <linux/module.h>
#include <linux/spinlock.h>

#include <linux/netlink.h>
#include <net/netlink.h>
#include <linux/rtnetlink.h>

#include <linux/skbuff.h>

#include <net/ip.h>
#include <net/ipv6.h>
#include <net/icmp.h>
#include <linux/icmpv6.h>
#include <linux/igmp.h>
#include <net/tcp.h>
#include <net/udp.h>
#include <net/ip6_checksum.h>

#include <net/act_api.h>

#include <linux/tc_act/tc_csum.h>
#include <net/tc_act/tc_csum.h>

#define CSUM_TAB_MASK 15

static const struct nla_policy csum_policy[TCA_CSUM_MAX + 1] = {
	[TCA_CSUM_PARMS] = { .len = sizeof(struct tc_csum), },
};

static int csum_net_id;

static int tcf_csum_init(struct net *net, struct nlattr *nla,
			 struct nlattr *est, struct tc_action *a, int ovr,
			 int bind)
{
	struct tc_action_net *tn = net_generic(net, csum_net_id);
	struct nlattr *tb[TCA_CSUM_MAX + 1];
	struct tc_csum *parm;
	struct tcf_csum *p;
	int ret = 0, err;

	if (nla == NULL)
		return -EINVAL;

	err = nla_parse_nested(tb, TCA_CSUM_MAX, nla, csum_policy);
	if (err < 0)
		return err;

	if (tb[TCA_CSUM_PARMS] == NULL)
		return -EINVAL;
	parm = nla_data(tb[TCA_CSUM_PARMS]);

	if (!tcf_hash_check(tn, parm->index, a, bind)) {
		ret = tcf_hash_create(tn, parm->index, est, a,
				      sizeof(*p), bind, false);
		if (ret)
			return ret;
		ret = ACT_P_CREATED;
	} else {
		if (bind)/* dont override defaults */
			return 0;
		tcf_hash_release(a, bind);
		if (!ovr)
			return -EEXIST;
	}

	p = to_tcf_csum(a);
	spin_lock_bh(&p->tcf_lock);
	p->tcf_action = parm->action;
	p->update_flags = parm->update_flags;
	spin_unlock_bh(&p->tcf_lock);

	if (ret == ACT_P_CREATED)
		tcf_hash_insert(tn, a);

	return ret;
}

/**
 * tcf_csum_skb_nextlayer - Get next layer pointer
 * @skb: sk_buff to use
 * @ihl: previous summed headers length
 * @ipl: complete packet length
 * @jhl: next header length
 *
 * Check the expected next layer availability in the specified sk_buff.
 * Return the next layer pointer if pass, NULL otherwise.
 */
static void *tcf_csum_skb_nextlayer(struct sk_buff *skb,
				    unsigned int ihl, unsigned int ipl,
				    unsigned int jhl)
{
	int ntkoff = skb_network_offset(skb);
	int hl = ihl + jhl;

	if (!pskb_may_pull(skb, ipl + ntkoff) || (ipl < hl) ||
	    skb_try_make_writable(skb, hl + ntkoff))
		return NULL;
	else
		return (void *)(skb_network_header(skb) + ihl);
}

static int tcf_csum_ipv4_icmp(struct sk_buff *skb,
			      unsigned int ihl, unsigned int ipl)
{
	struct icmphdr *icmph;

	icmph = tcf_csum_skb_nextlayer(skb, ihl, ipl, sizeof(*icmph));
	if (icmph == NULL)
		return 0;

	icmph->checksum = 0;
	skb->csum = csum_partial(icmph, ipl - ihl, 0);
	icmph->checksum = csum_fold(skb->csum);

	skb->ip_summed = CHECKSUM_NONE;

	return 1;
}

static int tcf_csum_ipv4_igmp(struct sk_buff *skb,
			      unsigned int ihl, unsigned int ipl)
{
	struct igmphdr *igmph;

	igmph = tcf_csum_skb_nextlayer(skb, ihl, ipl, sizeof(*igmph));
	if (igmph == NULL)
		return 0;

	igmph->csum = 0;
	skb->csum = csum_partial(igmph, ipl - ihl, 0);
	igmph->csum = csum_fold(skb->csum);

	skb->ip_summed = CHECKSUM_NONE;

	return 1;
}

static int tcf_csum_ipv6_icmp(struct sk_buff *skb,
			      unsigned int ihl, unsigned int ipl)
{
	struct icmp6hdr *icmp6h;
	const struct ipv6hdr *ip6h;

	icmp6h = tcf_csum_skb_nextlayer(skb, ihl, ipl, sizeof(*icmp6h));
	if (icmp6h == NULL)
		return 0;

	ip6h = ipv6_hdr(skb);
	icmp6h->icmp6_cksum = 0;
	skb->csum = csum_partial(icmp6h, ipl - ihl, 0);
	icmp6h->icmp6_cksum = csum_ipv6_magic(&ip6h->saddr, &ip6h->daddr,
					      ipl - ihl, IPPROTO_ICMPV6,
					      skb->csum);

	skb->ip_summed = CHECKSUM_NONE;

	return 1;
}

static int tcf_csum_ipv4_tcp(struct sk_buff *skb,
			     unsigned int ihl, unsigned int ipl)
{
	struct tcphdr *tcph;
	const struct iphdr *iph;

	tcph = tcf_csum_skb_nextlayer(skb, ihl, ipl, sizeof(*tcph));
	if (tcph == NULL)
		return 0;

	iph = ip_hdr(skb);
	tcph->check = 0;
	skb->csum = csum_partial(tcph, ipl - ihl, 0);
	tcph->check = tcp_v4_check(ipl - ihl,
				   iph->saddr, iph->daddr, skb->csum);

