xref: /linux/net/core/skmsg.c (revision d6869352cb3c3cf3450637a52349e2e87c1354aa)
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/skbuff.h>
6 #include <linux/scatterlist.h>
7 
8 #include <net/sock.h>
9 #include <net/tcp.h>
10 
11 static bool sk_msg_try_coalesce_ok(struct sk_msg *msg, int elem_first_coalesce)
12 {
13 	if (msg->sg.end > msg->sg.start &&
14 	    elem_first_coalesce < msg->sg.end)
15 		return true;
16 
17 	if (msg->sg.end < msg->sg.start &&
18 	    (elem_first_coalesce > msg->sg.start ||
19 	     elem_first_coalesce < msg->sg.end))
20 		return true;
21 
22 	return false;
23 }
24 
25 int sk_msg_alloc(struct sock *sk, struct sk_msg *msg, int len,
26 		 int elem_first_coalesce)
27 {
28 	struct page_frag *pfrag = sk_page_frag(sk);
29 	int ret = 0;
30 
31 	len -= msg->sg.size;
32 	while (len > 0) {
33 		struct scatterlist *sge;
34 		u32 orig_offset;
35 		int use, i;
36 
37 		if (!sk_page_frag_refill(sk, pfrag))
38 			return -ENOMEM;
39 
40 		orig_offset = pfrag->offset;
41 		use = min_t(int, len, pfrag->size - orig_offset);
42 		if (!sk_wmem_schedule(sk, use))
43 			return -ENOMEM;
44 
45 		i = msg->sg.end;
46 		sk_msg_iter_var_prev(i);
47 		sge = &msg->sg.data[i];
48 
49 		if (sk_msg_try_coalesce_ok(msg, elem_first_coalesce) &&
50 		    sg_page(sge) == pfrag->page &&
51 		    sge->offset + sge->length == orig_offset) {
52 			sge->length += use;
53 		} else {
54 			if (sk_msg_full(msg)) {
55 				ret = -ENOSPC;
56 				break;
57 			}
58 
59 			sge = &msg->sg.data[msg->sg.end];
60 			sg_unmark_end(sge);
61 			sg_set_page(sge, pfrag->page, use, orig_offset);
62 			get_page(pfrag->page);
63 			sk_msg_iter_next(msg, end);
64 		}
65 
66 		sk_mem_charge(sk, use);
67 		msg->sg.size += use;
68 		pfrag->offset += use;
69 		len -= use;
70 	}
71 
72 	return ret;
73 }
74 EXPORT_SYMBOL_GPL(sk_msg_alloc);
75 
76 int sk_msg_clone(struct sock *sk, struct sk_msg *dst, struct sk_msg *src,
77 		 u32 off, u32 len)
78 {
79 	int i = src->sg.start;
80 	struct scatterlist *sge = sk_msg_elem(src, i);
81 	struct scatterlist *sgd = NULL;
82 	u32 sge_len, sge_off;
83 
84 	while (off) {
85 		if (sge->length > off)
86 			break;
87 		off -= sge->length;
88 		sk_msg_iter_var_next(i);
89 		if (i == src->sg.end && off)
90 			return -ENOSPC;
91 		sge = sk_msg_elem(src, i);
92 	}
93 
94 	while (len) {
95 		sge_len = sge->length - off;
96 		if (sge_len > len)
97 			sge_len = len;
98 
99 		if (dst->sg.end)
100 			sgd = sk_msg_elem(dst, dst->sg.end - 1);
101 
102 		if (sgd &&
103 		    (sg_page(sge) == sg_page(sgd)) &&
104 		    (sg_virt(sge) + off == sg_virt(sgd) + sgd->length)) {
105 			sgd->length += sge_len;
106 			dst->sg.size += sge_len;
107 		} else if (!sk_msg_full(dst)) {
108 			sge_off = sge->offset + off;
109 			sk_msg_page_add(dst, sg_page(sge), sge_len, sge_off);
110 		} else {
111 			return -ENOSPC;
112 		}
113 
114 		off = 0;
115 		len -= sge_len;
116 		sk_mem_charge(sk, sge_len);
117 		sk_msg_iter_var_next(i);
118 		if (i == src->sg.end && len)
119 			return -ENOSPC;
120 		sge = sk_msg_elem(src, i);
121 	}
122 
123 	return 0;
124 }
125 EXPORT_SYMBOL_GPL(sk_msg_clone);
126 
127 void sk_msg_return_zero(struct sock *sk, struct sk_msg *msg, int bytes)
128 {
129 	int i = msg->sg.start;
130 
131 	do {
132 		struct scatterlist *sge = sk_msg_elem(msg, i);
133 
134 		if (bytes < sge->length) {
135 			sge->length -= bytes;
136 			sge->offset += bytes;
137 			sk_mem_uncharge(sk, bytes);
138 			break;
139 		}
140 
141 		sk_mem_uncharge(sk, sge->length);
142 		bytes -= sge->length;
143 		sge->length = 0;
144 		sge->offset = 0;
