[deliverable/linux.git] / net / ipv6 / ip6_offload.c

/*
 *	IPV6 GSO/GRO offload support
 *	Linux INET6 implementation
 *
 *	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/kernel.h>
#include <linux/socket.h>
#include <linux/netdevice.h>
#include <linux/skbuff.h>
#include <linux/printk.h>

#include <net/protocol.h>
#include <net/ipv6.h>

#include "ip6_offload.h"

static int ipv6_gso_pull_exthdrs(struct sk_buff *skb, int proto)
{
	const struct net_offload *ops = NULL;

	for (;;) {
		struct ipv6_opt_hdr *opth;
		int len;

		if (proto != NEXTHDR_HOP) {
			ops = rcu_dereference(inet6_offloads[proto]);

			if (unlikely(!ops))
				break;

			if (!(ops->flags & INET6_PROTO_GSO_EXTHDR))
				break;
		}

		if (unlikely(!pskb_may_pull(skb, 8)))
			break;

		opth = (void *)skb->data;
		len = ipv6_optlen(opth);

		if (unlikely(!pskb_may_pull(skb, len)))
			break;

		opth = (void *)skb->data;
		proto = opth->nexthdr;
		__skb_pull(skb, len);
	}

	return proto;
}

static struct sk_buff *ipv6_gso_segment(struct sk_buff *skb,
	netdev_features_t features)
{
	struct sk_buff *segs = ERR_PTR(-EINVAL);
	struct ipv6hdr *ipv6h;
	const struct net_offload *ops;
	int proto;
	struct frag_hdr *fptr;
	unsigned int unfrag_ip6hlen;
	u8 *prevhdr;
	int offset = 0;
	bool encap, udpfrag;
	int nhoff;

	if (unlikely(skb_shinfo(skb)->gso_type &
		     ~(SKB_GSO_UDP |
		       SKB_GSO_DODGY |
		       SKB_GSO_TCP_ECN |
		       SKB_GSO_GRE |
		       SKB_GSO_GRE_CSUM |
		       SKB_GSO_IPIP |
		       SKB_GSO_SIT |
		       SKB_GSO_UDP_TUNNEL |
		       SKB_GSO_UDP_TUNNEL_CSUM |
		       SKB_GSO_MPLS |
		       SKB_GSO_TCPV6 |
		       0)))
		goto out;

	skb_reset_network_header(skb);
	nhoff = skb_network_header(skb) - skb_mac_header(skb);
	if (unlikely(!pskb_may_pull(skb, sizeof(*ipv6h))))
		goto out;

	encap = SKB_GSO_CB(skb)->encap_level > 0;
	if (encap)
		features &= skb->dev->hw_enc_features;
	SKB_GSO_CB(skb)->encap_level += sizeof(*ipv6h);

	ipv6h = ipv6_hdr(skb);
	__skb_pull(skb, sizeof(*ipv6h));
	segs = ERR_PTR(-EPROTONOSUPPORT);

	proto = ipv6_gso_pull_exthdrs(skb, ipv6h->nexthdr);

	if (skb->encapsulation &&
	    skb_shinfo(skb)->gso_type & (SKB_GSO_SIT|SKB_GSO_IPIP))
		udpfrag = proto == IPPROTO_UDP && encap;
	else
		udpfrag = proto == IPPROTO_UDP && !skb->encapsulation;

	ops = rcu_dereference(inet6_offloads[proto]);
	if (likely(ops && ops->callbacks.gso_segment)) {
		skb_reset_transport_header(skb);
		segs = ops->callbacks.gso_segment(skb, features);
	}

	if (IS_ERR(segs))
		goto out;

	for (skb = segs; skb; skb = skb->next) {
		ipv6h = (struct ipv6hdr *)(skb_mac_header(skb) + nhoff);
		ipv6h->payload_len = htons(skb->len - nhoff - sizeof(*ipv6h));
		skb->network_header = (u8 *)ipv6h - skb->head;