	skb->ip_summed = CHECKSUM_NONE;

	return 1;
}

static int tcf_csum_ipv6_tcp(struct sk_buff *skb,
			     unsigned int ihl, unsigned int ipl)
{
	struct tcphdr *tcph;
	const struct ipv6hdr *ip6h;

	tcph = tcf_csum_skb_nextlayer(skb, ihl, ipl, sizeof(*tcph));
	if (tcph == NULL)
		return 0;

	ip6h = ipv6_hdr(skb);
	tcph->check = 0;
	skb->csum = csum_partial(tcph, ipl - ihl, 0);
	tcph->check = csum_ipv6_magic(&ip6h->saddr, &ip6h->daddr,
				      ipl - ihl, IPPROTO_TCP,
				      skb->csum);

	skb->ip_summed = CHECKSUM_NONE;

	return 1;
}

static int tcf_csum_ipv4_udp(struct sk_buff *skb,
			     unsigned int ihl, unsigned int ipl, int udplite)
{
	struct udphdr *udph;
	const struct iphdr *iph;
	u16 ul;

	/*
	 * Support both UDP and UDPLITE checksum algorithms, Don't use
	 * udph->len to get the real length without any protocol check,
	 * UDPLITE uses udph->len for another thing,
	 * Use iph->tot_len, or just ipl.
	 */

	udph = tcf_csum_skb_nextlayer(skb, ihl, ipl, sizeof(*udph));
	if (udph == NULL)
		return 0;

	iph = ip_hdr(skb);
	ul = ntohs(udph->len);

	if (udplite || udph->check) {

		udph->check = 0;

		if (udplite) {
			if (ul == 0)
				skb->csum = csum_partial(udph, ipl - ihl, 0);
			else if ((ul >= sizeof(*udph)) && (ul <= ipl - ihl))
				skb->csum = csum_partial(udph, ul, 0);
			else
				goto ignore_obscure_skb;
		} else {
			if (ul != ipl - ihl)
				goto ignore_obscure_skb;

			skb->csum = csum_partial(udph, ul, 0);
		}

		udph->check = csum_tcpudp_magic(iph->saddr, iph->daddr,
						ul, iph->protocol,
						skb->csum);

		if (!udph->check)
			udph->check = CSUM_MANGLED_0;
	}

	skb->ip_summed = CHECKSUM_NONE;

ignore_obscure_skb:
	return 1;
}

static int tcf_csum_ipv6_udp(struct sk_buff *skb,
			     unsigned int ihl, unsigned int ipl, int udplite)
{
	struct udphdr *udph;
	const struct ipv6hdr *ip6h;
	u16 ul;

	/*
	 * Support both UDP and UDPLITE checksum algorithms, Don't use
	 * udph->len to get the real length without any protocol check,
	 * UDPLITE uses udph->len for another thing,
	 * Use ip6h->payload_len + sizeof(*ip6h) ... , or just ipl.
	 */

	udph = tcf_csum_skb_nextlayer(skb, ihl, ipl, sizeof(*udph));
	if (udph == NULL)
		return 0;

	ip6h = ipv6_hdr(skb);
	ul = ntohs(udph->len);

	udph->check = 0;

	if (udplite) {
		if (ul == 0)
			skb->csum = csum_partial(udph, ipl - ihl, 0);

		else if ((ul >= sizeof(*udph)) && (ul <= ipl - ihl))
			skb->csum = csum_partial(udph, ul, 0);

		else
			goto ignore_obscure_skb;
	} else {
		if (ul != ipl - ihl)
			goto ignore_obscure_skb;

		skb->csum = csum_partial(udph, ul, 0);
	}

	udph->check = csum_ipv6_magic(&ip6h->saddr, &ip6h->daddr, ul,
				      udplite ? IPPROTO_UDPLITE : IPPROTO_UDP,
				      skb->csum);

	if (!udph->check)
		udph->check = CSUM_MANGLED_0;

	skb->ip_summed = CHECKSUM_NONE;

ignore_obscure_skb:
	return 1;
}

static int tcf_csum_ipv4(struct sk_buff *skb, u32 update_flags)
{
	const struct iphdr *iph;
	int ntkoff;

	ntkoff = skb_network_offset(skb);

	if (!pskb_may_pull(skb, sizeof(*iph) + ntkoff))
		goto fail;

	iph = ip_hdr(skb);

	switch (iph->frag_off & htons(IP_OFFSET) ? 0 : iph->protocol) {
	case IPPROTO_ICMP:
		if (update_flags & TCA_CSUM_UPDATE_FLAG_ICMP)
			if (!tcf_csum_ipv4_icmp(skb, iph->ihl * 4,
						ntohs(iph->tot_len)))
				goto fail;
		break;
	case IPPROTO_IGMP:
		if (update_flags & TCA_CSUM_UPDATE_FLAG_IGMP)
			if (!tcf_csum_ipv4_igmp(skb, iph->ihl * 4,
						ntohs(iph->tot_len)))
				goto fail;
		break;
	case IPPROTO_TCP:
		if (update_flags & TCA_CSUM_UPDATE_FLAG_TCP)
			if (!tcf_csum_ipv4_tcp(skb, iph->ihl * 4,
					       ntohs(iph->tot_len)))
				goto fail;
		break;
	case IPPROTO_UDP:
		if (update_flags & TCA_CSUM_UPDATE_FLAG_UDP)
			if (!tcf_csum_ipv4_udp(skb, iph->ihl * 4,
					       ntohs(iph->tot_len), 0))
				goto fail;
		break;
	case IPPROTO_UDPLITE:
		if (update_flags & TCA_CSUM_UPDATE_FLAG_UDPLITE)
			if (!tcf_csum_ipv4_udp(skb, iph->ihl * 4,
					       ntohs(iph->tot_len), 1))
				goto fail;
		break;
	}

