forked from ~ljy/RK356X_SDK_RELEASE
blame | history | patch | commit | commitdiff
hc
2024-02-20 ea08eeccae9297f7aabd2ef7f0c2517ac4549acc 
 kernel/net/ipv4/tcp_minisocks.c
....@@ -1,3 +1,4 @@
1
+// SPDX-License-Identifier: GPL-2.0-only
12 /*
23  * INET		An implementation of the TCP/IP protocol suite for the LINUX
34  *		operating system.  INET is implemented using the  BSD Socket
....@@ -179,7 +180,7 @@
179180 			 * Oh well... nobody has a sufficient solution to this
180181 			 * protocol bug yet.
181182 			 */
182
-			if (twsk_net(tw)->ipv4.sysctl_tcp_rfc1337 == 0) {
183
+			if (!READ_ONCE(twsk_net(tw)->ipv4.sysctl_tcp_rfc1337)) {
183184 kill:
184185 				inet_twsk_deschedule_put(tw);
185186 				return TCP_TW_SUCCESS;
....@@ -265,6 +266,7 @@
265266 
266267 		tw->tw_transparent	= inet->transparent;
267268 		tw->tw_mark		= sk->sk_mark;
269
+		tw->tw_priority		= sk->sk_priority;
268270 		tw->tw_rcv_wscale	= tp->rx_opt.rcv_wscale;
269271 		tcptw->tw_rcv_nxt	= tp->rcv_nxt;
270272 		tcptw->tw_snd_nxt	= tp->snd_nxt;
....@@ -273,7 +275,7 @@
273275 		tcptw->tw_ts_recent_stamp = tp->rx_opt.ts_recent_stamp;
274276 		tcptw->tw_ts_offset	= tp->tsoffset;
275277 		tcptw->tw_last_oow_ack_time = 0;
276
-
278
+		tcptw->tw_tx_delay	= tp->tcp_tx_delay;
277279 #if IS_ENABLED(CONFIG_IPV6)
278280 		if (tw->tw_family == PF_INET6) {
279281 			struct ipv6_pinfo *np = inet6_sk(sk);
....@@ -282,6 +284,7 @@
282284 			tw->tw_v6_rcv_saddr = sk->sk_v6_rcv_saddr;
283285 			tw->tw_tclass = np->tclass;
284286 			tw->tw_flowlabel = be32_to_cpu(np->flow_label & IPV6_FLOWLABEL_MASK);
287
+			tw->tw_txhash = sk->sk_txhash;
285288 			tw->tw_ipv6only = sk->sk_ipv6only;
286289 		}
287290 #endif
....@@ -294,12 +297,15 @@
294297 		 * so the timewait ack generating code has the key.
295298 		 */
296299 		do {
297
-			struct tcp_md5sig_key *key;
298300 			tcptw->tw_md5_key = NULL;
299
-			key = tp->af_specific->md5_lookup(sk, sk);
300
-			if (key) {
301
-				tcptw->tw_md5_key = kmemdup(key, sizeof(*key), GFP_ATOMIC);
302
-				BUG_ON(tcptw->tw_md5_key && !tcp_alloc_md5sig_pool());
301
+			if (static_branch_unlikely(&tcp_md5_needed)) {
302
+				struct tcp_md5sig_key *key;
303
+
304
+				key = tp->af_specific->md5_lookup(sk, sk);
305
+				if (key) {
306
+					tcptw->tw_md5_key = kmemdup(key, sizeof(*key), GFP_ATOMIC);
307
+					BUG_ON(tcptw->tw_md5_key && !tcp_alloc_md5sig_pool());
308
+				}
303309 			}
304310 		} while (0);
305311 #endif
....@@ -338,10 +344,12 @@
338344 void tcp_twsk_destructor(struct sock *sk)
339345 {
340346 #ifdef CONFIG_TCP_MD5SIG
341
-	struct tcp_timewait_sock *twsk = tcp_twsk(sk);
347
+	if (static_branch_unlikely(&tcp_md5_needed)) {
348
+		struct tcp_timewait_sock *twsk = tcp_twsk(sk);
342349 
343
-	if (twsk->tw_md5_key)
344
-		kfree_rcu(twsk->tw_md5_key, rcu);
350
+		if (twsk->tw_md5_key)
351
+			kfree_rcu(twsk->tw_md5_key, rcu);
352
+	}
345353 #endif
346354 }
347355 EXPORT_SYMBOL_GPL(tcp_twsk_destructor);
....@@ -406,7 +414,7 @@
406414 
407415 		rcu_read_lock();
408416 		ca = tcp_ca_find_key(ca_key);
409
-		if (likely(ca && try_module_get(ca->owner))) {
417
+		if (likely(ca && bpf_try_module_get(ca, ca->owner))) {
410418 			icsk->icsk_ca_dst_locked = tcp_ca_dst_locked(dst);
411419 			icsk->icsk_ca_ops = ca;
412420 			ca_got_dst = true;
....@@ -417,7 +425,7 @@
417425 	/* If no valid choice made yet, assign current system default ca. */
418426 	if (!ca_got_dst &&
419427 	    (!icsk->icsk_ca_setsockopt ||
420
-	     !try_module_get(icsk->icsk_ca_ops->owner)))
428
+	     !bpf_try_module_get(icsk->icsk_ca_ops, icsk->icsk_ca_ops->owner)))
421429 		tcp_assign_congestion_control(sk);
422430 
423431 	tcp_set_ca_state(sk, TCP_CA_Open);
