Loading...
1// SPDX-License-Identifier: GPL-2.0
2/* Copyright (c) 2017 - 2018 Covalent IO, Inc. http://covalent.io */
3
4#include <linux/skmsg.h>
5#include <linux/filter.h>
6#include <linux/bpf.h>
7#include <linux/init.h>
8#include <linux/wait.h>
9
10#include <net/inet_common.h>
11#include <net/tls.h>
12
13static bool tcp_bpf_stream_read(const struct sock *sk)
14{
15 struct sk_psock *psock;
16 bool empty = true;
17
18 rcu_read_lock();
19 psock = sk_psock(sk);
20 if (likely(psock))
21 empty = list_empty(&psock->ingress_msg);
22 rcu_read_unlock();
23 return !empty;
24}
25
26static int tcp_bpf_wait_data(struct sock *sk, struct sk_psock *psock,
27 int flags, long timeo, int *err)
28{
29 DEFINE_WAIT_FUNC(wait, woken_wake_function);
30 int ret = 0;
31
32 if (!timeo)
33 return ret;
34
35 add_wait_queue(sk_sleep(sk), &wait);
36 sk_set_bit(SOCKWQ_ASYNC_WAITDATA, sk);
37 ret = sk_wait_event(sk, &timeo,
38 !list_empty(&psock->ingress_msg) ||
39 !skb_queue_empty(&sk->sk_receive_queue), &wait);
40 sk_clear_bit(SOCKWQ_ASYNC_WAITDATA, sk);
41 remove_wait_queue(sk_sleep(sk), &wait);
42 return ret;
43}
44
45int __tcp_bpf_recvmsg(struct sock *sk, struct sk_psock *psock,
46 struct msghdr *msg, int len, int flags)
47{
48 struct iov_iter *iter = &msg->msg_iter;
49 int peek = flags & MSG_PEEK;
50 int i, ret, copied = 0;
51 struct sk_msg *msg_rx;
52
53 msg_rx = list_first_entry_or_null(&psock->ingress_msg,
54 struct sk_msg, list);
55
56 while (copied != len) {
57 struct scatterlist *sge;
58
59 if (unlikely(!msg_rx))
60 break;
61
62 i = msg_rx->sg.start;
63 do {
64 struct page *page;
65 int copy;
66
67 sge = sk_msg_elem(msg_rx, i);
68 copy = sge->length;
69 page = sg_page(sge);
70 if (copied + copy > len)
71 copy = len - copied;
72 ret = copy_page_to_iter(page, sge->offset, copy, iter);
73 if (ret != copy) {
74 msg_rx->sg.start = i;
75 return -EFAULT;
76 }
77
78 copied += copy;
79 if (likely(!peek)) {
80 sge->offset += copy;
81 sge->length -= copy;
82 sk_mem_uncharge(sk, copy);
83 msg_rx->sg.size -= copy;
84
85 if (!sge->length) {
86 sk_msg_iter_var_next(i);
87 if (!msg_rx->skb)
88 put_page(page);
89 }
90 } else {
91 sk_msg_iter_var_next(i);
92 }
93
94 if (copied == len)
95 break;
96 } while (i != msg_rx->sg.end);
97
98 if (unlikely(peek)) {
99 msg_rx = list_next_entry(msg_rx, list);
100 continue;
101 }
102
103 msg_rx->sg.start = i;
104 if (!sge->length && msg_rx->sg.start == msg_rx->sg.end) {
105 list_del(&msg_rx->list);
106 if (msg_rx->skb)
107 consume_skb(msg_rx->skb);
108 kfree(msg_rx);
109 }
110 msg_rx = list_first_entry_or_null(&psock->ingress_msg,
111 struct sk_msg, list);
112 }
113
114 return copied;
115}
116EXPORT_SYMBOL_GPL(__tcp_bpf_recvmsg);
117
118int tcp_bpf_recvmsg(struct sock *sk, struct msghdr *msg, size_t len,
119 int nonblock, int flags, int *addr_len)
120{
121 struct sk_psock *psock;
122 int copied, ret;
123
124 if (unlikely(flags & MSG_ERRQUEUE))
125 return inet_recv_error(sk, msg, len, addr_len);
126 if (!skb_queue_empty(&sk->sk_receive_queue))
127 return tcp_recvmsg(sk, msg, len, nonblock, flags, addr_len);
128
129 psock = sk_psock_get(sk);
130 if (unlikely(!psock))
131 return tcp_recvmsg(sk, msg, len, nonblock, flags, addr_len);
132 lock_sock(sk);
133msg_bytes_ready:
134 copied = __tcp_bpf_recvmsg(sk, psock, msg, len, flags);
135 if (!copied) {
136 int data, err = 0;
137 long timeo;
138
139 timeo = sock_rcvtimeo(sk, nonblock);
140 data = tcp_bpf_wait_data(sk, psock, flags, timeo, &err);
141 if (data) {
142 if (skb_queue_empty(&sk->sk_receive_queue))
143 goto msg_bytes_ready;
144 release_sock(sk);
145 sk_psock_put(sk, psock);
146 return tcp_recvmsg(sk, msg, len, nonblock, flags, addr_len);
147 }
148 if (err) {
149 ret = err;
150 goto out;
151 }
152 copied = -EAGAIN;
153 }
154 ret = copied;
155out:
156 release_sock(sk);
157 sk_psock_put(sk, psock);
158 return ret;
159}
160
161static int bpf_tcp_ingress(struct sock *sk, struct sk_psock *psock,
162 struct sk_msg *msg, u32 apply_bytes, int flags)
163{
164 bool apply = apply_bytes;
165 struct scatterlist *sge;
166 u32 size, copied = 0;
167 struct sk_msg *tmp;
168 int i, ret = 0;
169
170 tmp = kzalloc(sizeof(*tmp), __GFP_NOWARN | GFP_KERNEL);
171 if (unlikely(!tmp))
172 return -ENOMEM;
173
174 lock_sock(sk);
175 tmp->sg.start = msg->sg.start;
176 i = msg->sg.start;
177 do {
178 sge = sk_msg_elem(msg, i);
179 size = (apply && apply_bytes < sge->length) ?