	if (update_flags & TCA_CSUM_UPDATE_FLAG_IPV4HDR) {
		if (skb_try_make_writable(skb, sizeof(*iph) + ntkoff))
			goto fail;

		ip_send_check(ip_hdr(skb));
	}

	return 1;

fail:
	return 0;
}

static int tcf_csum_ipv6_hopopts(struct ipv6_opt_hdr *ip6xh,
				 unsigned int ixhl, unsigned int *pl)
{
	int off, len, optlen;
	unsigned char *xh = (void *)ip6xh;

	off = sizeof(*ip6xh);
	len = ixhl - off;

	while (len > 1) {
		switch (xh[off]) {
		case IPV6_TLV_PAD1:
			optlen = 1;
			break;
		case IPV6_TLV_JUMBO:
			optlen = xh[off + 1] + 2;
			if (optlen != 6 || len < 6 || (off & 3) != 2)
				/* wrong jumbo option length/alignment */
				return 0;
			*pl = ntohl(*(__be32 *)(xh + off + 2));
			goto done;
		default:
			optlen = xh[off + 1] + 2;
			if (optlen > len)
				/* ignore obscure options */
				goto done;
			break;
		}
		off += optlen;
		len -= optlen;
	}

done:
	return 1;
}

static int tcf_csum_ipv6(struct sk_buff *skb, u32 update_flags)
{
	struct ipv6hdr *ip6h;
	struct ipv6_opt_hdr *ip6xh;
	unsigned int hl, ixhl;
	unsigned int pl;
	int ntkoff;
	u8 nexthdr;

	ntkoff = skb_network_offset(skb);

	hl = sizeof(*ip6h);

	if (!pskb_may_pull(skb, hl + ntkoff))
		goto fail;

	ip6h = ipv6_hdr(skb);

	pl = ntohs(ip6h->payload_len);
	nexthdr = ip6h->nexthdr;

	do {
		switch (nexthdr) {
		case NEXTHDR_FRAGMENT:
			goto ignore_skb;
		case NEXTHDR_ROUTING:
		case NEXTHDR_HOP:
		case NEXTHDR_DEST:
			if (!pskb_may_pull(skb, hl + sizeof(*ip6xh) + ntkoff))
				goto fail;
			ip6xh = (void *)(skb_network_header(skb) + hl);
			ixhl = ipv6_optlen(ip6xh);
			if (!pskb_may_pull(skb, hl + ixhl + ntkoff))
				goto fail;
			ip6xh = (void *)(skb_network_header(skb) + hl);
			if ((nexthdr == NEXTHDR_HOP) &&
			    !(tcf_csum_ipv6_hopopts(ip6xh, ixhl, &pl)))
				goto fail;
			nexthdr = ip6xh->nexthdr;
			hl += ixhl;
			break;
		case IPPROTO_ICMPV6:
			if (update_flags & TCA_CSUM_UPDATE_FLAG_ICMP)
				if (!tcf_csum_ipv6_icmp(skb,
							hl, pl + sizeof(*ip6h)))
					goto fail;
			goto done;
		case IPPROTO_TCP:
			if (update_flags & TCA_CSUM_UPDATE_FLAG_TCP)
				if (!tcf_csum_ipv6_tcp(skb,
						       hl, pl + sizeof(*ip6h)))
					goto fail;
			goto done;
		case IPPROTO_UDP:
			if (update_flags & TCA_CSUM_UPDATE_FLAG_UDP)
				if (!tcf_csum_ipv6_udp(skb, hl,
						       pl + sizeof(*ip6h), 0))
					goto fail;
			goto done;
		case IPPROTO_UDPLITE:
			if (update_flags & TCA_CSUM_UPDATE_FLAG_UDPLITE)
				if (!tcf_csum_ipv6_udp(skb, hl,
						       pl + sizeof(*ip6h), 1))
					goto fail;
			goto done;
		default:
			goto ignore_skb;
		}
	} while (pskb_may_pull(skb, hl + 1 + ntkoff));

done:
ignore_skb:
	return 1;

fail:
	return 0;
}

static int tcf_csum(struct sk_buff *skb,
		    const struct tc_action *a, struct tcf_result *res)
{
	struct tcf_csum *p = a->priv;
	int action;
	u32 update_flags;

	spin_lock(&p->tcf_lock);
	p->tcf_tm.lastuse = jiffies;
	bstats_update(&p->tcf_bstats, skb);
	action = p->tcf_action;
	update_flags = p->update_flags;
	spin_unlock(&p->tcf_lock);

	if (unlikely(action == TC_ACT_SHOT))
		goto drop;

	switch (tc_skb_protocol(skb)) {
	case cpu_to_be16(ETH_P_IP):
		if (!tcf_csum_ipv4(skb, update_flags))
			goto drop;
		break;
	case cpu_to_be16(ETH_P_IPV6):
		if (!tcf_csum_ipv6(skb, update_flags))
			goto drop;
		break;
	}

	return action;

drop:
	spin_lock(&p->tcf_lock);
	p->tcf_qstats.drops++;
	spin_unlock(&p->tcf_lock);
	return TC_ACT_SHOT;
}

static int tcf_csum_dump(struct sk_buff *skb,
			 struct tc_action *a, int bind, int ref)
{
	unsigned char *b = skb_tail_pointer(skb);
	struct tcf_csum *p = a->priv;
	struct tc_csum opt = {
		.update_flags = p->update_flags,
		.index   = p->tcf_index,
		.action  = p->tcf_action,
		.refcnt  = p->tcf_refcnt - ref,
		.bindcnt = p->tcf_bindcnt - bind,
	};
	struct tcf_t t;

