[TCP] FRTO: Improve interoperability with other undo_marker users
Basically this change enables it, previously other undo_marker
users were left with nothing. Reverse undo_marker logic
completely to get it set right in CA_Loss. On the other hand,
when spurious RTO is detected, clear it. Clearing might be too
heavy for some scenarios but seems safe enough starting point
for now and shouldn't have much effect except in majority of
cases (if in any).
By adding a new FLAG_ we avoid looping through write_queue when
RTO occurs.
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index 8fe754b..0feb109 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -104,6 +104,7 @@
#define FLAG_ONLY_ORIG_SACKED 0x200 /* SACKs only non-rexmit sent before RTO */
#define FLAG_SND_UNA_ADVANCED 0x400 /* Snd_una was changed (!= FLAG_DATA_ACKED) */
#define FLAG_DSACKING_ACK 0x800 /* SACK blocks contained DSACK info */
+#define FLAG_NONHEAD_RETRANS_ACKED 0x1000 /* Non-head rexmitted data was ACKed */
#define FLAG_ACKED (FLAG_DATA_ACKED|FLAG_SYN_ACKED)
#define FLAG_NOT_DUP (FLAG_DATA|FLAG_WIN_UPDATE|FLAG_ACKED)
@@ -1594,6 +1595,8 @@
tp->undo_retrans = 0;
skb = tcp_write_queue_head(sk);
+ if (TCP_SKB_CB(skb)->sacked & TCPCB_RETRANS)
+ tp->undo_marker = 0;
if (TCP_SKB_CB(skb)->sacked & TCPCB_SACKED_RETRANS) {
TCP_SKB_CB(skb)->sacked &= ~TCPCB_SACKED_RETRANS;
tp->retrans_out -= tcp_skb_pcount(skb);
@@ -1643,6 +1646,8 @@
/* ...enter this if branch just for the first segment */
flag |= FLAG_DATA_ACKED;
} else {
+ if (TCP_SKB_CB(skb)->sacked & TCPCB_RETRANS)
+ tp->undo_marker = 0;
TCP_SKB_CB(skb)->sacked &= ~(TCPCB_LOST|TCPCB_SACKED_RETRANS);
}
@@ -1658,7 +1663,6 @@
tp->snd_cwnd = tcp_packets_in_flight(tp) + allowed_segments;
tp->snd_cwnd_cnt = 0;
tp->snd_cwnd_stamp = tcp_time_stamp;
- tp->undo_marker = 0;
tp->frto_counter = 0;
tp->reordering = min_t(unsigned int, tp->reordering,
@@ -2584,20 +2588,6 @@
end_seq = scb->end_seq;
}
- /* Initial outgoing SYN's get put onto the write_queue
- * just like anything else we transmit. It is not
- * true data, and if we misinform our callers that
- * this ACK acks real data, we will erroneously exit
- * connection startup slow start one packet too
- * quickly. This is severely frowned upon behavior.
- */
- if (!(scb->flags & TCPCB_FLAG_SYN)) {
- flag |= FLAG_DATA_ACKED;
- } else {
- flag |= FLAG_SYN_ACKED;
- tp->retrans_stamp = 0;
- }
-
/* MTU probing checks */
if (fully_acked && icsk->icsk_mtup.probe_size &&
!after(tp->mtu_probe.probe_seq_end, scb->end_seq)) {
@@ -2610,6 +2600,9 @@
tp->retrans_out -= packets_acked;
flag |= FLAG_RETRANS_DATA_ACKED;
seq_rtt = -1;
+ if ((flag & FLAG_DATA_ACKED) ||
+ (packets_acked > 1))
+ flag |= FLAG_NONHEAD_RETRANS_ACKED;
} else if (seq_rtt < 0) {
seq_rtt = now - scb->when;
if (fully_acked)
@@ -2631,6 +2624,20 @@
}
tp->packets_out -= packets_acked;
+ /* Initial outgoing SYN's get put onto the write_queue
+ * just like anything else we transmit. It is not
+ * true data, and if we misinform our callers that
+ * this ACK acks real data, we will erroneously exit
+ * connection startup slow start one packet too
+ * quickly. This is severely frowned upon behavior.
+ */
+ if (!(scb->flags & TCPCB_FLAG_SYN)) {
+ flag |= FLAG_DATA_ACKED;
+ } else {
+ flag |= FLAG_SYN_ACKED;
+ tp->retrans_stamp = 0;
+ }
+
if (!fully_acked)
break;
@@ -2852,6 +2859,10 @@
if (flag&FLAG_DATA_ACKED)
inet_csk(sk)->icsk_retransmits = 0;
+ if ((flag & FLAG_NONHEAD_RETRANS_ACKED) ||
+ ((tp->frto_counter >= 2) && (flag & FLAG_RETRANS_DATA_ACKED)))
+ tp->undo_marker = 0;
+
if (!before(tp->snd_una, tp->frto_highmark)) {
tcp_enter_frto_loss(sk, (tp->frto_counter == 1 ? 2 : 3), flag);
return 1;
@@ -2916,6 +2927,7 @@
break;
}
tp->frto_counter = 0;
+ tp->undo_marker = 0;
}
return 0;
}