Merge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6

Conflicts:

	include/asm-x86/statfs.h

1 files changed, 113 insertions, 109 deletions

diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c
index 8165f5aa8c7..990a5849323 100644
--- a/net/ipv4/tcp_output.c
+++ b/net/ipv4/tcp_output.c
@@ -345,6 +345,11 @@ static void tcp_init_nondata_skb(struct sk_buff *skb, u32 seq, u8 flags)
 	TCP_SKB_CB(skb)->end_seq = seq;
 }
 
+static inline int tcp_urg_mode(const struct tcp_sock *tp)
+{
+	return tp->snd_una != tp->snd_up;
+}
+
 #define OPTION_SACK_ADVERTISE	(1 << 0)
 #define OPTION_TS		(1 << 1)
 #define OPTION_MD5		(1 << 2)
@@ -646,7 +651,8 @@ static int tcp_transmit_skb(struct sock *sk, struct sk_buff *skb, int clone_it,
 	th->check		= 0;
 	th->urg_ptr		= 0;
 
-	if (unlikely(tp->urg_mode &&
+	/* The urg_mode check is necessary during a below snd_una win probe */
+	if (unlikely(tcp_urg_mode(tp) &&
 		     between(tp->snd_up, tcb->seq + 1, tcb->seq + 0xFFFF))) {
 		th->urg_ptr		= htons(tp->snd_up - tcb->seq);
 		th->urg			= 1;
@@ -1012,7 +1018,7 @@ unsigned int tcp_sync_mss(struct sock *sk, u32 pmtu)
 /* Compute the current effective MSS, taking SACKs and IP options,
  * and even PMTU discovery events into account.
  *
- * LARGESEND note: !urg_mode is overkill, only frames up to snd_up
+ * LARGESEND note: !tcp_urg_mode is overkill, only frames up to snd_up
  * cannot be large. However, taking into account rare use of URG, this
  * is not a big flaw.
  */
@@ -1029,7 +1035,7 @@ unsigned int tcp_current_mss(struct sock *sk, int large_allowed)
 
 	mss_now = tp->mss_cache;
 
-	if (large_allowed && sk_can_gso(sk) && !tp->urg_mode)
+	if (large_allowed && sk_can_gso(sk) && !tcp_urg_mode(tp))
 		doing_tso = 1;
 
 	if (dst) {
@@ -1193,7 +1199,7 @@ static inline int tcp_nagle_test(struct tcp_sock *tp, struct sk_buff *skb,
 	/* Don't use the nagle rule for urgent data (or for the final FIN).
 	 * Nagle can be ignored during F-RTO too (see RFC4138).
 	 */
-	if (tp->urg_mode || (tp->frto_counter == 2) ||
+	if (tcp_urg_mode(tp) || (tp->frto_counter == 2) ||
 	    (TCP_SKB_CB(skb)->flags & TCPCB_FLAG_FIN))
 		return 1;
 
@@ -1824,6 +1830,8 @@ static void tcp_retrans_try_collapse(struct sock *sk, struct sk_buff *skb,
 
 	/* changed transmit queue under us so clear hints */
 	tcp_clear_retrans_hints_partial(tp);
+	if (next_skb == tp->retransmit_skb_hint)
+		tp->retransmit_skb_hint = skb;
 
 	sk_wmem_free_skb(sk, next_skb);
 }
@@ -1838,7 +1846,7 @@ void tcp_simple_retransmit(struct sock *sk)
 	struct tcp_sock *tp = tcp_sk(sk);
 	struct sk_buff *skb;
 	unsigned int mss = tcp_current_mss(sk, 0);
-	int lost = 0;
+	u32 prior_lost = tp->lost_out;
 
 	tcp_for_write_queue(skb, sk) {
 		if (skb == tcp_send_head(sk))
@@ -1849,17 +1857,13 @@ void tcp_simple_retransmit(struct sock *sk)
 				TCP_SKB_CB(skb)->sacked &= ~TCPCB_SACKED_RETRANS;
 				tp->retrans_out -= tcp_skb_pcount(skb);
 			}
-			if (!(TCP_SKB_CB(skb)->sacked & TCPCB_LOST)) {
-				TCP_SKB_CB(skb)->sacked |= TCPCB_LOST;
-				tp->lost_out += tcp_skb_pcount(skb);
-				lost = 1;
-			}
+			tcp_skb_mark_lost_uncond_verify(tp, skb);
 		}
 	}
 