	if (nla_put(skb, TCA_CSUM_PARMS, sizeof(opt), &opt))
		goto nla_put_failure;
	t.install = jiffies_to_clock_t(jiffies - p->tcf_tm.install);
	t.lastuse = jiffies_to_clock_t(jiffies - p->tcf_tm.lastuse);
	t.expires = jiffies_to_clock_t(p->tcf_tm.expires);
	if (nla_put(skb, TCA_CSUM_TM, sizeof(t), &t))
		goto nla_put_failure;

	return skb->len;

nla_put_failure:
	nlmsg_trim(skb, b);
	return -1;
}

static int tcf_csum_walker(struct net *net, struct sk_buff *skb,
			   struct netlink_callback *cb, int type,
			   struct tc_action *a)
{
	struct tc_action_net *tn = net_generic(net, csum_net_id);

	return tcf_generic_walker(tn, skb, cb, type, a);
}

static int tcf_csum_search(struct net *net, struct tc_action *a, u32 index)
{
	struct tc_action_net *tn = net_generic(net, csum_net_id);

	return tcf_hash_search(tn, a, index);
}

static struct tc_action_ops act_csum_ops = {
	.kind		= "csum",
	.type		= TCA_ACT_CSUM,
	.owner		= THIS_MODULE,
	.act		= tcf_csum,
	.dump		= tcf_csum_dump,
	.init		= tcf_csum_init,
	.walk		= tcf_csum_walker,
	.lookup		= tcf_csum_search,
};

static __net_init int csum_init_net(struct net *net)
{
	struct tc_action_net *tn = net_generic(net, csum_net_id);

	return tc_action_net_init(tn, &act_csum_ops, CSUM_TAB_MASK);
}

static void __net_exit csum_exit_net(struct net *net)
{
	struct tc_action_net *tn = net_generic(net, csum_net_id);

	tc_action_net_exit(tn);
}

static struct pernet_operations csum_net_ops = {
	.init = csum_init_net,
	.exit = csum_exit_net,
	.id   = &csum_net_id,
	.size = sizeof(struct tc_action_net),
};

MODULE_DESCRIPTION("Checksum updating actions");
MODULE_LICENSE("GPL");

static int __init csum_init_module(void)
{
	return tcf_register_action(&act_csum_ops, &csum_net_ops);
}

static void __exit csum_cleanup_module(void)
{
	tcf_unregister_action(&act_csum_ops, &csum_net_ops);
}