180 apply_bytes : sge->length;
181 if (!sk_wmem_schedule(sk, size)) {
182 if (!copied)
183 ret = -ENOMEM;
184 break;
185 }
186
187 sk_mem_charge(sk, size);
188 sk_msg_xfer(tmp, msg, i, size);
189 copied += size;
190 if (sge->length)
191 get_page(sk_msg_page(tmp, i));
192 sk_msg_iter_var_next(i);
193 tmp->sg.end = i;
194 if (apply) {
195 apply_bytes -= size;
196 if (!apply_bytes)
197 break;
198 }
199 } while (i != msg->sg.end);
200
201 if (!ret) {
202 msg->sg.start = i;
203 msg->sg.size -= apply_bytes;
204 sk_psock_queue_msg(psock, tmp);
205 sk_psock_data_ready(sk, psock);
206 } else {
207 sk_msg_free(sk, tmp);
208 kfree(tmp);
209 }
210
211 release_sock(sk);
212 return ret;
213}
214
215static int tcp_bpf_push(struct sock *sk, struct sk_msg *msg, u32 apply_bytes,
216 int flags, bool uncharge)
217{
218 bool apply = apply_bytes;
219 struct scatterlist *sge;
220 struct page *page;
221 int size, ret = 0;
222 u32 off;
223
224 while (1) {
225 bool has_tx_ulp;
226
227 sge = sk_msg_elem(msg, msg->sg.start);
228 size = (apply && apply_bytes < sge->length) ?
229 apply_bytes : sge->length;
230 off = sge->offset;
231 page = sg_page(sge);
232
233 tcp_rate_check_app_limited(sk);
234retry:
235 has_tx_ulp = tls_sw_has_ctx_tx(sk);
236 if (has_tx_ulp) {
237 flags |= MSG_SENDPAGE_NOPOLICY;
238 ret = kernel_sendpage_locked(sk,
239 page, off, size, flags);
240 } else {
241 ret = do_tcp_sendpages(sk, page, off, size, flags);
242 }
243
244 if (ret <= 0)
245 return ret;
246 if (apply)
247 apply_bytes -= ret;
248 msg->sg.size -= ret;
249 sge->offset += ret;
250 sge->length -= ret;
251 if (uncharge)
252 sk_mem_uncharge(sk, ret);
253 if (ret != size) {
254 size -= ret;
255 off += ret;
256 goto retry;
257 }
258 if (!sge->length) {
259 put_page(page);
260 sk_msg_iter_next(msg, start);
261 sg_init_table(sge, 1);
262 if (msg->sg.start == msg->sg.end)
263 break;
264 }
265 if (apply && !apply_bytes)
266 break;
267 }
268
269 return 0;
270}
271
272static int tcp_bpf_push_locked(struct sock *sk, struct sk_msg *msg,
273 u32 apply_bytes, int flags, bool uncharge)
274{
275 int ret;
276
277 lock_sock(sk);
278 ret = tcp_bpf_push(sk, msg, apply_bytes, flags, uncharge);
279 release_sock(sk);
280 return ret;
281}
282
283int tcp_bpf_sendmsg_redir(struct sock *sk, struct sk_msg *msg,
284 u32 bytes, int flags)
285{
286 bool ingress = sk_msg_to_ingress(msg);
287 struct sk_psock *psock = sk_psock_get(sk);
288 int ret;
289
290 if (unlikely(!psock)) {
291 sk_msg_free(sk, msg);
292 return 0;
293 }
294 ret = ingress ? bpf_tcp_ingress(sk, psock, msg, bytes, flags) :
295 tcp_bpf_push_locked(sk, msg, bytes, flags, false);
296 sk_psock_put(sk, psock);
297 return ret;
298}
299EXPORT_SYMBOL_GPL(tcp_bpf_sendmsg_redir);
300
301static int tcp_bpf_send_verdict(struct sock *sk, struct sk_psock *psock,
302 struct sk_msg *msg, int *copied, int flags)
303{
304 bool cork = false, enospc = msg->sg.start == msg->sg.end;
305 struct sock *sk_redir;
306 u32 tosend, delta = 0;
307 int ret;
308
309more_data:
310 if (psock->eval == __SK_NONE) {
311 /* Track delta in msg size to add/subtract it on SK_DROP from
312 * returned to user copied size. This ensures user doesn't
313 * get a positive return code with msg_cut_data and SK_DROP
314 * verdict.
