[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 struct tcf_hashinfo csum_hash_info;

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

static int tcf_csum_init(struct net *n, struct nlattr *nla, struct nlattr *est,
			 struct tc_action *a, int ovr, int bind)
{
	struct nlattr *tb[TCA_CSUM_MAX + 1];
	struct tc_csum *parm;
	struct tcf_common *pc;
	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]);

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

	p = to_tcf_csum(pc);
	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(pc, a->ops->hinfo);

	return ret;
}

static int tcf_csum_cleanup(struct tc_action *a, int bind)
{
	struct tcf_csum *p = a->priv;
	return tcf_hash_release(&p->common, bind, &csum_hash_info);
}

/**
 * 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_cloned(skb) &&
	     !skb_clone_writable(skb, hl + ntkoff) &&
	     pskb_expand_head(skb, 0, 0, GFP_ATOMIC)))
		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_cloned(skb) &&
		    !skb_clone_writable(skb, sizeof(*iph) + ntkoff) &&
		    pskb_expand_head(skb, 0, 0, GFP_ATOMIC))
			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 (skb->protocol) {
	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 struct tc_action_ops act_csum_ops = {
	.kind		= "csum",
	.hinfo		= &csum_hash_info,
	.type		= TCA_ACT_CSUM,
	.owner		= THIS_MODULE,
	.act		= tcf_csum,
	.dump		= tcf_csum_dump,
	.cleanup	= tcf_csum_cleanup,
	.init		= tcf_csum_init,
};

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

static int __init csum_init_module(void)
{
	int err = tcf_hashinfo_init(&csum_hash_info, CSUM_TAB_MASK);
	if (err)
		return err;