		if (udpfrag) {
			unfrag_ip6hlen = ip6_find_1stfragopt(skb, &prevhdr);
			fptr = (struct frag_hdr *)((u8 *)ipv6h + unfrag_ip6hlen);
			fptr->frag_off = htons(offset);
			if (skb->next != NULL)
				fptr->frag_off |= htons(IP6_MF);
			offset += (ntohs(ipv6h->payload_len) -
				   sizeof(struct frag_hdr));
		}
		if (encap)
			skb_reset_inner_headers(skb);
	}

out:
	return segs;
}

/* Return the total length of all the extension hdrs, following the same
 * logic in ipv6_gso_pull_exthdrs() when parsing ext-hdrs.
 */
static int ipv6_exthdrs_len(struct ipv6hdr *iph,
			    const struct net_offload **opps)
{
	struct ipv6_opt_hdr *opth = (void *)iph;
	int len = 0, proto, optlen = sizeof(*iph);

	proto = iph->nexthdr;
	for (;;) {
		if (proto != NEXTHDR_HOP) {
			*opps = rcu_dereference(inet6_offloads[proto]);
			if (unlikely(!(*opps)))
				break;
			if (!((*opps)->flags & INET6_PROTO_GSO_EXTHDR))
				break;
		}
		opth = (void *)opth + optlen;
		optlen = ipv6_optlen(opth);
		len += optlen;
		proto = opth->nexthdr;
	}
	return len;
}

static struct sk_buff **ipv6_gro_receive(struct sk_buff **head,
					 struct sk_buff *skb)
{
	const struct net_offload *ops;
	struct sk_buff **pp = NULL;
	struct sk_buff *p;
	struct ipv6hdr *iph;
	unsigned int nlen;
	unsigned int hlen;
	unsigned int off;
	u16 flush = 1;
	int proto;

	off = skb_gro_offset(skb);
	hlen = off + sizeof(*iph);
	iph = skb_gro_header_fast(skb, off);
	if (skb_gro_header_hard(skb, hlen)) {
		iph = skb_gro_header_slow(skb, hlen, off);
		if (unlikely(!iph))
			goto out;
	}

	skb_set_network_header(skb, off);
	skb_gro_pull(skb, sizeof(*iph));
	skb_set_transport_header(skb, skb_gro_offset(skb));

	flush += ntohs(iph->payload_len) != skb_gro_len(skb);

	rcu_read_lock();
	proto = iph->nexthdr;
	ops = rcu_dereference(inet6_offloads[proto]);
	if (!ops || !ops->callbacks.gro_receive) {
		__pskb_pull(skb, skb_gro_offset(skb));
		proto = ipv6_gso_pull_exthdrs(skb, proto);
		skb_gro_pull(skb, -skb_transport_offset(skb));
		skb_reset_transport_header(skb);
		__skb_push(skb, skb_gro_offset(skb));

		ops = rcu_dereference(inet6_offloads[proto]);
		if (!ops || !ops->callbacks.gro_receive)
			goto out_unlock;

		iph = ipv6_hdr(skb);
	}

	NAPI_GRO_CB(skb)->proto = proto;

	flush--;
	nlen = skb_network_header_len(skb);

	for (p = *head; p; p = p->next) {
		const struct ipv6hdr *iph2;
		__be32 first_word; /* <Version:4><Traffic_Class:8><Flow_Label:20> */

		if (!NAPI_GRO_CB(p)->same_flow)
			continue;

		iph2 = (struct ipv6hdr *)(p->data + off);
		first_word = *(__be32 *)iph ^ *(__be32 *)iph2;

		/* All fields must match except length and Traffic Class.
		 * XXX skbs on the gro_list have all been parsed and pulled
		 * already so we don't need to compare nlen
		 * (nlen != (sizeof(*iph2) + ipv6_exthdrs_len(iph2, &ops)))
		 * memcmp() alone below is suffcient, right?
		 */
		 if ((first_word & htonl(0xF00FFFFF)) ||
		    memcmp(&iph->nexthdr, &iph2->nexthdr,
			   nlen - offsetof(struct ipv6hdr, nexthdr))) {
			NAPI_GRO_CB(p)->same_flow = 0;
			continue;
		}
		/* flush if Traffic Class fields are different */
		NAPI_GRO_CB(p)->flush |= !!(first_word & htonl(0x0FF00000));
		NAPI_GRO_CB(p)->flush |= flush;

