xref: /openbmc/linux/net/sched/act_csum.c (revision 74ce1896)
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>
32 #include <net/ip6_checksum.h>
33 #include <net/sctp/checksum.h>
34 
35 #include <net/act_api.h>
36 
37 #include <linux/tc_act/tc_csum.h>
38 #include <net/tc_act/tc_csum.h>
39 
40 static const struct nla_policy csum_policy[TCA_CSUM_MAX + 1] = {
41 	[TCA_CSUM_PARMS] = { .len = sizeof(struct tc_csum), },
42 };
43 
44 static unsigned int csum_net_id;
45 static struct tc_action_ops act_csum_ops;
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)
50 {
51 	struct tc_action_net *tn = net_generic(net, csum_net_id);
52 	struct nlattr *tb[TCA_CSUM_MAX + 1];
53 	struct tc_csum *parm;
54 	struct tcf_csum *p;
55 	int ret = 0, err;
56 
57 	if (nla == NULL)
58 		return -EINVAL;
59 
60 	err = nla_parse_nested(tb, TCA_CSUM_MAX, nla, csum_policy, NULL);
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 
68 	if (!tcf_idr_check(tn, parm->index, a, bind)) {
69 		ret = tcf_idr_create(tn, parm->index, est, a,
70 				     &act_csum_ops, bind, false);
71 		if (ret)
72 			return ret;
73 		ret = ACT_P_CREATED;
74 	} else {
75 		if (bind)/* dont override defaults */
76 			return 0;
77 		tcf_idr_release(*a, bind);
78 		if (!ovr)
79 			return -EEXIST;
80 	}
81 
82 	p = to_tcf_csum(*a);
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)
89 		tcf_idr_insert(tn, *a);
90 
91 	return ret;
92 }
93 
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) ||
112 	    skb_try_make_writable(skb, hl + ntkoff))
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, unsigned int ihl,
119 			      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 
154 static int tcf_csum_ipv6_icmp(struct sk_buff *skb, unsigned int ihl,
155 			      unsigned int ipl)
156 {
157 	struct icmp6hdr *icmp6h;
158 	const struct ipv6hdr *ip6h;
159 
160 	icmp6h = tcf_csum_skb_nextlayer(skb, ihl, ipl, sizeof(*icmp6h));
161 	if (icmp6h == NULL)
162 		return 0;
163 
164 	ip6h = ipv6_hdr(skb);
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 
176 static int tcf_csum_ipv4_tcp(struct sk_buff *skb, unsigned int ihl,
177 			     unsigned int ipl)
178 {
179 	struct tcphdr *tcph;
180 	const struct iphdr *iph;
181 
182 	if (skb_is_gso(skb) && skb_shinfo(skb)->gso_type & SKB_GSO_TCPV4)
183 		return 1;
184 
185 	tcph = tcf_csum_skb_nextlayer(skb, ihl, ipl, sizeof(*tcph));
186 	if (tcph == NULL)
187 		return 0;
188 
189 	iph = ip_hdr(skb);
190 	tcph->check = 0;
191 	skb->csum = csum_partial(tcph, ipl - ihl, 0);
192 	tcph->check = tcp_v4_check(ipl - ihl,
193 				   iph->saddr, iph->daddr, skb->csum);
194 
195 	skb->ip_summed = CHECKSUM_NONE;
196 
197 	return 1;
198 }
199 
200 static int tcf_csum_ipv6_tcp(struct sk_buff *skb, unsigned int ihl,
201 			     unsigned int ipl)
202 {
203 	struct tcphdr *tcph;
204 	const struct ipv6hdr *ip6h;
205 
206 	if (skb_is_gso(skb) && skb_shinfo(skb)->gso_type & SKB_GSO_TCPV6)
207 		return 1;
208 
209 	tcph = tcf_csum_skb_nextlayer(skb, ihl, ipl, sizeof(*tcph));
210 	if (tcph == NULL)
211 		return 0;
212 
213 	ip6h = ipv6_hdr(skb);
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 
225 static int tcf_csum_ipv4_udp(struct sk_buff *skb, unsigned int ihl,
226 			     unsigned int ipl, int udplite)
227 {