	return tcf_register_action(&act_csum_ops);
}

static void __exit csum_cleanup_module(void)
{
	tcf_unregister_action(&act_csum_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
369ba567	40	static struct tcf_hashinfo csum_hash_info;
eb4d4065 GB	41
	42	static const struct nla_policy csum_policy[TCA_CSUM_MAX + 1] = {
	43	[TCA_CSUM_PARMS] = { .len = sizeof(struct tc_csum), },
	44	};
	45
c1b52739	46	static int tcf_csum_init(struct net n, struct nlattr nla, struct nlattr *est,
eb4d4065 GB	47	struct tc_action *a, int ovr, int bind)
	48	{
	49	struct nlattr *tb[TCA_CSUM_MAX + 1];
	50	struct tc_csum *parm;
	51	struct tcf_common *pc;
	52	struct tcf_csum *p;
	53	int ret = 0, err;
	54
	55	if (nla == NULL)
	56	return -EINVAL;
	57
cc7ec456	58	err = nla_parse_nested(tb, TCA_CSUM_MAX, nla, csum_policy);
eb4d4065 GB	59	if (err < 0)
	60	return err;
	61
	62	if (tb[TCA_CSUM_PARMS] == NULL)
	63	return -EINVAL;
	64	parm = nla_data(tb[TCA_CSUM_PARMS]);
	65
c779f7af	66	pc = tcf_hash_check(parm->index, a, bind);
eb4d4065	67	if (!pc) {
c779f7af	68	pc = tcf_hash_create(parm->index, est, a, sizeof(*p), bind);
eb4d4065 GB	69	if (IS_ERR(pc))
eb4d4065 GB	70	return PTR_ERR(pc);
eb4d4065 GB	71	ret = ACT_P_CREATED;
eb4d4065 GB	72	} else {
1a29321e JHS	73	if (bind)/* dont override defaults */
1a29321e JHS	74	return 0;
c779f7af	75	tcf_hash_release(pc, bind, a->ops->hinfo);
1a29321e	76	if (!ovr)
eb4d4065	77	return -EEXIST;
eb4d4065 GB	78	}
eb4d4065 GB	79
1a29321e	80	p = to_tcf_csum(pc);
eb4d4065 GB	81	spin_lock_bh(&p->tcf_lock);
	82	p->tcf_action = parm->action;
	83	p->update_flags = parm->update_flags;
	84	spin_unlock_bh(&p->tcf_lock);
	85
	86	if (ret == ACT_P_CREATED)
c779f7af	87	tcf_hash_insert(pc, a->ops->hinfo);
eb4d4065 GB	88
	89	return ret;
	90	}
	91
	92	static int tcf_csum_cleanup(struct tc_action *a, int bind)
	93	{
	94	struct tcf_csum *p = a->priv;
	95	return tcf_hash_release(&p->common, bind, &csum_hash_info);
	96	}
	97
	98	/**
	99	* tcf_csum_skb_nextlayer - Get next layer pointer
	100	* @skb: sk_buff to use
	101	* @ihl: previous summed headers length
	102	* @ipl: complete packet length
	103	* @jhl: next header length
	104	*
	105	* Check the expected next layer availability in the specified sk_buff.
	106	* Return the next layer pointer if pass, NULL otherwise.
	107	*/
	108	static void tcf_csum_skb_nextlayer(struct sk_buff skb,
	109	unsigned int ihl, unsigned int ipl,
	110	unsigned int jhl)
	111	{
	112	int ntkoff = skb_network_offset(skb);
	113	int hl = ihl + jhl;
	114
	115	if (!pskb_may_pull(skb, ipl + ntkoff) \|\| (ipl < hl) \|\|
	116	(skb_cloned(skb) &&
	117	!skb_clone_writable(skb, hl + ntkoff) &&
	118	pskb_expand_head(skb, 0, 0, GFP_ATOMIC)))
	119	return NULL;
	120	else
	121	return (void *)(skb_network_header(skb) + ihl);
	122	}
	123
	124	static int tcf_csum_ipv4_icmp(struct sk_buff *skb,
	125	unsigned int ihl, unsigned int ipl)
	126	{
	127	struct icmphdr *icmph;
	128
	129	icmph = tcf_csum_skb_nextlayer(skb, ihl, ipl, sizeof(*icmph));
	130	if (icmph == NULL)
	131	return 0;
	132
	133	icmph->checksum = 0;
	134	skb->csum = csum_partial(icmph, ipl - ihl, 0);
	135	icmph->checksum = csum_fold(skb->csum);