		/* Clear flush_id, there's really no concept of ID in IPv6. */
		NAPI_GRO_CB(p)->flush_id = 0;
	}

	NAPI_GRO_CB(skb)->flush |= flush;

	skb_gro_postpull_rcsum(skb, iph, nlen);

	pp = ops->callbacks.gro_receive(head, skb);

out_unlock:
	rcu_read_unlock();

out:
	NAPI_GRO_CB(skb)->flush |= flush;

	return pp;
}

static int ipv6_gro_complete(struct sk_buff *skb, int nhoff)
{
	const struct net_offload *ops;
	struct ipv6hdr *iph = (struct ipv6hdr *)(skb->data + nhoff);
	int err = -ENOSYS;

	iph->payload_len = htons(skb->len - nhoff - sizeof(*iph));

	rcu_read_lock();

	nhoff += sizeof(*iph) + ipv6_exthdrs_len(iph, &ops);
	if (WARN_ON(!ops || !ops->callbacks.gro_complete))
		goto out_unlock;

	err = ops->callbacks.gro_complete(skb, nhoff);

out_unlock:
	rcu_read_unlock();

	return err;
}

static struct packet_offload ipv6_packet_offload __read_mostly = {
	.type = cpu_to_be16(ETH_P_IPV6),
	.callbacks = {
		.gso_segment = ipv6_gso_segment,
		.gro_receive = ipv6_gro_receive,
		.gro_complete = ipv6_gro_complete,
	},
};

static const struct net_offload sit_offload = {
	.callbacks = {
		.gso_segment	= ipv6_gso_segment,
		.gro_receive	= ipv6_gro_receive,
		.gro_complete	= ipv6_gro_complete,
	},
};

static int __init ipv6_offload_init(void)
{

	if (tcpv6_offload_init() < 0)
		pr_crit("%s: Cannot add TCP protocol offload\n", __func__);
	if (udp_offload_init() < 0)
		pr_crit("%s: Cannot add UDP protocol offload\n", __func__);
	if (ipv6_exthdrs_offload_init() < 0)
		pr_crit("%s: Cannot add EXTHDRS protocol offload\n", __func__);

	dev_add_offload(&ipv6_packet_offload);

	inet_add_offload(&sit_offload, IPPROTO_IPV6);

