1 files changed, 93 insertions, 52 deletions

diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c
index e3167ad96567..95618d0e78e4 100644
--- a/net/ipv4/tcp_output.c
+++ b/net/ipv4/tcp_output.c
@@ -39,11 +39,13 @@
 
 #include <net/tcp.h>
 #include <net/mptcp.h>
+#include <net/proto_memory.h>
 
 #include <linux/compiler.h>
 #include <linux/gfp.h>
 #include <linux/module.h>
 #include <linux/static_key.h>
+#include <linux/skbuff_ref.h>
 
 #include <trace/events/tcp.h>
 
@@ -203,16 +205,17 @@ static inline void tcp_event_ack_sent(struct sock *sk, u32 rcv_nxt)
  * This MUST be enforced by all callers.
  */
 void tcp_select_initial_window(const struct sock *sk, int __space, __u32 mss,
-			       __u32 *rcv_wnd, __u32 *window_clamp,
+			       __u32 *rcv_wnd, __u32 *__window_clamp,
 			       int wscale_ok, __u8 *rcv_wscale,
 			       __u32 init_rcv_wnd)
 {
 	unsigned int space = (__space < 0 ? 0 : __space);
+	u32 window_clamp = READ_ONCE(*__window_clamp);
 
 	/* If no clamp set the clamp to the max possible scaled window */
-	if (*window_clamp == 0)
-		(*window_clamp) = (U16_MAX << TCP_MAX_WSCALE);
-	space = min(*window_clamp, space);
+	if (window_clamp == 0)
+		window_clamp = (U16_MAX << TCP_MAX_WSCALE);
+	space = min(window_clamp, space);
 
 	/* Quantize space offering to a multiple of mss if possible. */
 	if (space > mss)
@@ -229,7 +232,7 @@ void tcp_select_initial_window(const struct sock *sk, int __space, __u32 mss,
 	if (READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_workaround_signed_windows))
 		(*rcv_wnd) = min(space, MAX_TCP_WINDOW);
 	else
-		(*rcv_wnd) = min_t(u32, space, U16_MAX);
+		(*rcv_wnd) = space;
 
 	if (init_rcv_wnd)
 		*rcv_wnd = min(*rcv_wnd, init_rcv_wnd * mss);
@@ -239,12 +242,13 @@ void tcp_select_initial_window(const struct sock *sk, int __space, __u32 mss,
 		/* Set window scaling on max possible window */
 		space = max_t(u32, space, READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_rmem[2]));
 		space = max_t(u32, space, READ_ONCE(sysctl_rmem_max));
-		space = min_t(u32, space, *window_clamp);
+		space = min_t(u32, space, window_clamp);
 		*rcv_wscale = clamp_t(int, ilog2(space) - 15,
 				      0, TCP_MAX_WSCALE);
 	}
 	/* Set the clamp no higher than max representable value */
-	(*window_clamp) = min_t(__u32, U16_MAX << (*rcv_wscale), *window_clamp);
+	WRITE_ONCE(*__window_clamp,
+		   min_t(__u32, U16_MAX << (*rcv_wscale), window_clamp));
 }
 EXPORT_SYMBOL(tcp_select_initial_window);
 
@@ -1499,18 +1503,22 @@ static void tcp_queue_skb(struct sock *sk, struct sk_buff *skb)
 }
 
 /* Initialize TSO segments for a packet. */
-static void tcp_set_skb_tso_segs(struct sk_buff *skb, unsigned int mss_now)
+static int tcp_set_skb_tso_segs(struct sk_buff *skb, unsigned int mss_now)
 {
+	int tso_segs;
+
 	if (skb->len <= mss_now) {
 		/* Avoid the costly divide in the normal
 		 * non-TSO case.
 		 */
-		tcp_skb_pcount_set(skb, 1);
 		TCP_SKB_CB(skb)->tcp_gso_size = 0;
-	} else {
-		tcp_skb_pcount_set(skb, DIV_ROUND_UP(skb->len, mss_now));
-		TCP_SKB_CB(skb)->tcp_gso_size = mss_now;
+		tcp_skb_pcount_set(skb, 1);
+		return 1;
 	}
+	TCP_SKB_CB(skb)->tcp_gso_size = mss_now;
+	tso_segs = DIV_ROUND_UP(skb->len, mss_now);
+	tcp_skb_pcount_set(skb, tso_segs);
+	return tso_segs;
 }
 