315 */
316 delta = msg->sg.size;
317 psock->eval = sk_psock_msg_verdict(sk, psock, msg);
318 if (msg->sg.size < delta)
319 delta -= msg->sg.size;
320 else
321 delta = 0;
322 }
323
324 if (msg->cork_bytes &&
325 msg->cork_bytes > msg->sg.size && !enospc) {
326 psock->cork_bytes = msg->cork_bytes - msg->sg.size;
327 if (!psock->cork) {
328 psock->cork = kzalloc(sizeof(*psock->cork),
329 GFP_ATOMIC | __GFP_NOWARN);
330 if (!psock->cork)
331 return -ENOMEM;
332 }
333 memcpy(psock->cork, msg, sizeof(*msg));
334 return 0;
335 }
336
337 tosend = msg->sg.size;
338 if (psock->apply_bytes && psock->apply_bytes < tosend)
339 tosend = psock->apply_bytes;
340
341 switch (psock->eval) {
342 case __SK_PASS:
343 ret = tcp_bpf_push(sk, msg, tosend, flags, true);
344 if (unlikely(ret)) {
345 *copied -= sk_msg_free(sk, msg);
346 break;
347 }
348 sk_msg_apply_bytes(psock, tosend);
349 break;
350 case __SK_REDIRECT:
351 sk_redir = psock->sk_redir;
352 sk_msg_apply_bytes(psock, tosend);
353 if (psock->cork) {
354 cork = true;
355 psock->cork = NULL;
356 }
357 sk_msg_return(sk, msg, tosend);
358 release_sock(sk);
359 ret = tcp_bpf_sendmsg_redir(sk_redir, msg, tosend, flags);
360 lock_sock(sk);
361 if (unlikely(ret < 0)) {
362 int free = sk_msg_free_nocharge(sk, msg);
363
364 if (!cork)
365 *copied -= free;
366 }
367 if (cork) {
368 sk_msg_free(sk, msg);
369 kfree(msg);
370 msg = NULL;
371 ret = 0;
372 }
373 break;
374 case __SK_DROP:
375 default:
376 sk_msg_free_partial(sk, msg, tosend);
377 sk_msg_apply_bytes(psock, tosend);
378 *copied -= (tosend + delta);
379 return -EACCES;
380 }
381
382 if (likely(!ret)) {
383 if (!psock->apply_bytes) {
384 psock->eval = __SK_NONE;
385 if (psock->sk_redir) {
386 sock_put(psock->sk_redir);
387 psock->sk_redir = NULL;
388 }
389 }
390 if (msg &&
391 msg->sg.data[msg->sg.start].page_link &&
392 msg->sg.data[msg->sg.start].length)
393 goto more_data;
394 }
395 return ret;
396}
397
398static int tcp_bpf_sendmsg(struct sock *sk, struct msghdr *msg, size_t size)
399{
400 struct sk_msg tmp, *msg_tx = NULL;
401 int copied = 0, err = 0;
402 struct sk_psock *psock;
403 long timeo;
404 int flags;
405
406 /* Don't let internal do_tcp_sendpages() flags through */
407 flags = (msg->msg_flags & ~MSG_SENDPAGE_DECRYPTED);
408 flags |= MSG_NO_SHARED_FRAGS;
409
410 psock = sk_psock_get(sk);
411 if (unlikely(!psock))
412 return tcp_sendmsg(sk, msg, size);
413
414 lock_sock(sk);
415 timeo = sock_sndtimeo(sk, msg->msg_flags & MSG_DONTWAIT);
416 while (msg_data_left(msg)) {
417 bool enospc = false;
418 u32 copy, osize;
419
420 if (sk->sk_err) {
421 err = -sk->sk_err;
422 goto out_err;
423 }
424
425 copy = msg_data_left(msg);
426 if (!sk_stream_memory_free(sk))
427 goto wait_for_sndbuf;
428 if (psock->cork) {
429 msg_tx = psock->cork;
430 } else {
431 msg_tx = &tmp;
432 sk_msg_init(msg_tx);
433 }
434
435 osize = msg_tx->sg.size;
436 err = sk_msg_alloc(sk, msg_tx, msg_tx->sg.size + copy, msg_tx->sg.end - 1);
437 if (err) {
438 if (err != -ENOSPC)
439 goto wait_for_memory;
440 enospc = true;
441 copy = msg_tx->sg.size - osize;
442 }
443
444 err = sk_msg_memcopy_from_iter(sk, &msg->msg_iter, msg_tx,
445 copy);
446 if (err < 0) {
447 sk_msg_trim(sk, msg_tx, osize);
448 goto out_err;
449 }
450
451 copied += copy;
452 if (psock->cork_bytes) {
453 if (size > psock->cork_bytes)
454 psock->cork_bytes = 0;
455 else
456 psock->cork_bytes -= size;
457 if (psock->cork_bytes && !enospc)
458 goto out_err;
459 /* All cork bytes are accounted, rerun the prog. */
460 psock->eval = __SK_NONE;
461 psock->cork_bytes = 0;
462 }
463
464 err = tcp_bpf_send_verdict(sk, psock, msg_tx, &copied, flags);
465 if (unlikely(err < 0))
466 goto out_err;
467 continue;
468wait_for_sndbuf:
469 set_bit(SOCK_NOSPACE, &sk->sk_socket->flags);
470wait_for_memory:
471 err = sk_stream_wait_memory(sk, &timeo);
472 if (err) {
473 if (msg_tx && msg_tx != psock->cork)
474 sk_msg_free(sk, msg_tx);
475 goto out_err;
476 }
477 }
478out_err:
479 if (err < 0)
480 err = sk_stream_error(sk, msg->msg_flags, err);
481 release_sock(sk);
482 sk_psock_put(sk, psock);
483 return copied ? copied : err;
484}
485
486static int tcp_bpf_sendpage(struct sock *sk, struct page *page, int offset,
487 size_t size, int flags)
488{
489 struct sk_msg tmp, *msg = NULL;
490 int err = 0, copied = 0;
491 struct sk_psock *psock;