-	tcp_clear_all_retrans_hints(tp);
+	tcp_clear_retrans_hints_partial(tp);
 
-	if (!lost)
+	if (prior_lost == tp->lost_out)
 		return;
 
 	if (tcp_is_reno(tp))
@@ -1934,8 +1938,8 @@ int tcp_retransmit_skb(struct sock *sk, struct sk_buff *skb)
 	/* Collapse two adjacent packets if worthwhile and we can. */
 	if (!(TCP_SKB_CB(skb)->flags & TCPCB_FLAG_SYN) &&
 	    (skb->len < (cur_mss >> 1)) &&
-	    (tcp_write_queue_next(sk, skb) != tcp_send_head(sk)) &&
 	    (!tcp_skb_is_last(sk, skb)) &&
+	    (tcp_write_queue_next(sk, skb) != tcp_send_head(sk)) &&
 	    (skb_shinfo(skb)->nr_frags == 0 &&
 	     skb_shinfo(tcp_write_queue_next(sk, skb))->nr_frags == 0) &&
 	    (tcp_skb_pcount(skb) == 1 &&
@@ -1996,86 +2000,18 @@ int tcp_retransmit_skb(struct sock *sk, struct sk_buff *skb)
 	return err;
 }
 
-/* This gets called after a retransmit timeout, and the initially
- * retransmitted data is acknowledged.  It tries to continue
- * resending the rest of the retransmit queue, until either
- * we've sent it all or the congestion window limit is reached.
- * If doing SACK, the first ACK which comes back for a timeout
- * based retransmit packet might feed us FACK information again.
- * If so, we use it to avoid unnecessarily retransmissions.
- */
-void tcp_xmit_retransmit_queue(struct sock *sk)
+static int tcp_can_forward_retransmit(struct sock *sk)
 {
 	const struct inet_connection_sock *icsk = inet_csk(sk);
 	struct tcp_sock *tp = tcp_sk(sk);
-	struct sk_buff *skb;
-	int packet_cnt;
-
-	if (tp->retransmit_skb_hint) {
-		skb = tp->retransmit_skb_hint;
-		packet_cnt = tp->retransmit_cnt_hint;
-	} else {
-		skb = tcp_write_queue_head(sk);
-		packet_cnt = 0;
-	}
-
-	/* First pass: retransmit lost packets. */
-	if (tp->lost_out) {
-		tcp_for_write_queue_from(skb, sk) {
-			__u8 sacked = TCP_SKB_CB(skb)->sacked;
-
-			if (skb == tcp_send_head(sk))
-				break;
-			/* we could do better than to assign each time */
-			tp->retransmit_skb_hint = skb;
-			tp->retransmit_cnt_hint = packet_cnt;
-
-			/* Assume this retransmit will generate
-			 * only one packet for congestion window
-			 * calculation purposes.  This works because
-			 * tcp_retransmit_skb() will chop up the
-			 * packet to be MSS sized and all the
-			 * packet counting works out.
-			 */
-			if (tcp_packets_in_flight(tp) >= tp->snd_cwnd)
-				return;
-
-			if (sacked & TCPCB_LOST) {
-				if (!(sacked & (TCPCB_SACKED_ACKED|TCPCB_SACKED_RETRANS))) {
-					int mib_idx;
-
-					if (tcp_retransmit_skb(sk, skb)) {
-						tp->retransmit_skb_hint = NULL;
-						return;
-					}
-					if (icsk->icsk_ca_state != TCP_CA_Loss)
-						mib_idx = LINUX_MIB_TCPFASTRETRANS;
-					else
-						mib_idx = LINUX_MIB_TCPSLOWSTARTRETRANS;
-					NET_INC_STATS_BH(sock_net(sk), mib_idx);
-
-					if (skb == tcp_write_queue_head(sk))
-						inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS,
-									  inet_csk(sk)->icsk_rto,
-									  TCP_RTO_MAX);
-				}
-
-				packet_cnt += tcp_skb_pcount(skb);
-				if (packet_cnt >= tp->lost_out)
-					break;
-			}
-		}
-	}
-
-	/* OK, demanded retransmission is finished. */
 
 	/* Forward retransmissions are possible only during Recovery. */
 	if (icsk->icsk_ca_state != TCP_CA_Recovery)
-		return;
+		return 0;
 
 	/* No forward retransmissions in Reno are possible. */
 	if (tcp_is_reno(tp))
-		return;
+		return 0;
 
