summaryrefslogtreecommitdiffstats
path: root/net/ipv4/inet_lro.c
blob: f17ea49b28fb48100968a92e9cade456527e9cb4 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
/*
 *  linux/net/ipv4/inet_lro.c
 *
 *  Large Receive Offload (ipv4 / tcp)
 *
 *  (C) Copyright IBM Corp. 2007
 *
 *  Authors:
 *       Jan-Bernd Themann <themann@de.ibm.com>
 *       Christoph Raisch <raisch@de.ibm.com>
 *
 *
 * 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, or (at your option)
 * any later version.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, write to the Free Software
 * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
 */


#include <linux/module.h>
#include <linux/if_vlan.h>
#include <linux/inet_lro.h>
#include <net/checksum.h>

MODULE_LICENSE("GPL");
MODULE_AUTHOR("Jan-Bernd Themann <themann@de.ibm.com>");
MODULE_DESCRIPTION("Large Receive Offload (ipv4 / tcp)");

#define TCP_HDR_LEN(tcph) (tcph->doff << 2)
#define IP_HDR_LEN(iph) (iph->ihl << 2)
#define TCP_PAYLOAD_LENGTH(iph, tcph) \
	(ntohs(iph->tot_len) - IP_HDR_LEN(iph) - TCP_HDR_LEN(tcph))

#define IPH_LEN_WO_OPTIONS 5
#define TCPH_LEN_WO_OPTIONS 5
#define TCPH_LEN_W_TIMESTAMP 8

#define LRO_MAX_PG_HLEN 64

#define LRO_INC_STATS(lro_mgr, attr) { lro_mgr->stats.attr++; }

/*
 * Basic tcp checks whether packet is suitable for LRO
 */

static int lro_tcp_ip_check(const struct iphdr *iph, const struct tcphdr *tcph,
			    int len, const struct net_lro_desc *lro_desc)
{
        /* check ip header: don't aggregate padded frames */
	if (ntohs(iph->tot_len) != len)
		return -1;

	if (TCP_PAYLOAD_LENGTH(iph, tcph) == 0)
		return -1;

	if (iph->ihl != IPH_LEN_WO_OPTIONS)
		return -1;

	if (tcph->cwr || tcph->ece || tcph->urg || !tcph->ack ||
	    tcph->rst || tcph->syn || tcph->fin)
		return -1;

	if (INET_ECN_is_ce(ipv4_get_dsfield(iph)))
		return -1;

	if (tcph->doff != TCPH_LEN_WO_OPTIONS &&
	    tcph->doff != TCPH_LEN_W_TIMESTAMP)
		return -1;

	/* check tcp options (only timestamp allowed) */
	if (tcph->doff == TCPH_LEN_W_TIMESTAMP) {
		__be32 *topt = (__be32 *)(tcph + 1);

		if (*topt != htonl((TCPOPT_NOP << 24) | (TCPOPT_NOP << 16)
				   | (TCPOPT_TIMESTAMP << 8)
				   | TCPOLEN_TIMESTAMP))
			return -1;

		/* timestamp should be in right order */
		topt++;
		if (lro_desc && after(ntohl(lro_desc->tcp_rcv_tsval),
				      ntohl(*topt)))
			return -1;

		/* timestamp reply should not be zero */
		topt++;
		if (*topt == 0)
			return -1;
	}

	return 0;
}

static void lro_update_tcp_ip_header(struct net_lro_desc *lro_desc)
{
	struct iphdr *iph = lro_desc->iph;
	struct tcphdr *tcph = lro_desc->tcph;
	__be32 *p;
	__wsum tcp_hdr_csum;

	tcph->ack_seq = lro_desc->tcp_ack;
	tcph->window = lro_desc->tcp_window;

	if (lro_desc->tcp_saw_tstamp) {
		p = (__be32 *)(tcph + 1);
		*(p+2) = lro_desc->tcp_rcv_tsecr;
	}

	csum_replace2(&iph->check, iph->tot_len, htons(lro_desc->ip_tot_len));
	iph->tot_len = htons(lro_desc->ip_tot_len);

	tcph->check = 0;
	tcp_hdr_csum = csum_partial(tcph, TCP_HDR_LEN(tcph), 0);
	lro_desc->data_csum = csum_add(lro_desc->data_csum, tcp_hdr_csum);
	tcph->check = csum_tcpudp_magic(iph->saddr, iph->daddr,
					lro_desc->ip_tot_len -
					IP_HDR_LEN(iph), IPPROTO_TCP,
					lro_desc->data_csum);
}