 /* Pcount in the middle of the write queue got changed, we need to do various
@@ -2070,16 +2078,10 @@ static unsigned int tcp_mss_split_point(const struct sock *sk,
 /* Can at least one segment of SKB be sent right now, according to the
  * congestion window rules?  If so, return how many segments are allowed.
  */
-static inline unsigned int tcp_cwnd_test(const struct tcp_sock *tp,
-					 const struct sk_buff *skb)
+static u32 tcp_cwnd_test(const struct tcp_sock *tp)
 {
 	u32 in_flight, cwnd, halfcwnd;
 
-	/* Don't be strict about the congestion window for the final FIN.  */
-	if ((TCP_SKB_CB(skb)->tcp_flags & TCPHDR_FIN) &&
-	    tcp_skb_pcount(skb) == 1)
-		return 1;
-
 	in_flight = tcp_packets_in_flight(tp);
 	cwnd = tcp_snd_cwnd(tp);
 	if (in_flight >= cwnd)
@@ -2100,10 +2102,9 @@ static int tcp_init_tso_segs(struct sk_buff *skb, unsigned int mss_now)
 {
 	int tso_segs = tcp_skb_pcount(skb);
 
-	if (!tso_segs || (tso_segs > 1 && tcp_skb_mss(skb) != mss_now)) {
-		tcp_set_skb_tso_segs(skb, mss_now);
-		tso_segs = tcp_skb_pcount(skb);
-	}
+	if (!tso_segs || (tso_segs > 1 && tcp_skb_mss(skb) != mss_now))
+		return tcp_set_skb_tso_segs(skb, mss_now);
+
 	return tso_segs;
 }
 
@@ -2403,6 +2404,21 @@ commit:
 	return 0;
 }
 
+/* tcp_mtu_probe() and tcp_grow_skb() can both eat an skb (src) if
+ * all its payload was moved to another one (dst).
+ * Make sure to transfer tcp_flags, eor, and tstamp.
+ */
+static void tcp_eat_one_skb(struct sock *sk,
+			    struct sk_buff *dst,
+			    struct sk_buff *src)
+{
+	TCP_SKB_CB(dst)->tcp_flags |= TCP_SKB_CB(src)->tcp_flags;
+	TCP_SKB_CB(dst)->eor = TCP_SKB_CB(src)->eor;
+	tcp_skb_collapse_tstamp(dst, src);
+	tcp_unlink_write_queue(src, sk);
+	tcp_wmem_free_skb(sk, src);
+}
+
 /* Create a new MTU probe if we are ready.
  * MTU probe is regularly attempting to increase the path MTU by
  * deliberately sending larger packets.  This discovers routing
@@ -2508,16 +2524,7 @@ static int tcp_mtu_probe(struct sock *sk)
 		copy = min_t(int, skb->len, probe_size - len);
 
 		if (skb->len <= copy) {
-			/* We've eaten all the data from this skb.
-			 * Throw it away. */
-			TCP_SKB_CB(nskb)->tcp_flags |= TCP_SKB_CB(skb)->tcp_flags;
-			/* If this is the last SKB we copy and eor is set
-			 * we need to propagate it to the new skb.
-			 */
-			TCP_SKB_CB(nskb)->eor = TCP_SKB_CB(skb)->eor;
-			tcp_skb_collapse_tstamp(nskb, skb);
-			tcp_unlink_write_queue(skb, sk);
-			tcp_wmem_free_skb(sk, skb);
+			tcp_eat_one_skb(sk, nskb, skb);
 		} else {
 			TCP_SKB_CB(nskb)->tcp_flags |= TCP_SKB_CB(skb)->tcp_flags &
 						   ~(TCPHDR_FIN|TCPHDR_PSH);
@@ -2683,6 +2690,35 @@ void tcp_chrono_stop(struct sock *sk, const enum tcp_chrono type)
 		tcp_chrono_set(tp, TCP_CHRONO_BUSY);
 }
 