	return 0;
}

fs_initcall(ipv6_offload_init);
Commit	Line	Data
d1da932e VY	1	/*
	2	* IPV6 GSO/GRO offload support
	3	* Linux INET6 implementation
	4	*
	5	* This program is free software; you can redistribute it and/or
	6	* modify it under the terms of the GNU General Public License
	7	* as published by the Free Software Foundation; either version
	8	* 2 of the License, or (at your option) any later version.
	9	*/
	10
	11	#include <linux/kernel.h>
	12	#include <linux/socket.h>
	13	#include <linux/netdevice.h>
	14	#include <linux/skbuff.h>
c6b641a4	15	#include <linux/printk.h>
d1da932e VY	16
	17	#include <net/protocol.h>
	18	#include <net/ipv6.h>
	19
	20	#include "ip6_offload.h"
	21
	22	static int ipv6_gso_pull_exthdrs(struct sk_buff *skb, int proto)
	23	{
	24	const struct net_offload *ops = NULL;
	25
	26	for (;;) {
	27	struct ipv6_opt_hdr *opth;
	28	int len;
	29
	30	if (proto != NEXTHDR_HOP) {
	31	ops = rcu_dereference(inet6_offloads[proto]);
	32
	33	if (unlikely(!ops))
	34	break;
	35
	36	if (!(ops->flags & INET6_PROTO_GSO_EXTHDR))
	37	break;
	38	}
	39
	40	if (unlikely(!pskb_may_pull(skb, 8)))
	41	break;
	42
	43	opth = (void *)skb->data;
	44	len = ipv6_optlen(opth);
	45
	46	if (unlikely(!pskb_may_pull(skb, len)))
	47	break;
	48
fc6fb41c	49	opth = (void *)skb->data;
d1da932e VY	50	proto = opth->nexthdr;
	51	__skb_pull(skb, len);
	52	}
	53
	54	return proto;
	55	}
	56
d1da932e VY	57	static struct sk_buff ipv6_gso_segment(struct sk_buff skb,
	58	netdev_features_t features)
	59	{
	60	struct sk_buff *segs = ERR_PTR(-EINVAL);
	61	struct ipv6hdr *ipv6h;
	62	const struct net_offload *ops;
	63	int proto;
	64	struct frag_hdr *fptr;
	65	unsigned int unfrag_ip6hlen;
	66	u8 *prevhdr;
	67	int offset = 0;
91a48a2e	68	bool encap, udpfrag;
d3e5e006	69	int nhoff;
d1da932e	70
d1da932e VY	71	if (unlikely(skb_shinfo(skb)->gso_type &
	72	~(SKB_GSO_UDP \|
	73	SKB_GSO_DODGY \|
	74	SKB_GSO_TCP_ECN \|
68c33163	75	SKB_GSO_GRE \|
4749c09c	76	SKB_GSO_GRE_CSUM \|
cb32f511	77	SKB_GSO_IPIP \|
61c1db7f	78	SKB_GSO_SIT \|
73136267	79	SKB_GSO_UDP_TUNNEL \|
0f4f4ffa	80	SKB_GSO_UDP_TUNNEL_CSUM \|
0d89d203	81	SKB_GSO_MPLS \|
d1da932e VY	82	SKB_GSO_TCPV6 \|
	83	0)))
	84	goto out;
	85
d3e5e006 ED	86	skb_reset_network_header(skb);
d3e5e006 ED	87	nhoff = skb_network_header(skb) - skb_mac_header(skb);
d1da932e VY	88	if (unlikely(!pskb_may_pull(skb, sizeof(*ipv6h))))
	89	goto out;
	90
91a48a2e HFS	91	encap = SKB_GSO_CB(skb)->encap_level > 0;
91a48a2e HFS	92	if (encap)
1e16aa3d	93	features &= skb->dev->hw_enc_features;
d3e5e006 ED	94	SKB_GSO_CB(skb)->encap_level += sizeof(*ipv6h);
d3e5e006 ED	95
d1da932e VY	96	ipv6h = ipv6_hdr(skb);
	97	__skb_pull(skb, sizeof(*ipv6h));
	98	segs = ERR_PTR(-EPROTONOSUPPORT);
	99
	100	proto = ipv6_gso_pull_exthdrs(skb, ipv6h->nexthdr);
b917eb15	101
91a48a2e HFS	102	if (skb->encapsulation &&
	103	skb_shinfo(skb)->gso_type & (SKB_GSO_SIT\|SKB_GSO_IPIP))
	104	udpfrag = proto == IPPROTO_UDP && encap;
	105	else
	106	udpfrag = proto == IPPROTO_UDP && !skb->encapsulation;
	107
d1da932e	108	ops = rcu_dereference(inet6_offloads[proto]);
f191a1d1	109	if (likely(ops && ops->callbacks.gso_segment)) {
d1da932e	110	skb_reset_transport_header(skb);
f191a1d1	111	segs = ops->callbacks.gso_segment(skb, features);
d1da932e	112	}
d1da932e VY	113