145 		sk_msg_iter_var_next(i);
146 	} while (bytes && i != msg->sg.end);
147 	msg->sg.start = i;
148 }
149 EXPORT_SYMBOL_GPL(sk_msg_return_zero);
150 
151 void sk_msg_return(struct sock *sk, struct sk_msg *msg, int bytes)
152 {
153 	int i = msg->sg.start;
154 
155 	do {
156 		struct scatterlist *sge = &msg->sg.data[i];
157 		int uncharge = (bytes < sge->length) ? bytes : sge->length;
158 
159 		sk_mem_uncharge(sk, uncharge);
160 		bytes -= uncharge;
161 		sk_msg_iter_var_next(i);
162 	} while (i != msg->sg.end);
163 }
164 EXPORT_SYMBOL_GPL(sk_msg_return);
165 
166 static int sk_msg_free_elem(struct sock *sk, struct sk_msg *msg, u32 i,
167 			    bool charge)
168 {
169 	struct scatterlist *sge = sk_msg_elem(msg, i);
170 	u32 len = sge->length;
171 
172 	if (charge)
173 		sk_mem_uncharge(sk, len);
174 	if (!msg->skb)
175 		put_page(sg_page(sge));
176 	memset(sge, 0, sizeof(*sge));
177 	return len;
178 }
179 
180 static int __sk_msg_free(struct sock *sk, struct sk_msg *msg, u32 i,
181 			 bool charge)
182 {
183 	struct scatterlist *sge = sk_msg_elem(msg, i);
184 	int freed = 0;
185 
186 	while (msg->sg.size) {
187 		msg->sg.size -= sge->length;
188 		freed += sk_msg_free_elem(sk, msg, i, charge);
189 		sk_msg_iter_var_next(i);
190 		sk_msg_check_to_free(msg, i, msg->sg.size);
191 		sge = sk_msg_elem(msg, i);
192 	}
193 	if (msg->skb)
194 		consume_skb(msg->skb);
195 	sk_msg_init(msg);
196 	return freed;
197 }
198 
199 int sk_msg_free_nocharge(struct sock *sk, struct sk_msg *msg)
200 {
201 	return __sk_msg_free(sk, msg, msg->sg.start, false);
202 }
203 EXPORT_SYMBOL_GPL(sk_msg_free_nocharge);
204 
205 int sk_msg_free(struct sock *sk, struct sk_msg *msg)
206 {
207 	return __sk_msg_free(sk, msg, msg->sg.start, true);
208 }
209 EXPORT_SYMBOL_GPL(sk_msg_free);
210 
211 static void __sk_msg_free_partial(struct sock *sk, struct sk_msg *msg,
212 				  u32 bytes, bool charge)
213 {
214 	struct scatterlist *sge;
215 	u32 i = msg->sg.start;
216 
217 	while (bytes) {
218 		sge = sk_msg_elem(msg, i);
219 		if (!sge->length)
220 			break;
221 		if (bytes < sge->length) {
222 			if (charge)
223 				sk_mem_uncharge(sk, bytes);
224 			sge->length -= bytes;
225 			sge->offset += bytes;
226 			msg->sg.size -= bytes;
227 			break;
228 		}
229 
230 		msg->sg.size -= sge->length;
231 		bytes -= sge->length;
232 		sk_msg_free_elem(sk, msg, i, charge);
233 		sk_msg_iter_var_next(i);
234 		sk_msg_check_to_free(msg, i, bytes);
235 	}
236 	msg->sg.start = i;
237 }
238 
239 void sk_msg_free_partial(struct sock *sk, struct sk_msg *msg, u32 bytes)
240 {
241 	__sk_msg_free_partial(sk, msg, bytes, true);
242 }
243 EXPORT_SYMBOL_GPL(sk_msg_free_partial);
244 
245 void sk_msg_free_partial_nocharge(struct sock *sk, struct sk_msg *msg,
246 				  u32 bytes)
247 {
248 	__sk_msg_free_partial(sk, msg, bytes, false);
249 }
250 
251 void sk_msg_trim(struct sock *sk, struct sk_msg *msg, int len)
252 {
253 	int trim = msg->sg.size - len;
254 	u32 i = msg->sg.end;
255 
256 	if (trim <= 0) {
257 		WARN_ON(trim < 0);
258 		return;
259 	}
260 
261 	sk_msg_iter_var_prev(i);
262 	msg->sg.size = len;
263 	while (msg->sg.data[i].length &&
264 	       trim >= msg->sg.data[i].length) {
265 		trim -= msg->sg.data[i].length;
266 		sk_msg_free_elem(sk, msg, i, true);
267 		sk_msg_iter_var_prev(i);
268 		if (!trim)
269 			goto out;
270 	}
271 
272 	msg->sg.data[i].length -= trim;
273 	sk_mem_uncharge(sk, trim);
274 out:
275 	/* If we trim data before curr pointer update copybreak and current
276 	 * so that any future copy operations start at new copy location.