static __wsum lro_tcp_data_csum(struct iphdr *iph, struct tcphdr *tcph, int len)
{
	__wsum tcp_csum;
	__wsum tcp_hdr_csum;
	__wsum tcp_ps_hdr_csum;

	tcp_csum = ~csum_unfold(tcph->check);
	tcp_hdr_csum = csum_partial(tcph, TCP_HDR_LEN(tcph), tcp_csum);

	tcp_ps_hdr_csum = csum_tcpudp_nofold(iph->saddr, iph->daddr,
					     len + TCP_HDR_LEN(tcph),
					     IPPROTO_TCP, 0);

	return csum_sub(csum_sub(tcp_csum, tcp_hdr_csum),
			tcp_ps_hdr_csum);
}

static void lro_init_desc(struct net_lro_desc *lro_desc, struct sk_buff *skb,
			  struct iphdr *iph, struct tcphdr *tcph)
{
	int nr_frags;
	__be32 *ptr;
	u32 tcp_data_len = TCP_PAYLOAD_LENGTH(iph, tcph);

	nr_frags = skb_shinfo(skb)->nr_frags;
	lro_desc->parent = skb;
	lro_desc->next_frag = &(skb_shinfo(skb)->frags[nr_frags]);
	lro_desc->iph = iph;
	lro_desc->tcph = tcph;
	lro_desc->tcp_next_seq = ntohl(tcph->seq) + tcp_data_len;
	lro_desc->tcp_ack = tcph->ack_seq;
	lro_desc->tcp_window = tcph->window;

	lro_desc->pkt_aggr_cnt = 1;
	lro_desc->ip_tot_len = ntohs(iph->tot_len);

	if (tcph->doff == 8) {
		ptr = (__be32 *)(tcph+1);
		lro_desc->tcp_saw_tstamp = 1;
		lro_desc->tcp_rcv_tsval = *(ptr+1);
		lro_desc->tcp_rcv_tsecr = *(ptr+2);
	}

	lro_desc->mss = tcp_data_len;
	lro_desc->active = 1;

	lro_desc->data_csum = lro_tcp_data_csum(iph, tcph,
						tcp_data_len);
}

static inline void lro_clear_desc(struct net_lro_desc *lro_desc)
{
	memset(lro_desc, 0, sizeof(struct net_lro_desc));
}

static void lro_add_common(struct net_lro_desc *lro_desc, struct iphdr *iph,
			   struct tcphdr *tcph, int tcp_data_len)
{
	struct sk_buff *parent = lro_desc->parent;
	__be32 *topt;

	lro_desc->pkt_aggr_cnt++;
	lro_desc->ip_tot_len += tcp_data_len;
	lro_desc->tcp_next_seq += tcp_data_len;
	lro_desc->tcp_window = tcph->window;
	lro_desc->tcp_ack = tcph->ack_seq;

	/* don't update tcp_rcv_tsval, would not work with PAWS */
	if (lro_desc->tcp_saw_tstamp) {
		topt = (__be32 *) (tcph + 1);
		lro_desc->tcp_rcv_tsecr = *(topt + 2);
	}

	lro_desc->data_csum = csum_block_add(lro_desc->data_csum,
					     lro_tcp_data_csum(iph, tcph,
							       tcp_data_len),
					     parent->len);

	parent->len += tcp_data_len;
	parent->data_len += tcp_data_len;
	if (tcp_data_len > lro_desc->mss)
		lro_desc->mss = tcp_data_len;
}

static void lro_add_packet(struct net_lro_desc *lro_desc, struct sk_buff *skb,
			   struct iphdr *iph, struct tcphdr *tcph)
{
	struct sk_buff *parent = lro_desc->parent;
	int tcp_data_len = TCP_PAYLOAD_LENGTH(iph, tcph);

	lro_add_common(lro_desc, iph, tcph, tcp_data_len);

	skb_pull(skb, (skb->len - tcp_data_len));
	parent->truesize += skb->truesize;

	if (lro_desc->last_skb)
		lro_desc->last_skb->next = skb;
	else
		skb_shinfo(parent)->frag_list = skb;

	lro_desc->last_skb = skb;
}


static int lro_check_tcp_conn(struct net_lro_desc *lro_desc,
			      struct iphdr *iph,
			      struct tcphdr *tcph)
{
	if ((lro_desc->iph->saddr != iph->saddr) ||
	    (lro_desc->iph->daddr != iph->daddr) ||
	    (lro_desc->tcph->source != tcph->source) ||
	    (lro_desc->tcph->dest != tcph->dest))
		return -1;
	return 0;
}