module_init(csum_init_module);
module_exit(csum_cleanup_module);
Commit	Line	Data
eb4d4065 GB	1	/*
	2	* Checksum updating actions
	3	*
	4	* Copyright (c) 2010 Gregoire Baron <baronchon@n7mm.org>
	5	*
	6	* This program is free software; you can redistribute it and/or modify it
	7	* under the terms of the GNU General Public License as published by the Free
	8	* Software Foundation; either version 2 of the License, or (at your option)
	9	* any later version.
	10	*
	11	*/
	12
	13	#include <linux/types.h>
	14	#include <linux/init.h>
	15	#include <linux/kernel.h>
	16	#include <linux/module.h>
	17	#include <linux/spinlock.h>
	18
	19	#include <linux/netlink.h>
	20	#include <net/netlink.h>
	21	#include <linux/rtnetlink.h>
	22
	23	#include <linux/skbuff.h>
	24
	25	#include <net/ip.h>
	26	#include <net/ipv6.h>
	27	#include <net/icmp.h>
	28	#include <linux/icmpv6.h>
	29	#include <linux/igmp.h>
	30	#include <net/tcp.h>
	31	#include <net/udp.h>
2436243a	32	#include <net/ip6_checksum.h>
eb4d4065 GB	33
	34	#include <net/act_api.h>
	35
	36	#include <linux/tc_act/tc_csum.h>
	37	#include <net/tc_act/tc_csum.h>
	38
	39	#define CSUM_TAB_MASK 15
eb4d4065 GB	40
	41	static const struct nla_policy csum_policy[TCA_CSUM_MAX + 1] = {
	42	[TCA_CSUM_PARMS] = { .len = sizeof(struct tc_csum), },
	43	};
	44
ddf97ccd WC	45	static int csum_net_id;
	46
	47	static int tcf_csum_init(struct net net, struct nlattr nla,
	48	struct nlattr est, struct tc_action a, int ovr,
	49	int bind)
eb4d4065	50	{
ddf97ccd	51	struct tc_action_net *tn = net_generic(net, csum_net_id);
eb4d4065 GB	52	struct nlattr *tb[TCA_CSUM_MAX + 1];
eb4d4065 GB	53	struct tc_csum *parm;
eb4d4065 GB	54	struct tcf_csum *p;
	55	int ret = 0, err;
	56
	57	if (nla == NULL)
	58	return -EINVAL;
	59
cc7ec456	60	err = nla_parse_nested(tb, TCA_CSUM_MAX, nla, csum_policy);
eb4d4065 GB	61	if (err < 0)
	62	return err;
	63
	64	if (tb[TCA_CSUM_PARMS] == NULL)
	65	return -EINVAL;
	66	parm = nla_data(tb[TCA_CSUM_PARMS]);
	67
ddf97ccd WC	68	if (!tcf_hash_check(tn, parm->index, a, bind)) {
	69	ret = tcf_hash_create(tn, parm->index, est, a,
	70	sizeof(*p), bind, false);
86062033 WC	71	if (ret)
86062033 WC	72	return ret;
eb4d4065 GB	73	ret = ACT_P_CREATED;
eb4d4065 GB	74	} else {
1a29321e JHS	75	if (bind)/* dont override defaults */
1a29321e JHS	76	return 0;
86062033	77	tcf_hash_release(a, bind);
1a29321e	78	if (!ovr)
eb4d4065	79	return -EEXIST;
eb4d4065 GB	80	}
eb4d4065 GB	81
86062033	82	p = to_tcf_csum(a);
eb4d4065 GB	83	spin_lock_bh(&p->tcf_lock);
	84	p->tcf_action = parm->action;
	85	p->update_flags = parm->update_flags;
	86	spin_unlock_bh(&p->tcf_lock);
	87
	88	if (ret == ACT_P_CREATED)
ddf97ccd	89	tcf_hash_insert(tn, a);
eb4d4065 GB	90
	91	return ret;
	92	}
	93
eb4d4065 GB	94	/**
	95	* tcf_csum_skb_nextlayer - Get next layer pointer
	96	* @skb: sk_buff to use
	97	* @ihl: previous summed headers length
	98	* @ipl: complete packet length
	99	* @jhl: next header length
	100	*
	101	* Check the expected next layer availability in the specified sk_buff.
	102	* Return the next layer pointer if pass, NULL otherwise.
	103	*/
	104	static void tcf_csum_skb_nextlayer(struct sk_buff skb,
	105	unsigned int ihl, unsigned int ipl,
	106	unsigned int jhl)
	107	{
	108	int ntkoff = skb_network_offset(skb);
	109	int hl = ihl + jhl;
	110
	111	if (!pskb_may_pull(skb, ipl + ntkoff) \|\| (ipl < hl) \|\|
3697649f	112	skb_try_make_writable(skb, hl + ntkoff))
eb4d4065 GB	113	return NULL;
	114	else
	115	return (void *)(skb_network_header(skb) + ihl);
	116	}
	117
	118	static int tcf_csum_ipv4_icmp(struct sk_buff *skb,
	119	unsigned int ihl, unsigned int ipl)
	120	{
	121	struct icmphdr *icmph;
	122
	123	icmph = tcf_csum_skb_nextlayer(skb, ihl, ipl, sizeof(*icmph));
	124	if (icmph == NULL)
	125	return 0;
	126
	127	icmph->checksum = 0;
	128	skb->csum = csum_partial(icmph, ipl - ihl, 0);
	129	icmph->checksum = csum_fold(skb->csum);
	130
	131	skb->ip_summed = CHECKSUM_NONE;
	132
	133	return 1;
	134	}
	135
	136	static int tcf_csum_ipv4_igmp(struct sk_buff *skb,
	137	unsigned int ihl, unsigned int ipl)
	138	{
	139	struct igmphdr *igmph;
	140
	141	igmph = tcf_csum_skb_nextlayer(skb, ihl, ipl, sizeof(*igmph));
	142	if (igmph == NULL)
	143	return 0;
	144
	145	igmph->csum = 0;
	146	skb->csum = csum_partial(igmph, ipl - ihl, 0);
	147	igmph->csum = csum_fold(skb->csum);
	148
	149	skb->ip_summed = CHECKSUM_NONE;