277 	 * However trimed data that has not yet been used in a copy op
278 	 * does not require an update.
279 	 */
280 	if (msg->sg.curr >= i) {
281 		msg->sg.curr = i;
282 		msg->sg.copybreak = msg->sg.data[i].length;
283 	}
284 	sk_msg_iter_var_next(i);
285 	msg->sg.end = i;
286 }
287 EXPORT_SYMBOL_GPL(sk_msg_trim);
288 
289 int sk_msg_zerocopy_from_iter(struct sock *sk, struct iov_iter *from,
290 			      struct sk_msg *msg, u32 bytes)
291 {
292 	int i, maxpages, ret = 0, num_elems = sk_msg_elem_used(msg);
293 	const int to_max_pages = MAX_MSG_FRAGS;
294 	struct page *pages[MAX_MSG_FRAGS];
295 	ssize_t orig, copied, use, offset;
296 
297 	orig = msg->sg.size;
298 	while (bytes > 0) {
299 		i = 0;
300 		maxpages = to_max_pages - num_elems;
301 		if (maxpages == 0) {
302 			ret = -EFAULT;
303 			goto out;
304 		}
305 
306 		copied = iov_iter_get_pages(from, pages, bytes, maxpages,
307 					    &offset);
308 		if (copied <= 0) {
309 			ret = -EFAULT;
310 			goto out;
311 		}
312 
313 		iov_iter_advance(from, copied);
314 		bytes -= copied;
315 		msg->sg.size += copied;
316 
317 		while (copied) {
318 			use = min_t(int, copied, PAGE_SIZE - offset);
319 			sg_set_page(&msg->sg.data[msg->sg.end],
320 				    pages[i], use, offset);
321 			sg_unmark_end(&msg->sg.data[msg->sg.end]);
322 			sk_mem_charge(sk, use);
323 
324 			offset = 0;
325 			copied -= use;
326 			sk_msg_iter_next(msg, end);
327 			num_elems++;
328 			i++;
329 		}
330 		/* When zerocopy is mixed with sk_msg_*copy* operations we
331 		 * may have a copybreak set in this case clear and prefer
332 		 * zerocopy remainder when possible.
333 		 */
334 		msg->sg.copybreak = 0;
335 		msg->sg.curr = msg->sg.end;
336 	}
337 out:
338 	/* Revert iov_iter updates, msg will need to use 'trim' later if it
339 	 * also needs to be cleared.