static struct net_lro_desc *lro_get_desc(struct net_lro_mgr *lro_mgr,
					 struct net_lro_desc *lro_arr,
					 struct iphdr *iph,
					 struct tcphdr *tcph)
{
	struct net_lro_desc *lro_desc = NULL;
	struct net_lro_desc *tmp;
	int max_desc = lro_mgr->max_desc;
	int i;

	for (i = 0; i < max_desc; i++) {
		tmp = &lro_arr[i];
		if (tmp->active)
			if (!lro_check_tcp_conn(tmp, iph, tcph)) {
				lro_desc = tmp;
				goto out;
			}
	}

	for (i = 0; i < max_desc; i++) {
		if (!lro_arr[i].active) {
			lro_desc = &lro_arr[i];
			goto out;
		}
	}

	LRO_INC_STATS(lro_mgr, no_desc);
out:
	return lro_desc;
}

static void lro_flush(struct net_lro_mgr *lro_mgr,
		      struct net_lro_desc *lro_desc)
{
	if (lro_desc->pkt_aggr_cnt > 1)
		lro_update_tcp_ip_header(lro_desc);

	skb_shinfo(lro_desc->parent)->gso_size = lro_desc->mss;

	if (lro_mgr->features & LRO_F_NAPI)
		netif_receive_skb(lro_desc->parent);
	else
		netif_rx(lro_desc->parent);

	LRO_INC_STATS(lro_mgr, flushed);
	lro_clear_desc(lro_desc);
}

static int __lro_proc_skb(struct net_lro_mgr *lro_mgr, struct sk_buff *skb,
			  void *priv)
{
	struct net_lro_desc *lro_desc;
	struct iphdr *iph;
	struct tcphdr *tcph;
	u64 flags;
	int vlan_hdr_len = 0;

	if (!lro_mgr->get_skb_header ||
	    lro_mgr->get_skb_header(skb, (void *)&iph, (void *)&tcph,
				    &flags, priv))
		goto out;

	if (!(flags & LRO_IPV4) || !(flags & LRO_TCP))
		goto out;

	lro_desc = lro_get_desc(lro_mgr, lro_mgr->lro_arr, iph, tcph);
	if (!lro_desc)
		goto out;

	if ((skb->protocol == htons(ETH_P_8021Q)) &&
	    !(lro_mgr->features & LRO_F_EXTRACT_VLAN_ID))
		vlan_hdr_len = VLAN_HLEN;

	if (!lro_desc->active) { /* start new lro session */
		if (lro_tcp_ip_check(iph, tcph, skb->len - vlan_hdr_len, NULL))
			goto out;

		skb->ip_summed = lro_mgr->ip_summed_aggr;
		lro_init_desc(lro_desc, skb, iph, tcph);
		LRO_INC_STATS(lro_mgr, aggregated);
		return 0;
	}

	if (lro_desc->tcp_next_seq != ntohl(tcph->seq))
		goto out2;

	if (lro_tcp_ip_check(iph, tcph, skb->len, lro_desc))
		goto out2;

	lro_add_packet(lro_desc, skb, iph, tcph);
	LRO_INC_STATS(lro_mgr, aggregated);

	if ((lro_desc->pkt_aggr_cnt >= lro_mgr->max_aggr) ||
	    lro_desc->parent->len > (0xFFFF - lro_mgr->dev->mtu))
		lro_flush(lro_mgr, lro_desc);

	return 0;

out2: /* send aggregated SKBs to stack */
	lro_flush(lro_mgr, lro_desc);

out:
	return 1;
}

void lro_receive_skb(struct net_lro_mgr *lro_mgr,
		     struct sk_buff *skb,
		     void *priv)
{
	if (__lro_proc_skb(lro_mgr, skb, priv)) {
		if (lro_mgr->features & LRO_F_NAPI)
			netif_receive_skb(skb);
		else
			netif_rx(skb);
	}
}
EXPORT_SYMBOL(lro_receive_skb);

void lro_flush_all(struct net_lro_mgr *lro_mgr)
{
	int i;
	struct net_lro_desc *lro_desc = lro_mgr->lro_arr;

	for (i = 0; i < lro_mgr->max_desc; i++) {
		if (lro_desc[i].active)
			lro_flush(lro_mgr, &lro_desc[i]);
	}
}
EXPORT_SYMBOL(lro_flush_all);
OpenPOWER on IntegriCloud