	136
	137	skb->ip_summed = CHECKSUM_NONE;
	138
	139	return 1;
	140	}
	141
	142	static int tcf_csum_ipv4_igmp(struct sk_buff *skb,
	143	unsigned int ihl, unsigned int ipl)
	144	{
	145	struct igmphdr *igmph;
	146
	147	igmph = tcf_csum_skb_nextlayer(skb, ihl, ipl, sizeof(*igmph));
	148	if (igmph == NULL)
	149	return 0;
	150
	151	igmph->csum = 0;
152	skb->csum = csum_partial(igmph, ipl - ihl, 0);
153	igmph->csum = csum_fold(skb->csum);
154
155	skb->ip_summed = CHECKSUM_NONE;
156
157	return 1;
158	}
159
d14a489a	160	static int tcf_csum_ipv6_icmp(struct sk_buff *skb,
eb4d4065 GB	161	unsigned int ihl, unsigned int ipl)
	162	{
	163	struct icmp6hdr *icmp6h;
d14a489a	164	const struct ipv6hdr *ip6h;
eb4d4065 GB	165
	166	icmp6h = tcf_csum_skb_nextlayer(skb, ihl, ipl, sizeof(*icmp6h));
	167	if (icmp6h == NULL)
	168	return 0;
	169
d14a489a	170	ip6h = ipv6_hdr(skb);
eb4d4065 GB	171	icmp6h->icmp6_cksum = 0;
	172	skb->csum = csum_partial(icmp6h, ipl - ihl, 0);
	173	icmp6h->icmp6_cksum = csum_ipv6_magic(&ip6h->saddr, &ip6h->daddr,
	174	ipl - ihl, IPPROTO_ICMPV6,
	175	skb->csum);
	176
	177	skb->ip_summed = CHECKSUM_NONE;
	178
	179	return 1;
	180	}
	181
d14a489a	182	static int tcf_csum_ipv4_tcp(struct sk_buff *skb,
eb4d4065 GB	183	unsigned int ihl, unsigned int ipl)
	184	{
	185	struct tcphdr *tcph;
d14a489a	186	const struct iphdr *iph;
eb4d4065 GB	187
	188	tcph = tcf_csum_skb_nextlayer(skb, ihl, ipl, sizeof(*tcph));
	189	if (tcph == NULL)
	190	return 0;
	191
d14a489a	192	iph = ip_hdr(skb);
eb4d4065 GB	193	tcph->check = 0;
	194	skb->csum = csum_partial(tcph, ipl - ihl, 0);
	195	tcph->check = tcp_v4_check(ipl - ihl,
	196	iph->saddr, iph->daddr, skb->csum);
	197
	198	skb->ip_summed = CHECKSUM_NONE;
	199
	200	return 1;
	201	}
	202
d14a489a	203	static int tcf_csum_ipv6_tcp(struct sk_buff *skb,
eb4d4065 GB	204	unsigned int ihl, unsigned int ipl)
	205	{
	206	struct tcphdr *tcph;
d14a489a	207	const struct ipv6hdr *ip6h;
eb4d4065 GB	208
	209	tcph = tcf_csum_skb_nextlayer(skb, ihl, ipl, sizeof(*tcph));
	210	if (tcph == NULL)
	211	return 0;
	212
d14a489a	213	ip6h = ipv6_hdr(skb);
eb4d4065 GB	214	tcph->check = 0;
	215	skb->csum = csum_partial(tcph, ipl - ihl, 0);
	216	tcph->check = csum_ipv6_magic(&ip6h->saddr, &ip6h->daddr,
	217	ipl - ihl, IPPROTO_TCP,
	218	skb->csum);
	219
	220	skb->ip_summed = CHECKSUM_NONE;
	221
	222	return 1;
	223	}
	224
d14a489a	225	static int tcf_csum_ipv4_udp(struct sk_buff *skb,
eb4d4065 GB	226	unsigned int ihl, unsigned int ipl, int udplite)
	227	{
	228	struct udphdr *udph;
d14a489a	229	const struct iphdr *iph;
eb4d4065 GB	230	u16 ul;
eb4d4065 GB	231
0eec32ff CG	232	/*
	233	* Support both UDP and UDPLITE checksum algorithms, Don't use
	234	* udph->len to get the real length without any protocol check,
eb4d4065 GB	235	* UDPLITE uses udph->len for another thing,
	236	* Use iph->tot_len, or just ipl.
	237	*/
	238
	239	udph = tcf_csum_skb_nextlayer(skb, ihl, ipl, sizeof(*udph));
	240	if (udph == NULL)
	241	return 0;
	242
d14a489a	243	iph = ip_hdr(skb);
eb4d4065 GB	244	ul = ntohs(udph->len);
	245
	246	if (udplite \|\| udph->check) {
	247
	248	udph->check = 0;
	249
	250	if (udplite) {