340 	 */
341 	if (ret)
342 		iov_iter_revert(from, msg->sg.size - orig);
343 	return ret;
344 }
345 EXPORT_SYMBOL_GPL(sk_msg_zerocopy_from_iter);
346 
347 int sk_msg_memcopy_from_iter(struct sock *sk, struct iov_iter *from,
348 			     struct sk_msg *msg, u32 bytes)
349 {
350 	int ret = -ENOSPC, i = msg->sg.curr;
351 	struct scatterlist *sge;
352 	u32 copy, buf_size;
353 	void *to;
354 
355 	do {
356 		sge = sk_msg_elem(msg, i);
357 		/* This is possible if a trim operation shrunk the buffer */
358 		if (msg->sg.copybreak >= sge->length) {
359 			msg->sg.copybreak = 0;
360 			sk_msg_iter_var_next(i);
361 			if (i == msg->sg.end)
362 				break;
363 			sge = sk_msg_elem(msg, i);
364 		}
365 
366 		buf_size = sge->length - msg->sg.copybreak;
367 		copy = (buf_size > bytes) ? bytes : buf_size;
368 		to = sg_virt(sge) + msg->sg.copybreak;
369 		msg->sg.copybreak += copy;
370 		if (sk->sk_route_caps & NETIF_F_NOCACHE_COPY)
371 			ret = copy_from_iter_nocache(to, copy, from);
372 		else
373 			ret = copy_from_iter(to, copy, from);
374 		if (ret != copy) {
375 			ret = -EFAULT;
376 			goto out;
377 		}
378 		bytes -= copy;
379 		if (!bytes)
380 			break;
381 		msg->sg.copybreak = 0;
382 		sk_msg_iter_var_next(i);
383 	} while (i != msg->sg.end);
384 out:
385 	msg->sg.curr = i;
386 	return ret;
387 }
388 EXPORT_SYMBOL_GPL(sk_msg_memcopy_from_iter);
389 
390 static int sk_psock_skb_ingress(struct sk_psock *psock, struct sk_buff *skb)
391 {
392 	struct sock *sk = psock->sk;
393 	int copied = 0, num_sge;
394 	struct sk_msg *msg;
395 
396 	msg = kzalloc(sizeof(*msg), __GFP_NOWARN | GFP_ATOMIC);
397 	if (unlikely(!msg))
398 		return -EAGAIN;
399 	if (!sk_rmem_schedule(sk, skb, skb->len)) {
400 		kfree(msg);
401 		return -EAGAIN;
402 	}
403 
404 	sk_msg_init(msg);
405 	num_sge = skb_to_sgvec(skb, msg->sg.data, 0, skb->len);
406 	if (unlikely(num_sge < 0)) {
407 		kfree(msg);
408 		return num_sge;
409 	}
410 
411 	sk_mem_charge(sk, skb->len);
412 	copied = skb->len;
413 	msg->sg.start = 0;
414 	msg->sg.size = copied;
415 	msg->sg.end = num_sge == MAX_MSG_FRAGS ? 0 : num_sge;
416 	msg->skb = skb;
417 
418 	sk_psock_queue_msg(psock, msg);
419 	sk_psock_data_ready(sk, psock);
420 	return copied;
421 }
422 
423 static int sk_psock_handle_skb(struct sk_psock *psock, struct sk_buff *skb,
424 			       u32 off, u32 len, bool ingress)
425 {
426 	if (ingress)
427 		return sk_psock_skb_ingress(psock, skb);
428 	else
429 		return skb_send_sock_locked(psock->sk, skb, off, len);
430 }
431 
432 static void sk_psock_backlog(struct work_struct *work)
433 {
434 	struct sk_psock *psock = container_of(work, struct sk_psock, work);
435 	struct sk_psock_work_state *state = &psock->work_state;
436 	struct sk_buff *skb;
437 	bool ingress;
438 	u32 len, off;
439 	int ret;
440 
441 	/* Lock sock to avoid losing sk_socket during loop. */
442 	lock_sock(psock->sk);
443 	if (state->skb) {
444 		skb = state->skb;
445 		len = state->len;
446 		off = state->off;
447 		state->skb = NULL;
448 		goto start;
449 	}
450 
451 	while ((skb = skb_dequeue(&psock->ingress_skb))) {
452 		len = skb->len;
453 		off = 0;
454 start:
455 		ingress = tcp_skb_bpf_ingress(skb);
456 		do {
457 			ret = -EIO;
458 			if (likely(psock->sk->sk_socket))
459 				ret = sk_psock_handle_skb(psock, skb, off,
460 							  len, ingress);
461 			if (ret <= 0) {
462 				if (ret == -EAGAIN) {
463 					state->skb = skb;