228 	struct udphdr *udph;
229 	const struct iphdr *iph;
230 	u16 ul;
231 
232 	/*
233 	 * Support both UDP and UDPLITE checksum algorithms, Don't use
234 	 * udph->len to get the real length without any protocol check,
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 
243 	iph = ip_hdr(skb);
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);
253 			else if ((ul >= sizeof(*udph)) && (ul <= ipl - ihl))
254 				skb->csum = csum_partial(udph, ul, 0);
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 
278 static int tcf_csum_ipv6_udp(struct sk_buff *skb, unsigned int ihl,
279 			     unsigned int ipl, int udplite)
280 {
281 	struct udphdr *udph;
282 	const struct ipv6hdr *ip6h;
283 	u16 ul;
284 
285 	/*
286 	 * Support both UDP and UDPLITE checksum algorithms, Don't use
287 	 * udph->len to get the real length without any protocol check,
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 
296 	ip6h = ipv6_hdr(skb);
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_sctp(struct sk_buff *skb, unsigned int ihl,
331 			 unsigned int ipl)
332 {
333 	struct sctphdr *sctph;
334 
335 	if (skb_is_gso(skb) && skb_shinfo(skb)->gso_type & SKB_GSO_SCTP)
336 		return 1;
337 
338 	sctph = tcf_csum_skb_nextlayer(skb, ihl, ipl, sizeof(*sctph));
339 	if (!sctph)
340 		return 0;
341 
342 	sctph->checksum = sctp_compute_cksum(skb,
343 					     skb_network_offset(skb) + ihl);
344 	skb->ip_summed = CHECKSUM_NONE;
345 	skb->csum_not_inet = 0;
346 
347 	return 1;
348 }
349 
350 static int tcf_csum_ipv4(struct sk_buff *skb, u32 update_flags)
351 {
352 	const struct iphdr *iph;
353 	int ntkoff;
354 
355 	ntkoff = skb_network_offset(skb);
356 
357 	if (!pskb_may_pull(skb, sizeof(*iph) + ntkoff))
358 		goto fail;
359 
360 	iph = ip_hdr(skb);
361 
362 	switch (iph->frag_off & htons(IP_OFFSET) ? 0 : iph->protocol) {
363 	case IPPROTO_ICMP:
364 		if (update_flags & TCA_CSUM_UPDATE_FLAG_ICMP)
365 			if (!tcf_csum_ipv4_icmp(skb, iph->ihl * 4,
366 						ntohs(iph->tot_len)))
367 				goto fail;
368 		break;
369 	case IPPROTO_IGMP:
370 		if (update_flags & TCA_CSUM_UPDATE_FLAG_IGMP)
371 			if (!tcf_csum_ipv4_igmp(skb, iph->ihl * 4,
372 						ntohs(iph->tot_len)))
373 				goto fail;
374 		break;
375 	case IPPROTO_TCP:
376 		if (update_flags & TCA_CSUM_UPDATE_FLAG_TCP)
377 			if (!tcf_csum_ipv4_tcp(skb, iph->ihl * 4,
378 					       ntohs(iph->tot_len)))
379 				goto fail;
380 		break;
381 	case IPPROTO_UDP:
382 		if (update_flags & TCA_CSUM_UPDATE_FLAG_UDP)
383 			if (!tcf_csum_ipv4_udp(skb, iph->ihl * 4,
384 					       ntohs(iph->tot_len), 0))
385 				goto fail;
386 		break;
387 	case IPPROTO_UDPLITE:
388 		if (update_flags & TCA_CSUM_UPDATE_FLAG_UDPLITE)
389 			if (!tcf_csum_ipv4_udp(skb, iph->ihl * 4,
390 					       ntohs(iph->tot_len), 1))
391 				goto fail;
392 		break;
393 	case IPPROTO_SCTP:
394 		if ((update_flags & TCA_CSUM_UPDATE_FLAG_SCTP) &&
395 		    !tcf_csum_sctp(skb, iph->ihl * 4, ntohs(iph->tot_len)))
396 			goto fail;
397 		break;
398 	}
399 
400 	if (update_flags & TCA_CSUM_UPDATE_FLAG_IPV4HDR) {
401 		if (skb_try_make_writable(skb, sizeof(*iph) + ntkoff))
402 			goto fail;
403 
404 		ip_send_check(ip_hdr(skb));
405 	}
406 
407 	return 1;
408 
409 fail:
410 	return 0;
411 }
412 