	251	if (ul == 0)
	252	skb->csum = csum_partial(udph, ipl - ihl, 0);
eb4d4065 GB	253	else if ((ul >= sizeof(*udph)) && (ul <= ipl - ihl))
eb4d4065 GB	254	skb->csum = csum_partial(udph, ul, 0);
eb4d4065 GB	255	else
	256	goto ignore_obscure_skb;
	257	} else {
	258	if (ul != ipl - ihl)
	259	goto ignore_obscure_skb;
	260
	261	skb->csum = csum_partial(udph, ul, 0);
	262	}
	263
	264	udph->check = csum_tcpudp_magic(iph->saddr, iph->daddr,
	265	ul, iph->protocol,
	266	skb->csum);
	267
	268	if (!udph->check)
	269	udph->check = CSUM_MANGLED_0;
	270	}
	271
	272	skb->ip_summed = CHECKSUM_NONE;
	273
	274	ignore_obscure_skb:
	275	return 1;
	276	}
	277
d14a489a	278	static int tcf_csum_ipv6_udp(struct sk_buff *skb,
eb4d4065 GB	279	unsigned int ihl, unsigned int ipl, int udplite)
	280	{
	281	struct udphdr *udph;
d14a489a	282	const struct ipv6hdr *ip6h;
eb4d4065 GB	283	u16 ul;
eb4d4065 GB	284
0eec32ff CG	285	/*
	286	* Support both UDP and UDPLITE checksum algorithms, Don't use
	287	* udph->len to get the real length without any protocol check,
eb4d4065 GB	288	* UDPLITE uses udph->len for another thing,
	289	* Use ip6h->payload_len + sizeof(*ip6h) ... , or just ipl.
	290	*/
	291
	292	udph = tcf_csum_skb_nextlayer(skb, ihl, ipl, sizeof(*udph));
	293	if (udph == NULL)
	294	return 0;
	295
d14a489a	296	ip6h = ipv6_hdr(skb);
eb4d4065 GB	297	ul = ntohs(udph->len);
	298
	299	udph->check = 0;
	300
	301	if (udplite) {
	302	if (ul == 0)
	303	skb->csum = csum_partial(udph, ipl - ihl, 0);
	304
	305	else if ((ul >= sizeof(*udph)) && (ul <= ipl - ihl))
	306	skb->csum = csum_partial(udph, ul, 0);
	307
	308	else
	309	goto ignore_obscure_skb;
	310	} else {
	311	if (ul != ipl - ihl)
	312	goto ignore_obscure_skb;
	313
	314	skb->csum = csum_partial(udph, ul, 0);
	315	}
	316
	317	udph->check = csum_ipv6_magic(&ip6h->saddr, &ip6h->daddr, ul,
	318	udplite ? IPPROTO_UDPLITE : IPPROTO_UDP,
	319	skb->csum);
	320
	321	if (!udph->check)
	322	udph->check = CSUM_MANGLED_0;
	323
	324	skb->ip_summed = CHECKSUM_NONE;
	325
	326	ignore_obscure_skb:
	327	return 1;
	328	}
	329
	330	static int tcf_csum_ipv4(struct sk_buff *skb, u32 update_flags)
	331	{
d14a489a	332	const struct iphdr *iph;
eb4d4065 GB	333	int ntkoff;
	334
	335	ntkoff = skb_network_offset(skb);
	336
	337	if (!pskb_may_pull(skb, sizeof(*iph) + ntkoff))
	338	goto fail;
	339
	340	iph = ip_hdr(skb);
	341
	342	switch (iph->frag_off & htons(IP_OFFSET) ? 0 : iph->protocol) {
	343	case IPPROTO_ICMP:
	344	if (update_flags & TCA_CSUM_UPDATE_FLAG_ICMP)
0eec32ff CG	345	if (!tcf_csum_ipv4_icmp(skb, iph->ihl * 4,
0eec32ff CG	346	ntohs(iph->tot_len)))
eb4d4065 GB	347	goto fail;
	348	break;
	349	case IPPROTO_IGMP:
	350	if (update_flags & TCA_CSUM_UPDATE_FLAG_IGMP)
0eec32ff CG	351	if (!tcf_csum_ipv4_igmp(skb, iph->ihl * 4,
0eec32ff CG	352	ntohs(iph->tot_len)))
eb4d4065 GB	353	goto fail;
	354	break;
	355	case IPPROTO_TCP:
	356	if (update_flags & TCA_CSUM_UPDATE_FLAG_TCP)
d14a489a	357	if (!tcf_csum_ipv4_tcp(skb, iph->ihl * 4,
0eec32ff	358	ntohs(iph->tot_len)))
eb4d4065 GB	359	goto fail;
	360	break;
	361	case IPPROTO_UDP:
	362	if (update_flags & TCA_CSUM_UPDATE_FLAG_UDP)