492 bool enospc = false;
493
494 psock = sk_psock_get(sk);
495 if (unlikely(!psock))
496 return tcp_sendpage(sk, page, offset, size, flags);
497
498 lock_sock(sk);
499 if (psock->cork) {
500 msg = psock->cork;
501 } else {
502 msg = &tmp;
503 sk_msg_init(msg);
504 }
505
506 /* Catch case where ring is full and sendpage is stalled. */
507 if (unlikely(sk_msg_full(msg)))
508 goto out_err;
509
510 sk_msg_page_add(msg, page, size, offset);
511 sk_mem_charge(sk, size);
512 copied = size;
513 if (sk_msg_full(msg))
514 enospc = true;
515 if (psock->cork_bytes) {
516 if (size > psock->cork_bytes)
517 psock->cork_bytes = 0;
518 else
519 psock->cork_bytes -= size;
520 if (psock->cork_bytes && !enospc)
521 goto out_err;
522 /* All cork bytes are accounted, rerun the prog. */
523 psock->eval = __SK_NONE;
524 psock->cork_bytes = 0;
525 }
526
527 err = tcp_bpf_send_verdict(sk, psock, msg, &copied, flags);
528out_err:
529 release_sock(sk);
530 sk_psock_put(sk, psock);
531 return copied ? copied : err;
532}
533
534static void tcp_bpf_remove(struct sock *sk, struct sk_psock *psock)
535{
536 struct sk_psock_link *link;
537
538 while ((link = sk_psock_link_pop(psock))) {
539 sk_psock_unlink(sk, link);
540 sk_psock_free_link(link);
541 }
542}
543
544static void tcp_bpf_unhash(struct sock *sk)
545{
546 void (*saved_unhash)(struct sock *sk);
547 struct sk_psock *psock;
548
549 rcu_read_lock();
550 psock = sk_psock(sk);
551 if (unlikely(!psock)) {
552 rcu_read_unlock();
553 if (sk->sk_prot->unhash)
554 sk->sk_prot->unhash(sk);
555 return;
556 }
557
558 saved_unhash = psock->saved_unhash;
559 tcp_bpf_remove(sk, psock);
560 rcu_read_unlock();
561 saved_unhash(sk);
562}
563
564static void tcp_bpf_close(struct sock *sk, long timeout)
565{
566 void (*saved_close)(struct sock *sk, long timeout);
567 struct sk_psock *psock;
568
569 lock_sock(sk);
570 rcu_read_lock();
571 psock = sk_psock(sk);
572 if (unlikely(!psock)) {
573 rcu_read_unlock();
574 release_sock(sk);
575 return sk->sk_prot->close(sk, timeout);
576 }
577
578 saved_close = psock->saved_close;
579 tcp_bpf_remove(sk, psock);
580 rcu_read_unlock();
581 release_sock(sk);
582 saved_close(sk, timeout);
583}
584
585enum {
586 TCP_BPF_IPV4,
587 TCP_BPF_IPV6,
588 TCP_BPF_NUM_PROTS,
589};
590
591enum {
592 TCP_BPF_BASE,
593 TCP_BPF_TX,
594 TCP_BPF_NUM_CFGS,
595};
596
597static struct proto *tcpv6_prot_saved __read_mostly;
598static DEFINE_SPINLOCK(tcpv6_prot_lock);
599static struct proto tcp_bpf_prots[TCP_BPF_NUM_PROTS][TCP_BPF_NUM_CFGS];
600
601static void tcp_bpf_rebuild_protos(struct proto prot[TCP_BPF_NUM_CFGS],
602 struct proto *base)
603{
604 prot[TCP_BPF_BASE] = *base;
605 prot[TCP_BPF_BASE].unhash = tcp_bpf_unhash;
606 prot[TCP_BPF_BASE].close = tcp_bpf_close;
607 prot[TCP_BPF_BASE].recvmsg = tcp_bpf_recvmsg;
608 prot[TCP_BPF_BASE].stream_memory_read = tcp_bpf_stream_read;
609
610 prot[TCP_BPF_TX] = prot[TCP_BPF_BASE];
611 prot[TCP_BPF_TX].sendmsg = tcp_bpf_sendmsg;
612 prot[TCP_BPF_TX].sendpage = tcp_bpf_sendpage;
613}
614
615static void tcp_bpf_check_v6_needs_rebuild(struct sock *sk, struct proto *ops)
616{
617 if (sk->sk_family == AF_INET6 &&
618 unlikely(ops != smp_load_acquire(&tcpv6_prot_saved))) {
619 spin_lock_bh(&tcpv6_prot_lock);
620 if (likely(ops != tcpv6_prot_saved)) {
621 tcp_bpf_rebuild_protos(tcp_bpf_prots[TCP_BPF_IPV6], ops);
622 smp_store_release(&tcpv6_prot_saved, ops);
623 }
624 spin_unlock_bh(&tcpv6_prot_lock);
625 }
626}
627
628static int __init tcp_bpf_v4_build_proto(void)
629{
630 tcp_bpf_rebuild_protos(tcp_bpf_prots[TCP_BPF_IPV4], &tcp_prot);
631 return 0;
632}
633core_initcall(tcp_bpf_v4_build_proto);
634
635static void tcp_bpf_update_sk_prot(struct sock *sk, struct sk_psock *psock)
636{
637 int family = sk->sk_family == AF_INET6 ? TCP_BPF_IPV6 : TCP_BPF_IPV4;
638 int config = psock->progs.msg_parser ? TCP_BPF_TX : TCP_BPF_BASE;
639
640 sk_psock_update_proto(sk, psock, &tcp_bpf_prots[family][config]);
641}
642
643static void tcp_bpf_reinit_sk_prot(struct sock *sk, struct sk_psock *psock)
644{
645 int family = sk->sk_family == AF_INET6 ? TCP_BPF_IPV6 : TCP_BPF_IPV4;
646 int config = psock->progs.msg_parser ? TCP_BPF_TX : TCP_BPF_BASE;
647
648 /* Reinit occurs when program types change e.g. TCP_BPF_TX is removed
649 * or added requiring sk_prot hook updates. We keep original saved
650 * hooks in this case.