464 					state->len = len;
465 					state->off = off;
466 					goto end;
467 				}
468 				/* Hard errors break pipe and stop xmit. */
469 				sk_psock_report_error(psock, ret ? -ret : EPIPE);
470 				sk_psock_clear_state(psock, SK_PSOCK_TX_ENABLED);
471 				kfree_skb(skb);
472 				goto end;
473 			}
474 			off += ret;
475 			len -= ret;
476 		} while (len);
477 
478 		if (!ingress)
479 			kfree_skb(skb);
480 	}
481 end:
482 	release_sock(psock->sk);
483 }
484 
485 struct sk_psock *sk_psock_init(struct sock *sk, int node)
486 {
487 	struct sk_psock *psock = kzalloc_node(sizeof(*psock),
488 					      GFP_ATOMIC | __GFP_NOWARN,
489 					      node);
490 	if (!psock)
491 		return NULL;
492 
493 	psock->sk = sk;
494 	psock->eval =  __SK_NONE;
495 
496 	INIT_LIST_HEAD(&psock->link);
497 	spin_lock_init(&psock->link_lock);
498 
499 	INIT_WORK(&psock->work, sk_psock_backlog);
500 	INIT_LIST_HEAD(&psock->ingress_msg);
501 	skb_queue_head_init(&psock->ingress_skb);
502 
503 	sk_psock_set_state(psock, SK_PSOCK_TX_ENABLED);
504 	refcount_set(&psock->refcnt, 1);
505 
506 	rcu_assign_sk_user_data(sk, psock);
507 	sock_hold(sk);
508 
509 	return psock;
510 }
511 EXPORT_SYMBOL_GPL(sk_psock_init);
512 
513 struct sk_psock_link *sk_psock_link_pop(struct sk_psock *psock)
514 {
515 	struct sk_psock_link *link;
516 
517 	spin_lock_bh(&psock->link_lock);
518 	link = list_first_entry_or_null(&psock->link, struct sk_psock_link,
519 					list);
520 	if (link)
521 		list_del(&link->list);
522 	spin_unlock_bh(&psock->link_lock);
523 	return link;
524 }
525 
526 void __sk_psock_purge_ingress_msg(struct sk_psock *psock)
527 {
528 	struct sk_msg *msg, *tmp;
529 
530 	list_for_each_entry_safe(msg, tmp, &psock->ingress_msg, list) {
531 		list_del(&msg->list);
532 		sk_msg_free(psock->sk, msg);
533 		kfree(msg);
534 	}
535 }
536 
537 static void sk_psock_zap_ingress(struct sk_psock *psock)
538 {
539 	__skb_queue_purge(&psock->ingress_skb);
540 	__sk_psock_purge_ingress_msg(psock);
541 }
542 
543 static void sk_psock_link_destroy(struct sk_psock *psock)
544 {
545 	struct sk_psock_link *link, *tmp;
546 
547 	list_for_each_entry_safe(link, tmp, &psock->link, list) {
548 		list_del(&link->list);
549 		sk_psock_free_link(link);
550 	}
551 }
552 
553 static void sk_psock_destroy_deferred(struct work_struct *gc)
554 {
555 	struct sk_psock *psock = container_of(gc, struct sk_psock, gc);
556 
557 	/* No sk_callback_lock since already detached. */
558 
559 	/* Parser has been stopped */
560 	if (psock->progs.skb_parser)
561 		strp_done(&psock->parser.strp);
562 
563 	cancel_work_sync(&psock->work);
564 
565 	psock_progs_drop(&psock->progs);
566 
567 	sk_psock_link_destroy(psock);
568 	sk_psock_cork_free(psock);
569 	sk_psock_zap_ingress(psock);
570 
571 	if (psock->sk_redir)
572 		sock_put(psock->sk_redir);
573 	sock_put(psock->sk);
574 	kfree(psock);
575 }
576 
577 void sk_psock_destroy(struct rcu_head *rcu)
578 {
579 	struct sk_psock *psock = container_of(rcu, struct sk_psock, rcu);
580 
581 	INIT_WORK(&psock->gc, sk_psock_destroy_deferred);
582 	schedule_work(&psock->gc);
583 }
584 EXPORT_SYMBOL_GPL(sk_psock_destroy);
585 
586 void sk_psock_drop(struct sock *sk, struct sk_psock *psock)
587 {
588 	rcu_assign_sk_user_data(sk, NULL);
589 	sk_psock_cork_free(psock);
590 	sk_psock_zap_ingress(psock);
591 	sk_psock_restore_proto(sk, psock);
592 
593 	write_lock_bh(&sk->sk_callback_lock);