d14a489a	363	if (!tcf_csum_ipv4_udp(skb, iph->ihl * 4,
0eec32ff	364	ntohs(iph->tot_len), 0))
eb4d4065 GB	365	goto fail;
	366	break;
	367	case IPPROTO_UDPLITE:
	368	if (update_flags & TCA_CSUM_UPDATE_FLAG_UDPLITE)
d14a489a	369	if (!tcf_csum_ipv4_udp(skb, iph->ihl * 4,
0eec32ff	370	ntohs(iph->tot_len), 1))
eb4d4065 GB	371	goto fail;
	372	break;
	373	}
	374
	375	if (update_flags & TCA_CSUM_UPDATE_FLAG_IPV4HDR) {
	376	if (skb_cloned(skb) &&
	377	!skb_clone_writable(skb, sizeof(*iph) + ntkoff) &&
	378	pskb_expand_head(skb, 0, 0, GFP_ATOMIC))
	379	goto fail;
	380
d14a489a	381	ip_send_check(ip_hdr(skb));
eb4d4065 GB	382	}
	383
	384	return 1;
	385
	386	fail:
	387	return 0;
	388	}
	389
	390	static int tcf_csum_ipv6_hopopts(struct ipv6_opt_hdr *ip6xh,
0eec32ff	391	unsigned int ixhl, unsigned int *pl)
eb4d4065 GB	392	{
	393	int off, len, optlen;
	394	unsigned char xh = (void )ip6xh;
	395
	396	off = sizeof(*ip6xh);
	397	len = ixhl - off;
	398
	399	while (len > 1) {
0eec32ff	400	switch (xh[off]) {
1de5a71c	401	case IPV6_TLV_PAD1:
eb4d4065 GB	402	optlen = 1;
	403	break;
	404	case IPV6_TLV_JUMBO:
	405	optlen = xh[off + 1] + 2;
	406	if (optlen != 6 \|\| len < 6 \|\| (off & 3) != 2)
	407	/* wrong jumbo option length/alignment */
	408	return 0;
	409	pl = ntohl((__be32 *)(xh + off + 2));
	410	goto done;
	411	default:
	412	optlen = xh[off + 1] + 2;
	413	if (optlen > len)
	414	/* ignore obscure options */
	415	goto done;
	416	break;
	417	}
	418	off += optlen;
	419	len -= optlen;
	420	}
	421
	422	done:
	423	return 1;
	424	}
	425
	426	static int tcf_csum_ipv6(struct sk_buff *skb, u32 update_flags)
	427	{
	428	struct ipv6hdr *ip6h;
	429	struct ipv6_opt_hdr *ip6xh;
	430	unsigned int hl, ixhl;
	431	unsigned int pl;
	432	int ntkoff;
	433	u8 nexthdr;
	434
	435	ntkoff = skb_network_offset(skb);
	436
	437	hl = sizeof(*ip6h);
	438
	439	if (!pskb_may_pull(skb, hl + ntkoff))
	440	goto fail;
	441
	442	ip6h = ipv6_hdr(skb);
	443
	444	pl = ntohs(ip6h->payload_len);
	445	nexthdr = ip6h->nexthdr;
	446
	447	do {
	448	switch (nexthdr) {
	449	case NEXTHDR_FRAGMENT:
	450	goto ignore_skb;
	451	case NEXTHDR_ROUTING:
	452	case NEXTHDR_HOP:
	453	case NEXTHDR_DEST:
	454	if (!pskb_may_pull(skb, hl + sizeof(*ip6xh) + ntkoff))
	455	goto fail;
	456	ip6xh = (void *)(skb_network_header(skb) + hl);
	457	ixhl = ipv6_optlen(ip6xh);
	458	if (!pskb_may_pull(skb, hl + ixhl + ntkoff))
	459	goto fail;
d14a489a	460	ip6xh = (void *)(skb_network_header(skb) + hl);
eb4d4065 GB	461	if ((nexthdr == NEXTHDR_HOP) &&
	462	!(tcf_csum_ipv6_hopopts(ip6xh, ixhl, &pl)))
	463	goto fail;
	464	nexthdr = ip6xh->nexthdr;
	465	hl += ixhl;
	466	break;
	467	case IPPROTO_ICMPV6:
	468	if (update_flags & TCA_CSUM_UPDATE_FLAG_ICMP)
d14a489a	469	if (!tcf_csum_ipv6_icmp(skb,
eb4d4065 GB	470	hl, pl + sizeof(*ip6h)))
	471	goto fail;
	472	goto done;
	473	case IPPROTO_TCP:
	474	if (update_flags & TCA_CSUM_UPDATE_FLAG_TCP)
d14a489a	475	if (!tcf_csum_ipv6_tcp(skb,
eb4d4065 GB	476	hl, pl + sizeof(*ip6h)))
	477	goto fail;
	478	goto done;
	479	case IPPROTO_UDP:
	480	if (update_flags & TCA_CSUM_UPDATE_FLAG_UDP)
d14a489a	481	if (!tcf_csum_ipv6_udp(skb, hl,
0eec32ff	482	pl + sizeof(*ip6h), 0))