+/* First skb in the write queue is smaller than ideal packet size.
+ * Check if we can move payload from the second skb in the queue.
+ */
+static void tcp_grow_skb(struct sock *sk, struct sk_buff *skb, int amount)
+{
+	struct sk_buff *next_skb = skb->next;
+	unsigned int nlen;
+
+	if (tcp_skb_is_last(sk, skb))
+		return;
+
+	if (!tcp_skb_can_collapse(skb, next_skb))
+		return;
+
+	nlen = min_t(u32, amount, next_skb->len);
+	if (!nlen || !skb_shift(skb, next_skb, nlen))
+		return;
+
+	TCP_SKB_CB(skb)->end_seq += nlen;
+	TCP_SKB_CB(next_skb)->seq += nlen;
+
+	if (!next_skb->len) {
+		/* In case FIN is set, we need to update end_seq */
+		TCP_SKB_CB(skb)->end_seq = TCP_SKB_CB(next_skb)->end_seq;
+
+		tcp_eat_one_skb(sk, skb, next_skb);
+	}
+}
+
 /* This routine writes packets to the network.  It advances the
  * send_head.  This happens as incoming acks open up the remote
  * window for us.
@@ -2703,10 +2739,9 @@ static bool tcp_write_xmit(struct sock *sk, unsigned int mss_now, int nonagle,
 	struct tcp_sock *tp = tcp_sk(sk);
 	struct sk_buff *skb;
 	unsigned int tso_segs, sent_pkts;
-	int cwnd_quota;
+	u32 cwnd_quota, max_segs;
 	int result;
 	bool is_cwnd_limited = false, is_rwnd_limited = false;
-	u32 max_segs;
 
 	sent_pkts = 0;
 
@@ -2724,6 +2759,7 @@ static bool tcp_write_xmit(struct sock *sk, unsigned int mss_now, int nonagle,
 	max_segs = tcp_tso_segs(sk, mss_now);
 	while ((skb = tcp_send_head(sk))) {
 		unsigned int limit;
+		int missing_bytes;
 
 		if (unlikely(tp->repair) && tp->repair_queue == TCP_SEND_QUEUE) {
 			/* "skb_mstamp_ns" is used as a start point for the retransmit timer */
@@ -2737,10 +2773,7 @@ static bool tcp_write_xmit(struct sock *sk, unsigned int mss_now, int nonagle,
 		if (tcp_pacing_check(sk))
 			break;
 
-		tso_segs = tcp_init_tso_segs(skb, mss_now);
-		BUG_ON(!tso_segs);
-
-		cwnd_quota = tcp_cwnd_test(tp, skb);
+		cwnd_quota = tcp_cwnd_test(tp);
 		if (!cwnd_quota) {
 			if (push_one == 2)
 				/* Force out a loss probe pkt. */
@@ -2748,6 +2781,12 @@ static bool tcp_write_xmit(struct sock *sk, unsigned int mss_now, int nonagle,
 			else
 				break;
 		}
+		cwnd_quota = min(cwnd_quota, max_segs);
+		missing_bytes = cwnd_quota * mss_now - skb->len;
+		if (missing_bytes > 0)
+			tcp_grow_skb(sk, skb, missing_bytes);
+
+		tso_segs = tcp_set_skb_tso_segs(skb, mss_now);
 
 		if (unlikely(!tcp_snd_wnd_test(tp, skb, mss_now))) {
 			is_rwnd_limited = true;
@@ -2769,9 +2808,7 @@ static bool tcp_write_xmit(struct sock *sk, unsigned int mss_now, int nonagle,
 		limit = mss_now;
 		if (tso_segs > 1 && !tcp_urg_mode(tp))
 			limit = tcp_mss_split_point(sk, skb, mss_now,
-						    min_t(unsigned int,
-							  cwnd_quota,
-							  max_segs),
+						    cwnd_quota,
 						    nonagle);
 
 		if (skb->len > limit &&
@@ -3387,11 +3424,6 @@ start:
 		err = tcp_transmit_skb(sk, skb, 1, GFP_ATOMIC);
 	}
 