	114	if (IS_ERR(segs))
	115	goto out;
	116
	117	for (skb = segs; skb; skb = skb->next) {
d3e5e006 ED	118	ipv6h = (struct ipv6hdr *)(skb_mac_header(skb) + nhoff);
d3e5e006 ED	119	ipv6h->payload_len = htons(skb->len - nhoff - sizeof(*ipv6h));
d3e5e006 ED	120	skb->network_header = (u8 *)ipv6h - skb->head;
d3e5e006 ED	121
91a48a2e	122	if (udpfrag) {
d1da932e	123	unfrag_ip6hlen = ip6_find_1stfragopt(skb, &prevhdr);
d3e5e006	124	fptr = (struct frag_hdr )((u8 )ipv6h + unfrag_ip6hlen);
d1da932e VY	125	fptr->frag_off = htons(offset);
	126	if (skb->next != NULL)
	127	fptr->frag_off \|= htons(IP6_MF);
	128	offset += (ntohs(ipv6h->payload_len) -
	129	sizeof(struct frag_hdr));
	130	}
91a48a2e HFS	131	if (encap)
91a48a2e HFS	132	skb_reset_inner_headers(skb);
d1da932e VY	133	}
	134
	135	out:
	136	return segs;
	137	}
	138
299603e8 JC	139	/* Return the total length of all the extension hdrs, following the same
	140	* logic in ipv6_gso_pull_exthdrs() when parsing ext-hdrs.
	141	*/
	142	static int ipv6_exthdrs_len(struct ipv6hdr *iph,
	143	const struct net_offload **opps)
	144	{
810c23a3 JC	145	struct ipv6_opt_hdr opth = (void )iph;
810c23a3 JC	146	int len = 0, proto, optlen = sizeof(*iph);
299603e8 JC	147
	148	proto = iph->nexthdr;
	149	for (;;) {
	150	if (proto != NEXTHDR_HOP) {
	151	*opps = rcu_dereference(inet6_offloads[proto]);
	152	if (unlikely(!(*opps)))
	153	break;
	154	if (!((*opps)->flags & INET6_PROTO_GSO_EXTHDR))
	155	break;
	156	}
810c23a3 JC	157	opth = (void *)opth + optlen;
810c23a3 JC	158	optlen = ipv6_optlen(opth);
299603e8 JC	159	len += optlen;
	160	proto = opth->nexthdr;
	161	}
	162	return len;
	163	}
	164
d1da932e VY	165	static struct sk_buff ipv6_gro_receive(struct sk_buff head,
	166	struct sk_buff *skb)
	167	{
	168	const struct net_offload *ops;
	169	struct sk_buff **pp = NULL;
	170	struct sk_buff *p;
	171	struct ipv6hdr *iph;
	172	unsigned int nlen;
	173	unsigned int hlen;
	174	unsigned int off;
bf5a755f	175	u16 flush = 1;
d1da932e	176	int proto;
d1da932e VY	177
	178	off = skb_gro_offset(skb);
	179	hlen = off + sizeof(*iph);
	180	iph = skb_gro_header_fast(skb, off);
	181	if (skb_gro_header_hard(skb, hlen)) {
	182	iph = skb_gro_header_slow(skb, hlen, off);
	183	if (unlikely(!iph))
	184	goto out;
	185	}
	186
299603e8	187	skb_set_network_header(skb, off);
d1da932e VY	188	skb_gro_pull(skb, sizeof(*iph));
	189	skb_set_transport_header(skb, skb_gro_offset(skb));
	190
	191	flush += ntohs(iph->payload_len) != skb_gro_len(skb);
	192
	193	rcu_read_lock();
	194	proto = iph->nexthdr;
	195	ops = rcu_dereference(inet6_offloads[proto]);
f191a1d1	196	if (!ops \|\| !ops->callbacks.gro_receive) {
d1da932e VY	197	__pskb_pull(skb, skb_gro_offset(skb));
	198	proto = ipv6_gso_pull_exthdrs(skb, proto);
	199	skb_gro_pull(skb, -skb_transport_offset(skb));
	200	skb_reset_transport_header(skb);
	201	__skb_push(skb, skb_gro_offset(skb));
	202
	203	ops = rcu_dereference(inet6_offloads[proto]);
f191a1d1	204	if (!ops \|\| !ops->callbacks.gro_receive)
d1da932e VY	205	goto out_unlock;
	206
	207	iph = ipv6_hdr(skb);
	208	}
	209
	210	NAPI_GRO_CB(skb)->proto = proto;
	211
	212	flush--;
	213	nlen = skb_network_header_len(skb);
	214
	215	for (p = *head; p; p = p->next) {
	216	const struct ipv6hdr *iph2;
	217	__be32 first_word; /* <Version:4><Traffic_Class:8><Flow_Label:20> */