651 */
652 sk->sk_prot = &tcp_bpf_prots[family][config];
653}
654
655static int tcp_bpf_assert_proto_ops(struct proto *ops)
656{
657 /* In order to avoid retpoline, we make assumptions when we call
658 * into ops if e.g. a psock is not present. Make sure they are
659 * indeed valid assumptions.
660 */
661 return ops->recvmsg == tcp_recvmsg &&
662 ops->sendmsg == tcp_sendmsg &&
663 ops->sendpage == tcp_sendpage ? 0 : -ENOTSUPP;
664}
665
666void tcp_bpf_reinit(struct sock *sk)
667{
668 struct sk_psock *psock;
669
670 sock_owned_by_me(sk);
671
672 rcu_read_lock();
673 psock = sk_psock(sk);
674 tcp_bpf_reinit_sk_prot(sk, psock);
675 rcu_read_unlock();
676}
677
678int tcp_bpf_init(struct sock *sk)
679{
680 struct proto *ops = READ_ONCE(sk->sk_prot);
681 struct sk_psock *psock;
682
683 sock_owned_by_me(sk);
684
685 rcu_read_lock();
686 psock = sk_psock(sk);
687 if (unlikely(!psock || psock->sk_proto ||
688 tcp_bpf_assert_proto_ops(ops))) {
689 rcu_read_unlock();
690 return -EINVAL;
691 }
692 tcp_bpf_check_v6_needs_rebuild(sk, ops);
693 tcp_bpf_update_sk_prot(sk, psock);
694 rcu_read_unlock();
695 return 0;
696}
1// SPDX-License-Identifier: GPL-2.0
2/* Copyright (c) 2017 - 2018 Covalent IO, Inc. http://covalent.io */
3
4#include <linux/skmsg.h>
5#include <linux/filter.h>
6#include <linux/bpf.h>
7#include <linux/init.h>
8#include <linux/wait.h>
9
10#include <net/inet_common.h>
11#include <net/tls.h>
12
13static int bpf_tcp_ingress(struct sock *sk, struct sk_psock *psock,
14 struct sk_msg *msg, u32 apply_bytes, int flags)
15{
16 bool apply = apply_bytes;
17 struct scatterlist *sge;
18 u32 size, copied = 0;
19 struct sk_msg *tmp;
20 int i, ret = 0;
21
22 tmp = kzalloc(sizeof(*tmp), __GFP_NOWARN | GFP_KERNEL);
23 if (unlikely(!tmp))
24 return -ENOMEM;
25
26 lock_sock(sk);
27 tmp->sg.start = msg->sg.start;
28 i = msg->sg.start;
29 do {
30 sge = sk_msg_elem(msg, i);
31 size = (apply && apply_bytes < sge->length) ?
32 apply_bytes : sge->length;
33 if (!sk_wmem_schedule(sk, size)) {
34 if (!copied)
35 ret = -ENOMEM;
36 break;
37 }
38
39 sk_mem_charge(sk, size);
40 sk_msg_xfer(tmp, msg, i, size);
41 copied += size;
42 if (sge->length)
43 get_page(sk_msg_page(tmp, i));
44 sk_msg_iter_var_next(i);
45 tmp->sg.end = i;
46 if (apply) {
47 apply_bytes -= size;
48 if (!apply_bytes)
49 break;
50 }
51 } while (i != msg->sg.end);
52
53 if (!ret) {
54 msg->sg.start = i;
55 sk_psock_queue_msg(psock, tmp);
56 sk_psock_data_ready(sk, psock);
57 } else {
58 sk_msg_free(sk, tmp);
59 kfree(tmp);
60 }
61
62 release_sock(sk);
63 return ret;
64}
65
66static int tcp_bpf_push(struct sock *sk, struct sk_msg *msg, u32 apply_bytes,
67 int flags, bool uncharge)
68{
69 bool apply = apply_bytes;
70 struct scatterlist *sge;
71 struct page *page;
72 int size, ret = 0;
73 u32 off;
74
75 while (1) {
76 bool has_tx_ulp;
77
78 sge = sk_msg_elem(msg, msg->sg.start);
79 size = (apply && apply_bytes < sge->length) ?