 	/* Yeah, we have to make difficult choice between forward transmission
 	 * and retransmission... Both ways have their merits...
@@ -2086,43 +2022,104 @@ void tcp_xmit_retransmit_queue(struct sock *sk)
 	 */
 
 	if (tcp_may_send_now(sk))
-		return;
+		return 0;
 
-	/* If nothing is SACKed, highest_sack in the loop won't be valid */
-	if (!tp->sacked_out)
-		return;
+	return 1;
+}
 
-	if (tp->forward_skb_hint)
-		skb = tp->forward_skb_hint;
-	else
+/* This gets called after a retransmit timeout, and the initially
+ * retransmitted data is acknowledged.  It tries to continue
+ * resending the rest of the retransmit queue, until either
+ * we've sent it all or the congestion window limit is reached.
+ * If doing SACK, the first ACK which comes back for a timeout
+ * based retransmit packet might feed us FACK information again.
+ * If so, we use it to avoid unnecessarily retransmissions.
+ */
+void tcp_xmit_retransmit_queue(struct sock *sk)
+{
+	const struct inet_connection_sock *icsk = inet_csk(sk);
+	struct tcp_sock *tp = tcp_sk(sk);
+	struct sk_buff *skb;
+	struct sk_buff *hole = NULL;
+	u32 last_lost;
+	int mib_idx;
+	int fwd_rexmitting = 0;
+
+	if (!tp->lost_out)
+		tp->retransmit_high = tp->snd_una;
+
+	if (tp->retransmit_skb_hint) {
+		skb = tp->retransmit_skb_hint;
+		last_lost = TCP_SKB_CB(skb)->end_seq;
+		if (after(last_lost, tp->retransmit_high))
+			last_lost = tp->retransmit_high;
+	} else {
 		skb = tcp_write_queue_head(sk);
+		last_lost = tp->snd_una;
+	}
 
+	/* First pass: retransmit lost packets. */
 	tcp_for_write_queue_from(skb, sk) {
-		if (skb == tcp_send_head(sk))
-			break;
-		tp->forward_skb_hint = skb;
+		__u8 sacked = TCP_SKB_CB(skb)->sacked;
 
-		if (!before(TCP_SKB_CB(skb)->seq, tcp_highest_sack_seq(tp)))
+		if (skb == tcp_send_head(sk))
 			break;
+		/* we could do better than to assign each time */
+		if (hole == NULL)
+			tp->retransmit_skb_hint = skb;
 
+		/* Assume this retransmit will generate
+		 * only one packet for congestion window
+		 * calculation purposes.  This works because
+		 * tcp_retransmit_skb() will chop up the
+		 * packet to be MSS sized and all the
+		 * packet counting works out.
+		 */
 		if (tcp_packets_in_flight(tp) >= tp->snd_cwnd)
-			break;
+			return;
+
+		if (fwd_rexmitting) {
+begin_fwd:
+			if (!before(TCP_SKB_CB(skb)->seq, tcp_highest_sack_seq(tp)))
+				break;
+			mib_idx = LINUX_MIB_TCPFORWARDRETRANS;
+
+		} else if (!before(TCP_SKB_CB(skb)->seq, tp->retransmit_high)) {
+			tp->retransmit_high = last_lost;
+			if (!tcp_can_forward_retransmit(sk))
+				break;
+			/* Backtrack if necessary to non-L'ed skb */
+			if (hole != NULL) {
+				skb = hole;
+				hole = NULL;
+			}
+			fwd_rexmitting = 1;
+			goto begin_fwd;
 
-		if (TCP_SKB_CB(skb)->sacked & TCPCB_TAGBITS)
+		} else if (!(sacked & TCPCB_LOST)) {
+			if (hole == NULL && !(sacked & TCPCB_SACKED_RETRANS))
+				hole = skb;
 			continue;
 
-		/* Ok, retransmit it. */
-		if (tcp_retransmit_skb(sk, skb)) {
-			tp->forward_skb_hint = NULL;
-			break;
+		} else {
+			last_lost = TCP_SKB_CB(skb)->end_seq;
+			if (icsk->icsk_ca_state != TCP_CA_Loss)
+				mib_idx = LINUX_MIB_TCPFASTRETRANS;
+			else
+				mib_idx = LINUX_MIB_TCPSLOWSTARTRETRANS;
 		}
 