	150
	151	return 1;
	152	}
	153
d14a489a	154	static int tcf_csum_ipv6_icmp(struct sk_buff *skb,
eb4d4065 GB	155	unsigned int ihl, unsigned int ipl)
	156	{
	157	struct icmp6hdr *icmp6h;
d14a489a	158	const struct ipv6hdr *ip6h;
eb4d4065 GB	159
	160	icmp6h = tcf_csum_skb_nextlayer(skb, ihl, ipl, sizeof(*icmp6h));
	161	if (icmp6h == NULL)
	162	return 0;
	163
d14a489a	164	ip6h = ipv6_hdr(skb);
eb4d4065 GB	165	icmp6h->icmp6_cksum = 0;
	166	skb->csum = csum_partial(icmp6h, ipl - ihl, 0);
	167	icmp6h->icmp6_cksum = csum_ipv6_magic(&ip6h->saddr, &ip6h->daddr,
	168	ipl - ihl, IPPROTO_ICMPV6,
	169	skb->csum);
	170
	171	skb->ip_summed = CHECKSUM_NONE;
	172
	173	return 1;
	174	}
	175
d14a489a	176	static int tcf_csum_ipv4_tcp(struct sk_buff *skb,
eb4d4065 GB	177	unsigned int ihl, unsigned int ipl)
	178	{
	179	struct tcphdr *tcph;
d14a489a	180	const struct iphdr *iph;
eb4d4065 GB	181
	182	tcph = tcf_csum_skb_nextlayer(skb, ihl, ipl, sizeof(*tcph));
	183	if (tcph == NULL)
	184	return 0;
	185
d14a489a	186	iph = ip_hdr(skb);
eb4d4065 GB	187	tcph->check = 0;
	188	skb->csum = csum_partial(tcph, ipl - ihl, 0);
	189	tcph->check = tcp_v4_check(ipl - ihl,
	190	iph->saddr, iph->daddr, skb->csum);
	191
	192	skb->ip_summed = CHECKSUM_NONE;
	193
	194	return 1;
	195	}
	196
d14a489a	197	static int tcf_csum_ipv6_tcp(struct sk_buff *skb,
eb4d4065 GB	198	unsigned int ihl, unsigned int ipl)
	199	{
	200	struct tcphdr *tcph;
d14a489a	201	const struct ipv6hdr *ip6h;
eb4d4065 GB	202
	203	tcph = tcf_csum_skb_nextlayer(skb, ihl, ipl, sizeof(*tcph));
	204	if (tcph == NULL)
	205	return 0;
	206
d14a489a	207	ip6h = ipv6_hdr(skb);
eb4d4065 GB	208	tcph->check = 0;
	209	skb->csum = csum_partial(tcph, ipl - ihl, 0);
	210	tcph->check = csum_ipv6_magic(&ip6h->saddr, &ip6h->daddr,
	211	ipl - ihl, IPPROTO_TCP,
	212	skb->csum);
	213
	214	skb->ip_summed = CHECKSUM_NONE;
	215
	216	return 1;
	217	}
	218
d14a489a	219	static int tcf_csum_ipv4_udp(struct sk_buff *skb,
eb4d4065 GB	220	unsigned int ihl, unsigned int ipl, int udplite)
	221	{
	222	struct udphdr *udph;
d14a489a	223	const struct iphdr *iph;
eb4d4065 GB	224	u16 ul;
eb4d4065 GB	225
0eec32ff CG	226	/*
	227	* Support both UDP and UDPLITE checksum algorithms, Don't use
	228	* udph->len to get the real length without any protocol check,
eb4d4065 GB	229	* UDPLITE uses udph->len for another thing,
	230	* Use iph->tot_len, or just ipl.
	231	*/
	232
	233	udph = tcf_csum_skb_nextlayer(skb, ihl, ipl, sizeof(*udph));
	234	if (udph == NULL)
	235	return 0;
	236
d14a489a	237	iph = ip_hdr(skb);
eb4d4065 GB	238	ul = ntohs(udph->len);
	239
	240	if (udplite \|\| udph->check) {
	241
	242	udph->check = 0;
	243
	244	if (udplite) {
	245	if (ul == 0)
	246	skb->csum = csum_partial(udph, ipl - ihl, 0);
eb4d4065 GB	247	else if ((ul >= sizeof(*udph)) && (ul <= ipl - ihl))
eb4d4065 GB	248	skb->csum = csum_partial(udph, ul, 0);
eb4d4065 GB	249	else
	250	goto ignore_obscure_skb;
	251	} else {
	252	if (ul != ipl - ihl)
	253	goto ignore_obscure_skb;
	254
	255	skb->csum = csum_partial(udph, ul, 0);
	256	}
	257
	258	udph->check = csum_tcpudp_magic(iph->saddr, iph->daddr,
	259	ul, iph->protocol,
	260	skb->csum);
	261
	262	if (!udph->check)
	263	udph->check = CSUM_MANGLED_0;
	264	}
	265
	266	skb->ip_summed = CHECKSUM_NONE;
	267
	268	ignore_obscure_skb:
	269	return 1;
	270	}
	271
d14a489a	272	static int tcf_csum_ipv6_udp(struct sk_buff *skb,
eb4d4065 GB	273	unsigned int ihl, unsigned int ipl, int udplite)
	274	{
	275	struct udphdr *udph;
d14a489a	276	const struct ipv6hdr *ip6h;
eb4d4065 GB	277	u16 ul;
eb4d4065 GB	278
0eec32ff CG	279	/*
	280	* Support both UDP and UDPLITE checksum algorithms, Don't use
	281	* udph->len to get the real length without any protocol check,
eb4d4065 GB	282	* UDPLITE uses udph->len for another thing,
	283	* Use ip6h->payload_len + sizeof(*ip6h) ... , or just ipl.
	284	*/
	285
	286	udph = tcf_csum_skb_nextlayer(skb, ihl, ipl, sizeof(*udph));
	287	if (udph == NULL)
	288	return 0;
	289
d14a489a	290	ip6h = ipv6_hdr(skb);
eb4d4065 GB	291	ul = ntohs(udph->len);
	292
	293	udph->check = 0;
	294
	295	if (udplite) {
	296	if (ul == 0)
	297	skb->csum = csum_partial(udph, ipl - ihl, 0);
	298
	299	else if ((ul >= sizeof(*udph)) && (ul <= ipl - ihl))
	300	skb->csum = csum_partial(udph, ul, 0);
	301
	302	else
	303	goto ignore_obscure_skb;
	304	} else {
	305	if (ul != ipl - ihl)
	306	goto ignore_obscure_skb;