80 apply_bytes : sge->length;
81 off = sge->offset;
82 page = sg_page(sge);
83
84 tcp_rate_check_app_limited(sk);
85retry:
86 has_tx_ulp = tls_sw_has_ctx_tx(sk);
87 if (has_tx_ulp) {
88 flags |= MSG_SENDPAGE_NOPOLICY;
89 ret = kernel_sendpage_locked(sk,
90 page, off, size, flags);
91 } else {
92 ret = do_tcp_sendpages(sk, page, off, size, flags);
93 }
94
95 if (ret <= 0)
96 return ret;
97 if (apply)
98 apply_bytes -= ret;
99 msg->sg.size -= ret;
100 sge->offset += ret;
101 sge->length -= ret;
102 if (uncharge)
103 sk_mem_uncharge(sk, ret);
104 if (ret != size) {
105 size -= ret;
106 off += ret;
107 goto retry;
108 }
109 if (!sge->length) {
110 put_page(page);
111 sk_msg_iter_next(msg, start);
112 sg_init_table(sge, 1);
113 if (msg->sg.start == msg->sg.end)
114 break;
115 }
116 if (apply && !apply_bytes)
117 break;
118 }
119
120 return 0;
121}
122
123static int tcp_bpf_push_locked(struct sock *sk, struct sk_msg *msg,
124 u32 apply_bytes, int flags, bool uncharge)
125{
126 int ret;
127
128 lock_sock(sk);
129 ret = tcp_bpf_push(sk, msg, apply_bytes, flags, uncharge);
130 release_sock(sk);
131 return ret;
132}
133
134int tcp_bpf_sendmsg_redir(struct sock *sk, struct sk_msg *msg,
135 u32 bytes, int flags)
136{
137 bool ingress = sk_msg_to_ingress(msg);
138 struct sk_psock *psock = sk_psock_get(sk);
139 int ret;
140
141 if (unlikely(!psock)) {
142 sk_msg_free(sk, msg);
143 return 0;
144 }
145 ret = ingress ? bpf_tcp_ingress(sk, psock, msg, bytes, flags) :
146 tcp_bpf_push_locked(sk, msg, bytes, flags, false);
147 sk_psock_put(sk, psock);
148 return ret;
149}
150EXPORT_SYMBOL_GPL(tcp_bpf_sendmsg_redir);
151
152#ifdef CONFIG_BPF_SYSCALL
153static bool tcp_bpf_stream_read(const struct sock *sk)
154{
155 struct sk_psock *psock;
156 bool empty = true;
157
158 rcu_read_lock();
159 psock = sk_psock(sk);
160 if (likely(psock))
161 empty = list_empty(&psock->ingress_msg);
162 rcu_read_unlock();
163 return !empty;
164}
165
166static int tcp_msg_wait_data(struct sock *sk, struct sk_psock *psock,
167 long timeo)
168{
169 DEFINE_WAIT_FUNC(wait, woken_wake_function);
170 int ret = 0;
171
172 if (sk->sk_shutdown & RCV_SHUTDOWN)
173 return 1;
174
175 if (!timeo)
176 return ret;
177
178 add_wait_queue(sk_sleep(sk), &wait);
179 sk_set_bit(SOCKWQ_ASYNC_WAITDATA, sk);
180 ret = sk_wait_event(sk, &timeo,
181 !list_empty(&psock->ingress_msg) ||
182 !skb_queue_empty(&sk->sk_receive_queue), &wait);
183 sk_clear_bit(SOCKWQ_ASYNC_WAITDATA, sk);
184 remove_wait_queue(sk_sleep(sk), &wait);
185 return ret;
186}
187
188static int tcp_bpf_recvmsg(struct sock *sk, struct msghdr *msg, size_t len,
189 int nonblock, int flags, int *addr_len)
190{
191 struct sk_psock *psock;
192 int copied, ret;
193
194 if (unlikely(flags & MSG_ERRQUEUE))
195 return inet_recv_error(sk, msg, len, addr_len);
196
197 psock = sk_psock_get(sk);
198 if (unlikely(!psock))
199 return tcp_recvmsg(sk, msg, len, nonblock, flags, addr_len);
200 if (!skb_queue_empty(&sk->sk_receive_queue) &&
201 sk_psock_queue_empty(psock)) {
202 sk_psock_put(sk, psock);
203 return tcp_recvmsg(sk, msg, len, nonblock, flags, addr_len);
204 }
205 lock_sock(sk);
206msg_bytes_ready:
207 copied = sk_msg_recvmsg(sk, psock, msg, len, flags);
208 if (!copied) {
209 long timeo;
210 int data;
211
212 timeo = sock_rcvtimeo(sk, nonblock);
213 data = tcp_msg_wait_data(sk, psock, timeo);
214 if (data) {
215 if (!sk_psock_queue_empty(psock))
216 goto msg_bytes_ready;
217 release_sock(sk);
218 sk_psock_put(sk, psock);
219 return tcp_recvmsg(sk, msg, len, nonblock, flags, addr_len);
220 }
221 copied = -EAGAIN;
222 }
223 ret = copied;
224 release_sock(sk);
225 sk_psock_put(sk, psock);
226 return ret;
227}
228
229static int tcp_bpf_send_verdict(struct sock *sk, struct sk_psock *psock,
230 struct sk_msg *msg, int *copied, int flags)
231{
232 bool cork = false, enospc = sk_msg_full(msg);
233 struct sock *sk_redir;
234 u32 tosend, delta = 0;
235 int ret;
236
237more_data:
238 if (psock->eval == __SK_NONE) {
239 /* Track delta in msg size to add/subtract it on SK_DROP from
240 * returned to user copied size. This ensures user doesn't
241 * get a positive return code with msg_cut_data and SK_DROP
242 * verdict.