eb4d4065 GB	483	goto fail;
	484	goto done;
	485	case IPPROTO_UDPLITE:
	486	if (update_flags & TCA_CSUM_UPDATE_FLAG_UDPLITE)
d14a489a	487	if (!tcf_csum_ipv6_udp(skb, hl,
0eec32ff	488	pl + sizeof(*ip6h), 1))
eb4d4065 GB	489	goto fail;
	490	goto done;
	491	default:
	492	goto ignore_skb;
	493	}
	494	} while (pskb_may_pull(skb, hl + 1 + ntkoff));
	495
	496	done:
	497	ignore_skb:
	498	return 1;
	499
	500	fail:
	501	return 0;
	502	}
	503
	504	static int tcf_csum(struct sk_buff *skb,
dc7f9f6e	505	const struct tc_action a, struct tcf_result res)
eb4d4065 GB	506	{
	507	struct tcf_csum *p = a->priv;
	508	int action;
	509	u32 update_flags;
	510
	511	spin_lock(&p->tcf_lock);
	512	p->tcf_tm.lastuse = jiffies;
bfe0d029	513	bstats_update(&p->tcf_bstats, skb);
eb4d4065 GB	514	action = p->tcf_action;
	515	update_flags = p->update_flags;
	516	spin_unlock(&p->tcf_lock);
	517
	518	if (unlikely(action == TC_ACT_SHOT))
	519	goto drop;
	520
	521	switch (skb->protocol) {
	522	case cpu_to_be16(ETH_P_IP):
	523	if (!tcf_csum_ipv4(skb, update_flags))
	524	goto drop;
	525	break;
	526	case cpu_to_be16(ETH_P_IPV6):
	527	if (!tcf_csum_ipv6(skb, update_flags))
	528	goto drop;
	529	break;
	530	}
	531
	532	return action;
	533
	534	drop:
	535	spin_lock(&p->tcf_lock);
	536	p->tcf_qstats.drops++;
	537	spin_unlock(&p->tcf_lock);
	538	return TC_ACT_SHOT;
	539	}
	540
	541	static int tcf_csum_dump(struct sk_buff *skb,
	542	struct tc_action *a, int bind, int ref)
	543	{
	544	unsigned char *b = skb_tail_pointer(skb);
	545	struct tcf_csum *p = a->priv;
	546	struct tc_csum opt = {
	547	.update_flags = p->update_flags,
eb4d4065 GB	548	.index = p->tcf_index,
	549	.action = p->tcf_action,
	550	.refcnt = p->tcf_refcnt - ref,
	551	.bindcnt = p->tcf_bindcnt - bind,
	552	};
	553	struct tcf_t t;
	554
1b34ec43 DM	555	if (nla_put(skb, TCA_CSUM_PARMS, sizeof(opt), &opt))
1b34ec43 DM	556	goto nla_put_failure;
eb4d4065 GB	557	t.install = jiffies_to_clock_t(jiffies - p->tcf_tm.install);
	558	t.lastuse = jiffies_to_clock_t(jiffies - p->tcf_tm.lastuse);
	559	t.expires = jiffies_to_clock_t(p->tcf_tm.expires);
1b34ec43 DM	560	if (nla_put(skb, TCA_CSUM_TM, sizeof(t), &t))
1b34ec43 DM	561	goto nla_put_failure;
eb4d4065 GB	562
	563	return skb->len;
	564
	565	nla_put_failure:
	566	nlmsg_trim(skb, b);
	567	return -1;
	568	}
	569
	570	static struct tc_action_ops act_csum_ops = {
0eec32ff CG	571	.kind = "csum",
	572	.hinfo = &csum_hash_info,
	573	.type = TCA_ACT_CSUM,
0eec32ff CG	574	.owner = THIS_MODULE,
	575	.act = tcf_csum,
	576	.dump = tcf_csum_dump,
	577	.cleanup = tcf_csum_cleanup,
0eec32ff	578	.init = tcf_csum_init,
eb4d4065 GB	579	};
	580
	581	MODULE_DESCRIPTION("Checksum updating actions");
	582	MODULE_LICENSE("GPL");
	583
	584	static int __init csum_init_module(void)
	585	{
568a153a	586	int err = tcf_hashinfo_init(&csum_hash_info, CSUM_TAB_MASK);
369ba567 WC	587	if (err)
	588	return err;
	589
eb4d4065 GB	590	return tcf_register_action(&act_csum_ops);
	591	}
	592
	593	static void __exit csum_cleanup_module(void)
	594	{
	595	tcf_unregister_action(&act_csum_ops);
	596	}
	597
	598	module_init(csum_init_module);
	599	module_exit(csum_cleanup_module);