594 	if (psock->progs.skb_parser)
595 		sk_psock_stop_strp(sk, psock);
596 	write_unlock_bh(&sk->sk_callback_lock);
597 	sk_psock_clear_state(psock, SK_PSOCK_TX_ENABLED);
598 
599 	call_rcu(&psock->rcu, sk_psock_destroy);
600 }
601 EXPORT_SYMBOL_GPL(sk_psock_drop);
602 
603 static int sk_psock_map_verd(int verdict, bool redir)
604 {
605 	switch (verdict) {
606 	case SK_PASS:
607 		return redir ? __SK_REDIRECT : __SK_PASS;
608 	case SK_DROP:
609 	default:
610 		break;
611 	}
612 
613 	return __SK_DROP;
614 }
615 
616 int sk_psock_msg_verdict(struct sock *sk, struct sk_psock *psock,
617 			 struct sk_msg *msg)
618 {
619 	struct bpf_prog *prog;
620 	int ret;
621 
622 	preempt_disable();
623 	rcu_read_lock();
624 	prog = READ_ONCE(psock->progs.msg_parser);
625 	if (unlikely(!prog)) {
626 		ret = __SK_PASS;
627 		goto out;
628 	}
629 
630 	sk_msg_compute_data_pointers(msg);
631 	msg->sk = sk;
632 	ret = BPF_PROG_RUN(prog, msg);
633 	ret = sk_psock_map_verd(ret, msg->sk_redir);
634 	psock->apply_bytes = msg->apply_bytes;
635 	if (ret == __SK_REDIRECT) {
636 		if (psock->sk_redir)
637 			sock_put(psock->sk_redir);
638 		psock->sk_redir = msg->sk_redir;
639 		if (!psock->sk_redir) {
640 			ret = __SK_DROP;
641 			goto out;
642 		}
643 		sock_hold(psock->sk_redir);
644 	}
645 out:
646 	rcu_read_unlock();
647 	preempt_enable();
648 	return ret;
649 }
650 EXPORT_SYMBOL_GPL(sk_psock_msg_verdict);
651 
652 static int sk_psock_bpf_run(struct sk_psock *psock, struct bpf_prog *prog,
653 			    struct sk_buff *skb)
654 {
655 	int ret;
656 
657 	skb->sk = psock->sk;
658 	bpf_compute_data_end_sk_skb(skb);
659 	preempt_disable();
660 	ret = BPF_PROG_RUN(prog, skb);
661 	preempt_enable();
662 	/* strparser clones the skb before handing it to a upper layer,
663 	 * meaning skb_orphan has been called. We NULL sk on the way out
664 	 * to ensure we don't trigger a BUG_ON() in skb/sk operations
665 	 * later and because we are not charging the memory of this skb
666 	 * to any socket yet.
667 	 */
668 	skb->sk = NULL;
669 	return ret;
670 }
671 
672 static struct sk_psock *sk_psock_from_strp(struct strparser *strp)
673 {
674 	struct sk_psock_parser *parser;
675 
676 	parser = container_of(strp, struct sk_psock_parser, strp);
677 	return container_of(parser, struct sk_psock, parser);
678 }
679 
680 static void sk_psock_verdict_apply(struct sk_psock *psock,
681 				   struct sk_buff *skb, int verdict)
682 {
683 	struct sk_psock *psock_other;
684 	struct sock *sk_other;
685 	bool ingress;
686 
687 	switch (verdict) {
688 	case __SK_PASS:
689 		sk_other = psock->sk;
690 		if (sock_flag(sk_other, SOCK_DEAD) ||
691 		    !sk_psock_test_state(psock, SK_PSOCK_TX_ENABLED)) {
692 			goto out_free;
693 		}
694 		if (atomic_read(&sk_other->sk_rmem_alloc) <=
695 		    sk_other->sk_rcvbuf) {
696 			struct tcp_skb_cb *tcp = TCP_SKB_CB(skb);
697 
698 			tcp->bpf.flags |= BPF_F_INGRESS;
699 			skb_queue_tail(&psock->ingress_skb, skb);
700 			schedule_work(&psock->work);
701 			break;
702 		}
703 		goto out_free;
704 	case __SK_REDIRECT:
705 		sk_other = tcp_skb_bpf_redirect_fetch(skb);
706 		if (unlikely(!sk_other))
707 			goto out_free;
708 		psock_other = sk_psock(sk_other);
709 		if (!psock_other || sock_flag(sk_other, SOCK_DEAD) ||
710 		    !sk_psock_test_state(psock_other, SK_PSOCK_TX_ENABLED))
711 			goto out_free;
712 		ingress = tcp_skb_bpf_ingress(skb);