243 */
244 delta = msg->sg.size;
245 psock->eval = sk_psock_msg_verdict(sk, psock, msg);
246 delta -= msg->sg.size;
247 }
248
249 if (msg->cork_bytes &&
250 msg->cork_bytes > msg->sg.size && !enospc) {
251 psock->cork_bytes = msg->cork_bytes - msg->sg.size;
252 if (!psock->cork) {
253 psock->cork = kzalloc(sizeof(*psock->cork),
254 GFP_ATOMIC | __GFP_NOWARN);
255 if (!psock->cork)
256 return -ENOMEM;
257 }
258 memcpy(psock->cork, msg, sizeof(*msg));
259 return 0;
260 }
261
262 tosend = msg->sg.size;
263 if (psock->apply_bytes && psock->apply_bytes < tosend)
264 tosend = psock->apply_bytes;
265
266 switch (psock->eval) {
267 case __SK_PASS:
268 ret = tcp_bpf_push(sk, msg, tosend, flags, true);
269 if (unlikely(ret)) {
270 *copied -= sk_msg_free(sk, msg);
271 break;
272 }
273 sk_msg_apply_bytes(psock, tosend);
274 break;
275 case __SK_REDIRECT:
276 sk_redir = psock->sk_redir;
277 sk_msg_apply_bytes(psock, tosend);
278 if (psock->cork) {
279 cork = true;
280 psock->cork = NULL;
281 }
282 sk_msg_return(sk, msg, tosend);
283 release_sock(sk);
284 ret = tcp_bpf_sendmsg_redir(sk_redir, msg, tosend, flags);
285 lock_sock(sk);
286 if (unlikely(ret < 0)) {
287 int free = sk_msg_free_nocharge(sk, msg);
288
289 if (!cork)
290 *copied -= free;
291 }
292 if (cork) {
293 sk_msg_free(sk, msg);
294 kfree(msg);
295 msg = NULL;
296 ret = 0;
297 }
298 break;
299 case __SK_DROP:
300 default:
301 sk_msg_free_partial(sk, msg, tosend);
302 sk_msg_apply_bytes(psock, tosend);
303 *copied -= (tosend + delta);
304 return -EACCES;
305 }
306
307 if (likely(!ret)) {
308 if (!psock->apply_bytes) {
309 psock->eval = __SK_NONE;
310 if (psock->sk_redir) {
311 sock_put(psock->sk_redir);
312 psock->sk_redir = NULL;
313 }
314 }
315 if (msg &&
316 msg->sg.data[msg->sg.start].page_link &&
317 msg->sg.data[msg->sg.start].length)
318 goto more_data;
319 }
320 return ret;
321}
322
323static int tcp_bpf_sendmsg(struct sock *sk, struct msghdr *msg, size_t size)
324{
325 struct sk_msg tmp, *msg_tx = NULL;
326 int copied = 0, err = 0;
327 struct sk_psock *psock;
328 long timeo;
329 int flags;
330
331 /* Don't let internal do_tcp_sendpages() flags through */
332 flags = (msg->msg_flags & ~MSG_SENDPAGE_DECRYPTED);
333 flags |= MSG_NO_SHARED_FRAGS;
334
335 psock = sk_psock_get(sk);
336 if (unlikely(!psock))
337 return tcp_sendmsg(sk, msg, size);
338
339 lock_sock(sk);
340 timeo = sock_sndtimeo(sk, msg->msg_flags & MSG_DONTWAIT);
341 while (msg_data_left(msg)) {
342 bool enospc = false;
343 u32 copy, osize;
344
345 if (sk->sk_err) {
346 err = -sk->sk_err;
347 goto out_err;
348 }
349
350 copy = msg_data_left(msg);
351 if (!sk_stream_memory_free(sk))
352 goto wait_for_sndbuf;
353 if (psock->cork) {
354 msg_tx = psock->cork;
355 } else {
356 msg_tx = &tmp;
357 sk_msg_init(msg_tx);
358 }
359
360 osize = msg_tx->sg.size;
361 err = sk_msg_alloc(sk, msg_tx, msg_tx->sg.size + copy, msg_tx->sg.end - 1);
362 if (err) {
363 if (err != -ENOSPC)
364 goto wait_for_memory;
365 enospc = true;
366 copy = msg_tx->sg.size - osize;
367 }
368
369 err = sk_msg_memcopy_from_iter(sk, &msg->msg_iter, msg_tx,
370 copy);
371 if (err < 0) {
372 sk_msg_trim(sk, msg_tx, osize);
373 goto out_err;
374 }
375
376 copied += copy;
377 if (psock->cork_bytes) {
378 if (size > psock->cork_bytes)
379 psock->cork_bytes = 0;
380 else
381 psock->cork_bytes -= size;
382 if (psock->cork_bytes && !enospc)
383 goto out_err;
384 /* All cork bytes are accounted, rerun the prog. */
385 psock->eval = __SK_NONE;
386 psock->cork_bytes = 0;
387 }
388
389 err = tcp_bpf_send_verdict(sk, psock, msg_tx, &copied, flags);
390 if (unlikely(err < 0))
391 goto out_err;
392 continue;
393wait_for_sndbuf:
394 set_bit(SOCK_NOSPACE, &sk->sk_socket->flags);
395wait_for_memory:
396 err = sk_stream_wait_memory(sk, &timeo);
397 if (err) {
398 if (msg_tx && msg_tx != psock->cork)
399 sk_msg_free(sk, msg_tx);
400 goto out_err;
401 }
402 }
403out_err:
404 if (err < 0)
405 err = sk_stream_error(sk, msg->msg_flags, err);
406 release_sock(sk);
407 sk_psock_put(sk, psock);
408 return copied ? copied : err;
409}
410
411static int tcp_bpf_sendpage(struct sock *sk, struct page *page, int offset,
412 size_t size, int flags)
413{
414 struct sk_msg tmp, *msg = NULL;
415 int err = 0, copied = 0;
416 struct sk_psock *psock;
417 bool enospc = false;
418
419 psock = sk_psock_get(sk);
420 if (unlikely(!psock))
421 return tcp_sendpage(sk, page, offset, size, flags);
422
423 lock_sock(sk);