	307
	308	skb->csum = csum_partial(udph, ul, 0);
	309	}
	310
	311	udph->check = csum_ipv6_magic(&ip6h->saddr, &ip6h->daddr, ul,
	312	udplite ? IPPROTO_UDPLITE : IPPROTO_UDP,
	313	skb->csum);
	314
	315	if (!udph->check)
	316	udph->check = CSUM_MANGLED_0;
	317
	318	skb->ip_summed = CHECKSUM_NONE;
	319
	320	ignore_obscure_skb:
	321	return 1;
	322	}
	323
	324	static int tcf_csum_ipv4(struct sk_buff *skb, u32 update_flags)
	325	{
d14a489a	326	const struct iphdr *iph;
eb4d4065 GB	327	int ntkoff;
	328
	329	ntkoff = skb_network_offset(skb);
	330
	331	if (!pskb_may_pull(skb, sizeof(*iph) + ntkoff))
	332	goto fail;
	333
	334	iph = ip_hdr(skb);
	335
	336	switch (iph->frag_off & htons(IP_OFFSET) ? 0 : iph->protocol) {
	337	case IPPROTO_ICMP:
	338	if (update_flags & TCA_CSUM_UPDATE_FLAG_ICMP)
0eec32ff CG	339	if (!tcf_csum_ipv4_icmp(skb, iph->ihl * 4,
0eec32ff CG	340	ntohs(iph->tot_len)))
eb4d4065 GB	341	goto fail;
	342	break;
	343	case IPPROTO_IGMP:
	344	if (update_flags & TCA_CSUM_UPDATE_FLAG_IGMP)
0eec32ff CG	345	if (!tcf_csum_ipv4_igmp(skb, iph->ihl * 4,
0eec32ff CG	346	ntohs(iph->tot_len)))
eb4d4065 GB	347	goto fail;
	348	break;
	349	case IPPROTO_TCP:
	350	if (update_flags & TCA_CSUM_UPDATE_FLAG_TCP)
d14a489a	351	if (!tcf_csum_ipv4_tcp(skb, iph->ihl * 4,
0eec32ff	352	ntohs(iph->tot_len)))
eb4d4065 GB	353	goto fail;
	354	break;
	355	case IPPROTO_UDP:
	356	if (update_flags & TCA_CSUM_UPDATE_FLAG_UDP)
d14a489a	357	if (!tcf_csum_ipv4_udp(skb, iph->ihl * 4,
0eec32ff	358	ntohs(iph->tot_len), 0))
eb4d4065 GB	359	goto fail;
	360	break;
	361	case IPPROTO_UDPLITE:
	362	if (update_flags & TCA_CSUM_UPDATE_FLAG_UDPLITE)
d14a489a	363	if (!tcf_csum_ipv4_udp(skb, iph->ihl * 4,
0eec32ff	364	ntohs(iph->tot_len), 1))
eb4d4065 GB	365	goto fail;
	366	break;
	367	}
	368
	369	if (update_flags & TCA_CSUM_UPDATE_FLAG_IPV4HDR) {
3697649f	370	if (skb_try_make_writable(skb, sizeof(*iph) + ntkoff))
eb4d4065 GB	371	goto fail;
eb4d4065 GB	372
d14a489a	373	ip_send_check(ip_hdr(skb));
eb4d4065 GB	374	}
	375
	376	return 1;
	377
	378	fail:
	379	return 0;
	380	}
	381
	382	static int tcf_csum_ipv6_hopopts(struct ipv6_opt_hdr *ip6xh,
0eec32ff	383	unsigned int ixhl, unsigned int *pl)
eb4d4065 GB	384	{
	385	int off, len, optlen;
	386	unsigned char xh = (void )ip6xh;
	387
	388	off = sizeof(*ip6xh);
	389	len = ixhl - off;
	390
	391	while (len > 1) {
0eec32ff	392	switch (xh[off]) {
1de5a71c	393	case IPV6_TLV_PAD1:
eb4d4065 GB	394	optlen = 1;
	395	break;
	396	case IPV6_TLV_JUMBO:
	397	optlen = xh[off + 1] + 2;
	398	if (optlen != 6 \|\| len < 6 \|\| (off & 3) != 2)
	399	/* wrong jumbo option length/alignment */
	400	return 0;
	401	pl = ntohl((__be32 *)(xh + off + 2));
	402	goto done;
	403	default:
	404	optlen = xh[off + 1] + 2;
	405	if (optlen > len)
	406	/* ignore obscure options */
	407	goto done;
	408	break;
	409	}
	410	off += optlen;
	411	len -= optlen;
	412	}
	413
	414	done:
	415	return 1;
	416	}
	417
	418	static int tcf_csum_ipv6(struct sk_buff *skb, u32 update_flags)
	419	{
	420	struct ipv6hdr *ip6h;
	421	struct ipv6_opt_hdr *ip6xh;
	422	unsigned int hl, ixhl;
	423	unsigned int pl;
	424	int ntkoff;
	425	u8 nexthdr;
	426
	427	ntkoff = skb_network_offset(skb);
	428
	429	hl = sizeof(*ip6h);
	430
	431	if (!pskb_may_pull(skb, hl + ntkoff))
	432	goto fail;
	433
	434	ip6h = ipv6_hdr(skb);
	435
	436	pl = ntohs(ip6h->payload_len);
	437	nexthdr = ip6h->nexthdr;
	438
	439	do {
	440	switch (nexthdr) {
	441	case NEXTHDR_FRAGMENT:
	442	goto ignore_skb;
	443	case NEXTHDR_ROUTING:
	444	case NEXTHDR_HOP:
	445	case NEXTHDR_DEST:
	446	if (!pskb_may_pull(skb, hl + sizeof(*ip6xh) + ntkoff))
	447	goto fail;
	448	ip6xh = (void *)(skb_network_header(skb) + hl);
	449	ixhl = ipv6_optlen(ip6xh);
	450	if (!pskb_may_pull(skb, hl + ixhl + ntkoff))
	451	goto fail;
d14a489a	452	ip6xh = (void *)(skb_network_header(skb) + hl);
eb4d4065 GB	453	if ((nexthdr == NEXTHDR_HOP) &&
	454	!(tcf_csum_ipv6_hopopts(ip6xh, ixhl, &pl)))
	455	goto fail;
	456	nexthdr = ip6xh->nexthdr;
	457	hl += ixhl;
	458	break;
	459	case IPPROTO_ICMPV6:
	460	if (update_flags & TCA_CSUM_UPDATE_FLAG_ICMP)
d14a489a	461	if (!tcf_csum_ipv6_icmp(skb,
eb4d4065 GB	462	hl, pl + sizeof(*ip6h)))
	463	goto fail;
	464	goto done;
	465	case IPPROTO_TCP:
	466	if (update_flags & TCA_CSUM_UPDATE_FLAG_TCP)
d14a489a	467	if (!tcf_csum_ipv6_tcp(skb,
eb4d4065 GB	468	hl, pl + sizeof(*ip6h)))
	469	goto fail;
	470	goto done;