713 		if ((!ingress && sock_writeable(sk_other)) ||
714 		    (ingress &&
715 		     atomic_read(&sk_other->sk_rmem_alloc) <=
716 		     sk_other->sk_rcvbuf)) {
717 			if (!ingress)
718 				skb_set_owner_w(skb, sk_other);
719 			skb_queue_tail(&psock_other->ingress_skb, skb);
720 			schedule_work(&psock_other->work);
721 			break;
722 		}
723 		/* fall-through */
724 	case __SK_DROP:
725 		/* fall-through */
726 	default:
727 out_free:
728 		kfree_skb(skb);
729 	}
730 }
731 
732 static void sk_psock_strp_read(struct strparser *strp, struct sk_buff *skb)
733 {
734 	struct sk_psock *psock = sk_psock_from_strp(strp);
735 	struct bpf_prog *prog;
736 	int ret = __SK_DROP;
737 
738 	rcu_read_lock();
739 	prog = READ_ONCE(psock->progs.skb_verdict);
740 	if (likely(prog)) {
741 		skb_orphan(skb);
742 		tcp_skb_bpf_redirect_clear(skb);
743 		ret = sk_psock_bpf_run(psock, prog, skb);
744 		ret = sk_psock_map_verd(ret, tcp_skb_bpf_redirect_fetch(skb));
745 	}
746 	rcu_read_unlock();
747 	sk_psock_verdict_apply(psock, skb, ret);
748 }
749 
750 static int sk_psock_strp_read_done(struct strparser *strp, int err)
751 {
752 	return err;
753 }
754 
755 static int sk_psock_strp_parse(struct strparser *strp, struct sk_buff *skb)
756 {
757 	struct sk_psock *psock = sk_psock_from_strp(strp);
758 	struct bpf_prog *prog;
759 	int ret = skb->len;
760 
761 	rcu_read_lock();
762 	prog = READ_ONCE(psock->progs.skb_parser);
763 	if (likely(prog))
764 		ret = sk_psock_bpf_run(psock, prog, skb);
765 	rcu_read_unlock();
766 	return ret;
767 }
768 
769 /* Called with socket lock held. */
770 static void sk_psock_strp_data_ready(struct sock *sk)
771 {
772 	struct sk_psock *psock;
773 
774 	rcu_read_lock();
775 	psock = sk_psock(sk);
776 	if (likely(psock)) {
777 		write_lock_bh(&sk->sk_callback_lock);
778 		strp_data_ready(&psock->parser.strp);
779 		write_unlock_bh(&sk->sk_callback_lock);
780 	}
781 	rcu_read_unlock();
782 }
783 
784 static void sk_psock_write_space(struct sock *sk)
785 {
786 	struct sk_psock *psock;
787 	void (*write_space)(struct sock *sk);
788 
789 	rcu_read_lock();
790 	psock = sk_psock(sk);
791 	if (likely(psock && sk_psock_test_state(psock, SK_PSOCK_TX_ENABLED)))
792 		schedule_work(&psock->work);
793 	write_space = psock->saved_write_space;
794 	rcu_read_unlock();
795 	write_space(sk);
796 }
797 
798 int sk_psock_init_strp(struct sock *sk, struct sk_psock *psock)
799 {
800 	static const struct strp_callbacks cb = {
801 		.rcv_msg	= sk_psock_strp_read,
802 		.read_sock_done	= sk_psock_strp_read_done,
803 		.parse_msg	= sk_psock_strp_parse,
804 	};
805 
806 	psock->parser.enabled = false;
807 	return strp_init(&psock->parser.strp, sk, &cb);
808 }
809 
810 void sk_psock_start_strp(struct sock *sk, struct sk_psock *psock)
811 {
812 	struct sk_psock_parser *parser = &psock->parser;
813 
814 	if (parser->enabled)
815 		return;
816 
817 	parser->saved_data_ready = sk->sk_data_ready;
818 	sk->sk_data_ready = sk_psock_strp_data_ready;
819 	sk->sk_write_space = sk_psock_write_space;
820 	parser->enabled = true;
821 }
822 
823 void sk_psock_stop_strp(struct sock *sk, struct sk_psock *psock)
824 {
825 	struct sk_psock_parser *parser = &psock->parser;
826 
827 	if (!parser->enabled)
828 		return;
829 
830 	sk->sk_data_ready = parser->saved_data_ready;
831 	parser->saved_data_ready = NULL;
832 	strp_stop(&parser->strp);
833 	parser->enabled = false;
834 }
835