-	/* To avoid taking spuriously low RTT samples based on a timestamp
-	 * for a transmit that never happened, always mark EVER_RETRANS
-	 */
-	TCP_SKB_CB(skb)->sacked |= TCPCB_EVER_RETRANS;
-
 	if (BPF_SOCK_OPS_TEST_FLAG(tp, BPF_SOCK_OPS_RETRANS_CB_FLAG))
 		tcp_call_bpf_3arg(sk, BPF_SOCK_OPS_RETRANS_CB,
 				  TCP_SKB_CB(skb)->seq, segs, err);
@@ -3401,6 +3433,12 @@ start:
 	} else if (err != -EBUSY) {
 		NET_ADD_STATS(sock_net(sk), LINUX_MIB_TCPRETRANSFAIL, segs);
 	}
+
+	/* To avoid taking spuriously low RTT samples based on a timestamp
+	 * for a transmit that never happened, always mark EVER_RETRANS
+	 */
+	TCP_SKB_CB(skb)->sacked |= TCPCB_EVER_RETRANS;
+
 	return err;
 }
 
@@ -3563,7 +3601,9 @@ void tcp_send_fin(struct sock *sk)
 			return;
 		}
 	} else {
-		skb = alloc_skb_fclone(MAX_TCP_HEADER, sk->sk_allocation);
+		skb = alloc_skb_fclone(MAX_TCP_HEADER,
+				       sk_gfp_mask(sk, GFP_ATOMIC |
+						       __GFP_NOWARN));
 		if (unlikely(!skb))
 			return;
 
@@ -3583,7 +3623,8 @@ void tcp_send_fin(struct sock *sk)
  * was unread data in the receive queue.  This behavior is recommended
  * by RFC 2525, section 2.17.  -DaveM
  */
-void tcp_send_active_reset(struct sock *sk, gfp_t priority)
+void tcp_send_active_reset(struct sock *sk, gfp_t priority,
+			   enum sk_rst_reason reason)
 {
 	struct sk_buff *skb;
 
@@ -3608,7 +3649,7 @@ void tcp_send_active_reset(struct sock *sk, gfp_t priority)
 	/* skb of trace_tcp_send_reset() keeps the skb that caused RST,
 	 * skb here is different to the troublesome skb, so use NULL
 	 */
-	trace_tcp_send_reset(sk, NULL);
+	trace_tcp_send_reset(sk, NULL, SK_RST_REASON_NOT_SPECIFIED);
 }
 
 /* Send a crossed SYN-ACK during socket establishment.
@@ -3855,7 +3896,7 @@ static void tcp_connect_init(struct sock *sk)
 	tcp_ca_dst_init(sk, dst);
 
 	if (!tp->window_clamp)
-		tp->window_clamp = dst_metric(dst, RTAX_WINDOW);
+		WRITE_ONCE(tp->window_clamp, dst_metric(dst, RTAX_WINDOW));
 	tp->advmss = tcp_mss_clamp(tp, dst_metric_advmss(dst));
 
 	tcp_initialize_rcv_mss(sk);
@@ -3863,7 +3904,7 @@ static void tcp_connect_init(struct sock *sk)
 	/* limit the window selection if the user enforce a smaller rx buffer */
 	if (sk->sk_userlocks & SOCK_RCVBUF_LOCK &&
 	    (tp->window_clamp > tcp_full_space(sk) || tp->window_clamp == 0))
-		tp->window_clamp = tcp_full_space(sk);
+		WRITE_ONCE(tp->window_clamp, tcp_full_space(sk));
 
 	rcv_wnd = tcp_rwnd_init_bpf(sk);
 	if (rcv_wnd == 0)