424 if (psock->cork) {
425 msg = psock->cork;
426 } else {
427 msg = &tmp;
428 sk_msg_init(msg);
429 }
430
431 /* Catch case where ring is full and sendpage is stalled. */
432 if (unlikely(sk_msg_full(msg)))
433 goto out_err;
434
435 sk_msg_page_add(msg, page, size, offset);
436 sk_mem_charge(sk, size);
437 copied = size;
438 if (sk_msg_full(msg))
439 enospc = true;
440 if (psock->cork_bytes) {
441 if (size > psock->cork_bytes)
442 psock->cork_bytes = 0;
443 else
444 psock->cork_bytes -= size;
445 if (psock->cork_bytes && !enospc)
446 goto out_err;
447 /* All cork bytes are accounted, rerun the prog. */
448 psock->eval = __SK_NONE;
449 psock->cork_bytes = 0;
450 }
451
452 err = tcp_bpf_send_verdict(sk, psock, msg, &copied, flags);
453out_err:
454 release_sock(sk);
455 sk_psock_put(sk, psock);
456 return copied ? copied : err;
457}
458
459enum {
460 TCP_BPF_IPV4,
461 TCP_BPF_IPV6,
462 TCP_BPF_NUM_PROTS,
463};
464
465enum {
466 TCP_BPF_BASE,
467 TCP_BPF_TX,
468 TCP_BPF_NUM_CFGS,
469};
470
471static struct proto *tcpv6_prot_saved __read_mostly;
472static DEFINE_SPINLOCK(tcpv6_prot_lock);
473static struct proto tcp_bpf_prots[TCP_BPF_NUM_PROTS][TCP_BPF_NUM_CFGS];
474
475static void tcp_bpf_rebuild_protos(struct proto prot[TCP_BPF_NUM_CFGS],
476 struct proto *base)
477{
478 prot[TCP_BPF_BASE] = *base;
479 prot[TCP_BPF_BASE].unhash = sock_map_unhash;
480 prot[TCP_BPF_BASE].close = sock_map_close;
481 prot[TCP_BPF_BASE].recvmsg = tcp_bpf_recvmsg;
482 prot[TCP_BPF_BASE].stream_memory_read = tcp_bpf_stream_read;
483
484 prot[TCP_BPF_TX] = prot[TCP_BPF_BASE];
485 prot[TCP_BPF_TX].sendmsg = tcp_bpf_sendmsg;
486 prot[TCP_BPF_TX].sendpage = tcp_bpf_sendpage;
487}
488
489static void tcp_bpf_check_v6_needs_rebuild(struct proto *ops)
490{
491 if (unlikely(ops != smp_load_acquire(&tcpv6_prot_saved))) {
492 spin_lock_bh(&tcpv6_prot_lock);
493 if (likely(ops != tcpv6_prot_saved)) {
494 tcp_bpf_rebuild_protos(tcp_bpf_prots[TCP_BPF_IPV6], ops);
495 smp_store_release(&tcpv6_prot_saved, ops);
496 }
497 spin_unlock_bh(&tcpv6_prot_lock);
498 }
499}
500
501static int __init tcp_bpf_v4_build_proto(void)
502{
503 tcp_bpf_rebuild_protos(tcp_bpf_prots[TCP_BPF_IPV4], &tcp_prot);
504 return 0;
505}
506late_initcall(tcp_bpf_v4_build_proto);
507
508static int tcp_bpf_assert_proto_ops(struct proto *ops)
509{
510 /* In order to avoid retpoline, we make assumptions when we call
511 * into ops if e.g. a psock is not present. Make sure they are
512 * indeed valid assumptions.
513 */
514 return ops->recvmsg == tcp_recvmsg &&
515 ops->sendmsg == tcp_sendmsg &&
516 ops->sendpage == tcp_sendpage ? 0 : -ENOTSUPP;
517}
518
519int tcp_bpf_update_proto(struct sock *sk, struct sk_psock *psock, bool restore)
520{
521 int family = sk->sk_family == AF_INET6 ? TCP_BPF_IPV6 : TCP_BPF_IPV4;
522 int config = psock->progs.msg_parser ? TCP_BPF_TX : TCP_BPF_BASE;
523
524 if (restore) {
525 if (inet_csk_has_ulp(sk)) {
526 /* TLS does not have an unhash proto in SW cases,
527 * but we need to ensure we stop using the sock_map
528 * unhash routine because the associated psock is being
529 * removed. So use the original unhash handler.
530 */
531 WRITE_ONCE(sk->sk_prot->unhash, psock->saved_unhash);
532 tcp_update_ulp(sk, psock->sk_proto, psock->saved_write_space);
533 } else {
534 sk->sk_write_space = psock->saved_write_space;
535 /* Pairs with lockless read in sk_clone_lock() */
536 WRITE_ONCE(sk->sk_prot, psock->sk_proto);
537 }
538 return 0;
539 }
540
541 if (inet_csk_has_ulp(sk))
542 return -EINVAL;
543
544 if (sk->sk_family == AF_INET6) {
545 if (tcp_bpf_assert_proto_ops(psock->sk_proto))
546 return -EINVAL;
547
548 tcp_bpf_check_v6_needs_rebuild(psock->sk_proto);
549 }
550
551 /* Pairs with lockless read in sk_clone_lock() */
552 WRITE_ONCE(sk->sk_prot, &tcp_bpf_prots[family][config]);
553 return 0;
554}
555EXPORT_SYMBOL_GPL(tcp_bpf_update_proto);
556
557/* If a child got cloned from a listening socket that had tcp_bpf
558 * protocol callbacks installed, we need to restore the callbacks to
559 * the default ones because the child does not inherit the psock state
560 * that tcp_bpf callbacks expect.
561 */
562void tcp_bpf_clone(const struct sock *sk, struct sock *newsk)
563{
564 int family = sk->sk_family == AF_INET6 ? TCP_BPF_IPV6 : TCP_BPF_IPV4;
565 struct proto *prot = newsk->sk_prot;
566
567 if (prot == &tcp_bpf_prots[family][TCP_BPF_BASE])
568 newsk->sk_prot = sk->sk_prot_creator;
569}
570#endif /* CONFIG_BPF_SYSCALL */