	471	case IPPROTO_UDP:
	472	if (update_flags & TCA_CSUM_UPDATE_FLAG_UDP)
d14a489a	473	if (!tcf_csum_ipv6_udp(skb, hl,
0eec32ff	474	pl + sizeof(*ip6h), 0))
eb4d4065 GB	475	goto fail;
	476	goto done;
	477	case IPPROTO_UDPLITE:
	478	if (update_flags & TCA_CSUM_UPDATE_FLAG_UDPLITE)
d14a489a	479	if (!tcf_csum_ipv6_udp(skb, hl,
0eec32ff	480	pl + sizeof(*ip6h), 1))
eb4d4065 GB	481	goto fail;
	482	goto done;
	483	default:
	484	goto ignore_skb;
	485	}
	486	} while (pskb_may_pull(skb, hl + 1 + ntkoff));
	487
	488	done:
	489	ignore_skb:
	490	return 1;
	491
	492	fail:
	493	return 0;
	494	}
	495
	496	static int tcf_csum(struct sk_buff *skb,
dc7f9f6e	497	const struct tc_action a, struct tcf_result res)
eb4d4065 GB	498	{
	499	struct tcf_csum *p = a->priv;
	500	int action;
	501	u32 update_flags;
	502
	503	spin_lock(&p->tcf_lock);
	504	p->tcf_tm.lastuse = jiffies;
bfe0d029	505	bstats_update(&p->tcf_bstats, skb);
eb4d4065 GB	506	action = p->tcf_action;
	507	update_flags = p->update_flags;
	508	spin_unlock(&p->tcf_lock);
	509
	510	if (unlikely(action == TC_ACT_SHOT))
	511	goto drop;
	512
d8b9605d	513	switch (tc_skb_protocol(skb)) {
eb4d4065 GB	514	case cpu_to_be16(ETH_P_IP):
	515	if (!tcf_csum_ipv4(skb, update_flags))
	516	goto drop;
	517	break;
	518	case cpu_to_be16(ETH_P_IPV6):
	519	if (!tcf_csum_ipv6(skb, update_flags))
	520	goto drop;
	521	break;
	522	}
	523
	524	return action;
	525
	526	drop:
	527	spin_lock(&p->tcf_lock);
	528	p->tcf_qstats.drops++;
	529	spin_unlock(&p->tcf_lock);
	530	return TC_ACT_SHOT;
	531	}
	532
	533	static int tcf_csum_dump(struct sk_buff *skb,
	534	struct tc_action *a, int bind, int ref)
	535	{
	536	unsigned char *b = skb_tail_pointer(skb);
	537	struct tcf_csum *p = a->priv;
	538	struct tc_csum opt = {
	539	.update_flags = p->update_flags,
eb4d4065 GB	540	.index = p->tcf_index,
	541	.action = p->tcf_action,
	542	.refcnt = p->tcf_refcnt - ref,
	543	.bindcnt = p->tcf_bindcnt - bind,
	544	};
	545	struct tcf_t t;
	546
1b34ec43 DM	547	if (nla_put(skb, TCA_CSUM_PARMS, sizeof(opt), &opt))
1b34ec43 DM	548	goto nla_put_failure;
eb4d4065 GB	549	t.install = jiffies_to_clock_t(jiffies - p->tcf_tm.install);
	550	t.lastuse = jiffies_to_clock_t(jiffies - p->tcf_tm.lastuse);
	551	t.expires = jiffies_to_clock_t(p->tcf_tm.expires);
1b34ec43 DM	552	if (nla_put(skb, TCA_CSUM_TM, sizeof(t), &t))
1b34ec43 DM	553	goto nla_put_failure;
eb4d4065 GB	554
	555	return skb->len;
	556
	557	nla_put_failure:
	558	nlmsg_trim(skb, b);
	559	return -1;
	560	}
	561
ddf97ccd WC	562	static int tcf_csum_walker(struct net net, struct sk_buff skb,
	563	struct netlink_callback *cb, int type,
	564	struct tc_action *a)
	565	{
	566	struct tc_action_net *tn = net_generic(net, csum_net_id);
	567
	568	return tcf_generic_walker(tn, skb, cb, type, a);
	569	}
	570
	571	static int tcf_csum_search(struct net net, struct tc_action a, u32 index)
	572	{
	573	struct tc_action_net *tn = net_generic(net, csum_net_id);
	574
	575	return tcf_hash_search(tn, a, index);
	576	}
	577
eb4d4065	578	static struct tc_action_ops act_csum_ops = {
0eec32ff	579	.kind = "csum",
0eec32ff	580	.type = TCA_ACT_CSUM,
0eec32ff CG	581	.owner = THIS_MODULE,
	582	.act = tcf_csum,
	583	.dump = tcf_csum_dump,
0eec32ff	584	.init = tcf_csum_init,
ddf97ccd WC	585	.walk = tcf_csum_walker,
	586	.lookup = tcf_csum_search,
	587	};
	588
	589	static __net_init int csum_init_net(struct net *net)
	590	{
	591	struct tc_action_net *tn = net_generic(net, csum_net_id);
	592
	593	return tc_action_net_init(tn, &act_csum_ops, CSUM_TAB_MASK);
	594	}
	595
	596	static void __net_exit csum_exit_net(struct net *net)
	597	{
	598	struct tc_action_net *tn = net_generic(net, csum_net_id);
	599
	600	tc_action_net_exit(tn);
	601	}
	602
	603	static struct pernet_operations csum_net_ops = {
	604	.init = csum_init_net,
	605	.exit = csum_exit_net,
	606	.id = &csum_net_id,
	607	.size = sizeof(struct tc_action_net),
eb4d4065 GB	608	};
	609
	610	MODULE_DESCRIPTION("Checksum updating actions");
	611	MODULE_LICENSE("GPL");
	612
	613	static int __init csum_init_module(void)
	614	{
ddf97ccd	615	return tcf_register_action(&act_csum_ops, &csum_net_ops);
eb4d4065 GB	616	}
	617
	618	static void __exit csum_cleanup_module(void)
	619	{
ddf97ccd	620	tcf_unregister_action(&act_csum_ops, &csum_net_ops);
eb4d4065 GB	621	}
	622
	623	module_init(csum_init_module);
	624	module_exit(csum_cleanup_module);