....@@ -474,50 +482,25 @@
474482 	WRITE_ONCE(newtp->rcv_nxt, seq);
475483 	newtp->segs_in = 1;
476484 
477
-	newtp->snd_sml = newtp->snd_una =
478
-	newtp->snd_nxt = newtp->snd_up = treq->snt_isn + 1;
485
+	seq = treq->snt_isn + 1;
486
+	newtp->snd_sml = newtp->snd_una = seq;
487
+	WRITE_ONCE(newtp->snd_nxt, seq);
488
+	newtp->snd_up = seq;
479489 
480490 	INIT_LIST_HEAD(&newtp->tsq_node);
481491 	INIT_LIST_HEAD(&newtp->tsorted_sent_queue);
482492 
483493 	tcp_init_wl(newtp, treq->rcv_isn);
484494 
485
-	newtp->srtt_us = 0;
486
-	newtp->mdev_us = jiffies_to_usecs(TCP_TIMEOUT_INIT);
487495 	minmax_reset(&newtp->rtt_min, tcp_jiffies32, ~0U);
488
-	newicsk->icsk_rto = TCP_TIMEOUT_INIT;
489496 	newicsk->icsk_ack.lrcvtime = tcp_jiffies32;
490497 
491
-	newtp->packets_out = 0;
492
-	newtp->retrans_out = 0;
493
-	newtp->sacked_out = 0;
494
-	newtp->snd_ssthresh = TCP_INFINITE_SSTHRESH;
495
-	newtp->tlp_high_seq = 0;
496498 	newtp->lsndtime = tcp_jiffies32;
497499 	newsk->sk_txhash = treq->txhash;
498
-	newtp->last_oow_ack_time = 0;
499500 	newtp->total_retrans = req->num_retrans;
500
-
501
-	/* So many TCP implementations out there (incorrectly) count the
502
-	 * initial SYN frame in their delayed-ACK and congestion control
503
-	 * algorithms that we must have the following bandaid to talk
504
-	 * efficiently to them.  -DaveM
505
-	 */
506
-	newtp->snd_cwnd = TCP_INIT_CWND;
507
-	newtp->snd_cwnd_cnt = 0;
508
-
509
-	/* There's a bubble in the pipe until at least the first ACK. */
510
-	newtp->app_limited = ~0U;
511501 
512502 	tcp_init_xmit_timers(newsk);
513503 	WRITE_ONCE(newtp->write_seq, newtp->pushed_seq = treq->snt_isn + 1);
514
-
515
-	newtp->rx_opt.saw_tstamp = 0;
516
-
517
-	newtp->rx_opt.dsack = 0;
518
-	newtp->rx_opt.num_sacks = 0;
519
-
520
-	newtp->urg_data = 0;
521504 
522505 	if (sock_flag(newsk, SOCK_KEEPOPEN))
523506 		inet_csk_reset_keepalive_timer(newsk,
....@@ -540,17 +523,22 @@
540523 	newtp->max_window = newtp->snd_wnd;
541524 
542525 	if (newtp->rx_opt.tstamp_ok) {
543
-		newtp->rx_opt.ts_recent = req->ts_recent;
526
+		newtp->rx_opt.ts_recent = READ_ONCE(req->ts_recent);
544527 		newtp->rx_opt.ts_recent_stamp = ktime_get_seconds();
545528 		newtp->tcp_header_len = sizeof(struct tcphdr) + TCPOLEN_TSTAMP_ALIGNED;
546529 	} else {
547530 		newtp->rx_opt.ts_recent_stamp = 0;
548531 		newtp->tcp_header_len = sizeof(struct tcphdr);
549532 	}
533
+	if (req->num_timeout) {
534
+		newtp->undo_marker = treq->snt_isn;
535
+		newtp->retrans_stamp = div_u64(treq->snt_synack,
536
+					       USEC_PER_SEC / TCP_TS_HZ);
537
+	}
550538 	newtp->tsoffset = treq->ts_off;
551539 #ifdef CONFIG_TCP_MD5SIG
552540 	newtp->md5sig_info = NULL;	/*XXX*/
553
-	if (newtp->af_specific->md5_lookup(sk, newsk))
541
+	if (treq->af_specific->req_md5_lookup(sk, req_to_sk(req)))
554542 		newtp->tcp_header_len += TCPOLEN_MD5SIG_ALIGNED;
555543 #endif
556544 	if (skb->len >= TCP_MSS_DEFAULT + newtp->tcp_header_len)
....@@ -558,14 +546,9 @@
558546 	newtp->rx_opt.mss_clamp = req->mss;
559547 	tcp_ecn_openreq_child(newtp, req);
560548 	newtp->fastopen_req = NULL;
561
-	newtp->fastopen_rsk = NULL;
562
-	newtp->syn_data_acked = 0;
563
-	newtp->rack.mstamp = 0;
564
-	newtp->rack.advanced = 0;
565
-	newtp->rack.reo_wnd_steps = 1;
566
-	newtp->rack.last_delivered = 0;
567
-	newtp->rack.reo_wnd_persist = 0;
568
-	newtp->rack.dsack_seen = 0;
549
+	RCU_INIT_POINTER(newtp->fastopen_rsk, NULL);
550
+
551
+	tcp_bpf_clone(sk, newsk);
569552 
570553 	__TCP_INC_STATS(sock_net(sk), TCP_MIB_PASSIVEOPENS);
571554 
....@@ -582,6 +565,9 @@
582565  * validation and inside tcp_v4_reqsk_send_ack(). Can we do better?
