net: rose: mark expected switch fall-throughs
[linux-2.6-block.git] / net / sched / act_csum.c
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);