413 static int tcf_csum_ipv6_hopopts(struct ipv6_opt_hdr *ip6xh, unsigned int ixhl,
414 				 unsigned int *pl)
415 {
416 	int off, len, optlen;
417 	unsigned char *xh = (void *)ip6xh;
418 
419 	off = sizeof(*ip6xh);
420 	len = ixhl - off;
421 
422 	while (len > 1) {
423 		switch (xh[off]) {
424 		case IPV6_TLV_PAD1:
425 			optlen = 1;
426 			break;
427 		case IPV6_TLV_JUMBO:
428 			optlen = xh[off + 1] + 2;
429 			if (optlen != 6 || len < 6 || (off & 3) != 2)
430 				/* wrong jumbo option length/alignment */
431 				return 0;
432 			*pl = ntohl(*(__be32 *)(xh + off + 2));
433 			goto done;
434 		default:
435 			optlen = xh[off + 1] + 2;
436 			if (optlen > len)
437 				/* ignore obscure options */
438 				goto done;
439 			break;
440 		}
441 		off += optlen;
442 		len -= optlen;
443 	}
444 
445 done:
446 	return 1;
447 }
448 
449 static int tcf_csum_ipv6(struct sk_buff *skb, u32 update_flags)
450 {
451 	struct ipv6hdr *ip6h;
452 	struct ipv6_opt_hdr *ip6xh;
453 	unsigned int hl, ixhl;
454 	unsigned int pl;
455 	int ntkoff;
456 	u8 nexthdr;
457 
458 	ntkoff = skb_network_offset(skb);
459 
460 	hl = sizeof(*ip6h);
461 
462 	if (!pskb_may_pull(skb, hl + ntkoff))
463 		goto fail;
464 
465 	ip6h = ipv6_hdr(skb);
466 
467 	pl = ntohs(ip6h->payload_len);
468 	nexthdr = ip6h->nexthdr;
469 
470 	do {
471 		switch (nexthdr) {
472 		case NEXTHDR_FRAGMENT:
473 			goto ignore_skb;
474 		case NEXTHDR_ROUTING:
475 		case NEXTHDR_HOP:
476 		case NEXTHDR_DEST:
477 			if (!pskb_may_pull(skb, hl + sizeof(*ip6xh) + ntkoff))
478 				goto fail;
479 			ip6xh = (void *)(skb_network_header(skb) + hl);
480 			ixhl = ipv6_optlen(ip6xh);
481 			if (!pskb_may_pull(skb, hl + ixhl + ntkoff))
482 				goto fail;
483 			ip6xh = (void *)(skb_network_header(skb) + hl);
484 			if ((nexthdr == NEXTHDR_HOP) &&
485 			    !(tcf_csum_ipv6_hopopts(ip6xh, ixhl, &pl)))
486 				goto fail;
487 			nexthdr = ip6xh->nexthdr;
488 			hl += ixhl;
489 			break;
490 		case IPPROTO_ICMPV6:
491 			if (update_flags & TCA_CSUM_UPDATE_FLAG_ICMP)
492 				if (!tcf_csum_ipv6_icmp(skb,
493 							hl, pl + sizeof(*ip6h)))
494 					goto fail;
495 			goto done;
496 		case IPPROTO_TCP:
497 			if (update_flags & TCA_CSUM_UPDATE_FLAG_TCP)
498 				if (!tcf_csum_ipv6_tcp(skb,
499 						       hl, pl + sizeof(*ip6h)))
500 					goto fail;
501 			goto done;
502 		case IPPROTO_UDP:
503 			if (update_flags & TCA_CSUM_UPDATE_FLAG_UDP)
504 				if (!tcf_csum_ipv6_udp(skb, hl,
505 						       pl + sizeof(*ip6h), 0))
506 					goto fail;
507 			goto done;
508 		case IPPROTO_UDPLITE:
509 			if (update_flags & TCA_CSUM_UPDATE_FLAG_UDPLITE)
510 				if (!tcf_csum_ipv6_udp(skb, hl,
511 						       pl + sizeof(*ip6h), 1))
512 					goto fail;
513 			goto done;
514 		case IPPROTO_SCTP:
515 			if ((update_flags & TCA_CSUM_UPDATE_FLAG_SCTP) &&
516 			    !tcf_csum_sctp(skb, hl, pl + sizeof(*ip6h)))
517 				goto fail;
518 			goto done;
519 		default:
520 			goto ignore_skb;
521 		}
522 	} while (pskb_may_pull(skb, hl + 1 + ntkoff));
523 
524 done:
525 ignore_skb:
526 	return 1;
527 
528 fail:
529 	return 0;
530 }
531 
532 static int tcf_csum(struct sk_buff *skb, const struct tc_action *a,
533 		    struct tcf_result *res)
534 {
535 	struct tcf_csum *p = to_tcf_csum(a);
536 	int action;
537 	u32 update_flags;
538 