	218
	219	if (!NAPI_GRO_CB(p)->same_flow)
	220	continue;
	221
299603e8	222	iph2 = (struct ipv6hdr *)(p->data + off);
67ba4152	223	first_word = (__be32 )iph ^ (__be32 )iph2;
d1da932e	224
299603e8 JC	225	/* All fields must match except length and Traffic Class.
	226	* XXX skbs on the gro_list have all been parsed and pulled
	227	* already so we don't need to compare nlen
	228	* (nlen != (sizeof(*iph2) + ipv6_exthdrs_len(iph2, &ops)))
	229	* memcmp() alone below is suffcient, right?
	230	*/
	231	if ((first_word & htonl(0xF00FFFFF)) \|\|
d1da932e VY	232	memcmp(&iph->nexthdr, &iph2->nexthdr,
	233	nlen - offsetof(struct ipv6hdr, nexthdr))) {
	234	NAPI_GRO_CB(p)->same_flow = 0;
	235	continue;
	236	}
	237	/* flush if Traffic Class fields are different */
	238	NAPI_GRO_CB(p)->flush \|= !!(first_word & htonl(0x0FF00000));
	239	NAPI_GRO_CB(p)->flush \|= flush;
03d56daa TH	240
	241	/* Clear flush_id, there's really no concept of ID in IPv6. */
	242	NAPI_GRO_CB(p)->flush_id = 0;
d1da932e VY	243	}
	244
	245	NAPI_GRO_CB(skb)->flush \|= flush;
	246
4de462ab	247	skb_gro_postpull_rcsum(skb, iph, nlen);
d1da932e	248
f191a1d1	249	pp = ops->callbacks.gro_receive(head, skb);
d1da932e	250
d1da932e VY	251	out_unlock:
	252	rcu_read_unlock();
	253
	254	out:
	255	NAPI_GRO_CB(skb)->flush \|= flush;
	256
	257	return pp;
	258	}
	259
299603e8	260	static int ipv6_gro_complete(struct sk_buff *skb, int nhoff)
d1da932e VY	261	{
d1da932e VY	262	const struct net_offload *ops;
299603e8	263	struct ipv6hdr iph = (struct ipv6hdr )(skb->data + nhoff);
d1da932e VY	264	int err = -ENOSYS;
d1da932e VY	265
299603e8	266	iph->payload_len = htons(skb->len - nhoff - sizeof(*iph));
d1da932e VY	267
d1da932e VY	268	rcu_read_lock();
299603e8 JC	269
299603e8 JC	270	nhoff += sizeof(*iph) + ipv6_exthdrs_len(iph, &ops);
f191a1d1	271	if (WARN_ON(!ops \|\| !ops->callbacks.gro_complete))
d1da932e VY	272	goto out_unlock;
d1da932e VY	273
299603e8	274	err = ops->callbacks.gro_complete(skb, nhoff);
d1da932e VY	275
	276	out_unlock:
	277	rcu_read_unlock();
	278
	279	return err;
	280	}
	281
	282	static struct packet_offload ipv6_packet_offload __read_mostly = {
	283	.type = cpu_to_be16(ETH_P_IPV6),
f191a1d1	284	.callbacks = {
f191a1d1 VY	285	.gso_segment = ipv6_gso_segment,
	286	.gro_receive = ipv6_gro_receive,
	287	.gro_complete = ipv6_gro_complete,
	288	},
d1da932e VY	289	};
d1da932e VY	290
61c1db7f ED	291	static const struct net_offload sit_offload = {
61c1db7f ED	292	.callbacks = {
61c1db7f	293	.gso_segment = ipv6_gso_segment,
19424e05 TH	294	.gro_receive = ipv6_gro_receive,
19424e05 TH	295	.gro_complete = ipv6_gro_complete,
61c1db7f ED	296	},
	297	};
	298
c6b641a4	299	static int __init ipv6_offload_init(void)
d1da932e	300	{
c6b641a4 VY	301
	302	if (tcpv6_offload_init() < 0)
	303	pr_crit("%s: Cannot add TCP protocol offload\n", __func__);
	304	if (udp_offload_init() < 0)
	305	pr_crit("%s: Cannot add UDP protocol offload\n", __func__);
	306	if (ipv6_exthdrs_offload_init() < 0)
	307	pr_crit("%s: Cannot add EXTHDRS protocol offload\n", __func__);
	308
d1da932e	309	dev_add_offload(&ipv6_packet_offload);
61c1db7f ED	310
	311	inet_add_offload(&sit_offload, IPPROTO_IPV6);
	312
c6b641a4	313	return 0;
d1da932e VY	314	}
d1da932e VY	315
c6b641a4	316	fs_initcall(ipv6_offload_init);