+		if (sacked & (TCPCB_SACKED_ACKED|TCPCB_SACKED_RETRANS))
+			continue;
+
+		if (tcp_retransmit_skb(sk, skb))
+			return;
+		NET_INC_STATS_BH(sock_net(sk), mib_idx);
+
 		if (skb == tcp_write_queue_head(sk))
 			inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS,
 						  inet_csk(sk)->icsk_rto,
 						  TCP_RTO_MAX);
-
-		NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_TCPFORWARDRETRANS);
 	}
 }
 
@@ -2241,6 +2238,7 @@ struct sk_buff *tcp_make_synack(struct sock *sk, struct dst_entry *dst,
 	struct sk_buff *skb;
 	struct tcp_md5sig_key *md5;
 	__u8 *md5_hash_location;
+	int mss;
 
 	skb = sock_wmalloc(sk, MAX_TCP_HEADER + 15, 1, GFP_ATOMIC);
 	if (skb == NULL)
@@ -2251,13 +2249,17 @@ struct sk_buff *tcp_make_synack(struct sock *sk, struct dst_entry *dst,
 
 	skb->dst = dst_clone(dst);
 
+	mss = dst_metric(dst, RTAX_ADVMSS);
+	if (tp->rx_opt.user_mss && tp->rx_opt.user_mss < mss)
+		mss = tp->rx_opt.user_mss;
+
 	if (req->rcv_wnd == 0) { /* ignored for retransmitted syns */
 		__u8 rcv_wscale;
 		/* Set this up on the first call only */
 		req->window_clamp = tp->window_clamp ? : dst_metric(dst, RTAX_WINDOW);
 		/* tcp_full_space because it is guaranteed to be the first packet */
 		tcp_select_initial_window(tcp_full_space(sk),
-			dst_metric(dst, RTAX_ADVMSS) - (ireq->tstamp_ok ? TCPOLEN_TSTAMP_ALIGNED : 0),
+			mss - (ireq->tstamp_ok ? TCPOLEN_TSTAMP_ALIGNED : 0),
 			&req->rcv_wnd,
 			&req->window_clamp,
 			ireq->wscale_ok,
@@ -2267,8 +2269,7 @@ struct sk_buff *tcp_make_synack(struct sock *sk, struct dst_entry *dst,
 
 	memset(&opts, 0, sizeof(opts));
 	TCP_SKB_CB(skb)->when = tcp_time_stamp;
-	tcp_header_size = tcp_synack_options(sk, req,
-					     dst_metric(dst, RTAX_ADVMSS),
+	tcp_header_size = tcp_synack_options(sk, req, mss,
 					     skb, &opts, &md5) +
 			  sizeof(struct tcphdr);
 
@@ -2280,7 +2281,7 @@ struct sk_buff *tcp_make_synack(struct sock *sk, struct dst_entry *dst,
 	th->syn = 1;
 	th->ack = 1;
 	TCP_ECN_make_synack(req, th);
-	th->source = inet_sk(sk)->sport;
+	th->source = ireq->loc_port;
 	th->dest = ireq->rmt_port;
 	/* Setting of flags are superfluous here for callers (and ECE is
 	 * not even correctly set)
@@ -2342,6 +2343,9 @@ static void tcp_connect_init(struct sock *sk)
 	if (!tp->window_clamp)
 		tp->window_clamp = dst_metric(dst, RTAX_WINDOW);
 	tp->advmss = dst_metric(dst, RTAX_ADVMSS);
+	if (tp->rx_opt.user_mss && tp->rx_opt.user_mss < tp->advmss)
+		tp->advmss = tp->rx_opt.user_mss;
+
 	tcp_initialize_rcv_mss(sk);
 
 	tcp_select_initial_window(tcp_full_space(sk),
@@ -2360,6 +2364,7 @@ static void tcp_connect_init(struct sock *sk)
 	tcp_init_wl(tp, tp->write_seq, 0);
 	tp->snd_una = tp->write_seq;
 	tp->snd_sml = tp->write_seq;
+	tp->snd_up = tp->write_seq;
 	tp->rcv_nxt = 0;
 	tp->rcv_wup = 0;
 	tp->copied_seq = 0;
@@ -2569,8 +2574,7 @@ int tcp_write_wakeup(struct sock *sk)
 			tcp_event_new_data_sent(sk, skb);
 		return err;
 	} else {
-		if (tp->urg_mode &&
-		    between(tp->snd_up, tp->snd_una + 1, tp->snd_una + 0xFFFF))
+		if (between(tp->snd_up, tp->snd_una + 1, tp->snd_una + 0xFFFF))
 			tcp_xmit_probe_skb(sk, 1);
 		return tcp_xmit_probe_skb(sk, 0);
 	}