583566  *
584567  * We don't need to initialize tmp_opt.sack_ok as we don't use the results
568
+ *
569
+ * Note: If @fastopen is true, this can be called from process context.
570
+ *       Otherwise, this is from BH context.
585571  */
586572 
587573 struct sock *tcp_check_req(struct sock *sk, struct sk_buff *skb,
....@@ -600,7 +586,7 @@
600586 		tcp_parse_options(sock_net(sk), skb, &tmp_opt, 0, NULL);
601587 
602588 		if (tmp_opt.saw_tstamp) {
603
-			tmp_opt.ts_recent = req->ts_recent;
589
+			tmp_opt.ts_recent = READ_ONCE(req->ts_recent);
604590 			if (tmp_opt.rcv_tsecr)
605591 				tmp_opt.rcv_tsecr -= tcp_rsk(req)->ts_off;
606592 			/* We do not store true stamp, but it is not required,
....@@ -734,14 +720,17 @@
734720 					  &tcp_rsk(req)->last_oow_ack_time))
735721 			req->rsk_ops->send_ack(sk, skb, req);
736722 		if (paws_reject)
737
-			__NET_INC_STATS(sock_net(sk), LINUX_MIB_PAWSESTABREJECTED);
723
+			NET_INC_STATS(sock_net(sk), LINUX_MIB_PAWSESTABREJECTED);
738724 		return NULL;
739725 	}
740726 
741727 	/* In sequence, PAWS is OK. */
742728 
729
+	/* TODO: We probably should defer ts_recent change once
730
+	 * we take ownership of @req.
731
+	 */
743732 	if (tmp_opt.saw_tstamp && !after(TCP_SKB_CB(skb)->seq, tcp_rsk(req)->rcv_nxt))
744
-		req->ts_recent = tmp_opt.rcv_tsval;
733
+		WRITE_ONCE(req->ts_recent, tmp_opt.rcv_tsval);
745734 
746735 	if (TCP_SKB_CB(skb)->seq == tcp_rsk(req)->rcv_isn) {
747736 		/* Truncate SYN, it is out of window starting
....@@ -753,7 +742,7 @@
753742 	 *	   "fourth, check the SYN bit"
754743 	 */
755744 	if (flg & (TCP_FLAG_RST|TCP_FLAG_SYN)) {
756
-		__TCP_INC_STATS(sock_net(sk), TCP_MIB_ATTEMPTFAILS);
745
+		TCP_INC_STATS(sock_net(sk), TCP_MIB_ATTEMPTFAILS);
757746 		goto embryonic_reset;
758747 	}
759748 
....@@ -790,6 +779,12 @@
790779 							 req, &own_req);
791780 	if (!child)
792781 		goto listen_overflow;
782
+
783
+	if (own_req && rsk_drop_req(req)) {
784
+		reqsk_queue_removed(&inet_csk(sk)->icsk_accept_queue, req);
785
+		inet_csk_reqsk_queue_drop_and_put(sk, req);
786
+		return child;
787
+	}
793788 
794789 	sock_rps_save_rxhash(child, skb);
795790 	tcp_synack_rtt_meas(child, req);
....@@ -839,6 +834,7 @@
839834 
840835 int tcp_child_process(struct sock *parent, struct sock *child,
841836 		      struct sk_buff *skb)
837
+	__releases(&((child)->sk_lock.slock))
842838 {
843839 	int ret = 0;
844840 	int state = child->sk_state;