539 	spin_lock(&p->tcf_lock);
540 	tcf_lastuse_update(&p->tcf_tm);
541 	bstats_update(&p->tcf_bstats, skb);
542 	action = p->tcf_action;
543 	update_flags = p->update_flags;
544 	spin_unlock(&p->tcf_lock);
545 
546 	if (unlikely(action == TC_ACT_SHOT))
547 		goto drop;
548 
549 	switch (tc_skb_protocol(skb)) {
550 	case cpu_to_be16(ETH_P_IP):
551 		if (!tcf_csum_ipv4(skb, update_flags))
552 			goto drop;
553 		break;
554 	case cpu_to_be16(ETH_P_IPV6):
555 		if (!tcf_csum_ipv6(skb, update_flags))
556 			goto drop;
557 		break;
558 	}
559 
560 	return action;
561 
562 drop:
563 	spin_lock(&p->tcf_lock);
564 	p->tcf_qstats.drops++;
565 	spin_unlock(&p->tcf_lock);
566 	return TC_ACT_SHOT;
567 }
568 
569 static int tcf_csum_dump(struct sk_buff *skb, struct tc_action *a, int bind,
570 			 int ref)
571 {
572 	unsigned char *b = skb_tail_pointer(skb);
573 	struct tcf_csum *p = to_tcf_csum(a);
574 	struct tc_csum opt = {
575 		.update_flags = p->update_flags,
576 		.index   = p->tcf_index,
577 		.action  = p->tcf_action,
578 		.refcnt  = p->tcf_refcnt - ref,
579 		.bindcnt = p->tcf_bindcnt - bind,
580 	};
581 	struct tcf_t t;
582 
583 	if (nla_put(skb, TCA_CSUM_PARMS, sizeof(opt), &opt))
584 		goto nla_put_failure;
585 
586 	tcf_tm_dump(&t, &p->tcf_tm);
587 	if (nla_put_64bit(skb, TCA_CSUM_TM, sizeof(t), &t, TCA_CSUM_PAD))
588 		goto nla_put_failure;
589 
590 	return skb->len;
591 
592 nla_put_failure:
593 	nlmsg_trim(skb, b);
594 	return -1;
595 }
596 
597 static int tcf_csum_walker(struct net *net, struct sk_buff *skb,
598 			   struct netlink_callback *cb, int type,
599 			   const struct tc_action_ops *ops)
600 {
601 	struct tc_action_net *tn = net_generic(net, csum_net_id);
602 
603 	return tcf_generic_walker(tn, skb, cb, type, ops);
604 }
605 
606 static int tcf_csum_search(struct net *net, struct tc_action **a, u32 index)
607 {
608 	struct tc_action_net *tn = net_generic(net, csum_net_id);
609 
610 	return tcf_idr_search(tn, a, index);
611 }
612 
613 static struct tc_action_ops act_csum_ops = {
614 	.kind		= "csum",
615 	.type		= TCA_ACT_CSUM,
616 	.owner		= THIS_MODULE,
617 	.act		= tcf_csum,
618 	.dump		= tcf_csum_dump,
619 	.init		= tcf_csum_init,
620 	.walk		= tcf_csum_walker,
621 	.lookup		= tcf_csum_search,
622 	.size		= sizeof(struct tcf_csum),
623 };
624 
625 static __net_init int csum_init_net(struct net *net)
626 {
627 	struct tc_action_net *tn = net_generic(net, csum_net_id);
628 
629 	return tc_action_net_init(tn, &act_csum_ops);
630 }
631 
632 static void __net_exit csum_exit_net(struct net *net)
633 {
634 	struct tc_action_net *tn = net_generic(net, csum_net_id);
635 
636 	tc_action_net_exit(tn);
637 }
638 
639 static struct pernet_operations csum_net_ops = {
640 	.init = csum_init_net,
641 	.exit = csum_exit_net,
642 	.id   = &csum_net_id,
643 	.size = sizeof(struct tc_action_net),
644 };
645 
646 MODULE_DESCRIPTION("Checksum updating actions");
647 MODULE_LICENSE("GPL");
648 
649 static int __init csum_init_module(void)
650 {
651 	return tcf_register_action(&act_csum_ops, &csum_net_ops);
652 }
653 
654 static void __exit csum_cleanup_module(void)
655 {
656 	tcf_unregister_action(&act_csum_ops, &csum_net_ops);
657 }
658 
659 module_init(csum_init_module);
660 module_exit(csum_cleanup_module);
661