sctp: Update max.burst implementation

Current implementation of max.burst ends up limiting new
data during cwnd decay period.  The decay is happening becuase
the connection is idle and we are allowed to fill the congestion
window.  The point of max.burst is to limit micro-bursts in response
to large acks.  This still happens, as max.burst is still applied
to each transmit opportunity.  It will also apply if a very large
send is made (greater then allowed by burst).

Tested-by: Florian Niederbacher <florian.niederbacher@student.uibk.ac.at>
Signed-off-by: Vlad Yasevich <vladislav.yasevich@hp.com>

5 files changed, 58 insertions, 23 deletions

diff --git a/include/net/sctp/structs.h b/include/net/sctp/structs.h
index ac794a6823eb..bc3f8d879c5c 100644
--- a/include/net/sctp/structs.h
+++ b/include/net/sctp/structs.h
@@ -942,6 +942,8 @@ struct sctp_transport {
 	/* Data that has been sent, but not acknowledged. */
 	__u32 flight_size;
 
+	__u32 burst_limited;	/* Holds old cwnd when max.burst is applied */
+
 	/* TSN marking the fast recovery exit point */
 	__u32 fast_recovery_exit;
 
@@ -1070,6 +1072,8 @@ void sctp_transport_put(struct sctp_transport *);
 void sctp_transport_update_rto(struct sctp_transport *, __u32);
 void sctp_transport_raise_cwnd(struct sctp_transport *, __u32, __u32);
 void sctp_transport_lower_cwnd(struct sctp_transport *, sctp_lower_cwnd_t);
+void sctp_transport_burst_limited(struct sctp_transport *);
+void sctp_transport_burst_reset(struct sctp_transport *);
 unsigned long sctp_transport_timeout(struct sctp_transport *);
 void sctp_transport_reset(struct sctp_transport *);
 void sctp_transport_update_pmtu(struct sctp_transport *, u32);
diff --git a/net/sctp/associola.c b/net/sctp/associola.c
index 37e982510bea..880dae2ca87b 100644
--- a/net/sctp/associola.c
+++ b/net/sctp/associola.c
@@ -738,6 +738,7 @@ struct sctp_transport *sctp_assoc_add_peer(struct sctp_association *asoc,
 
 	peer->partial_bytes_acked = 0;
 	peer->flight_size = 0;
+	peer->burst_limited = 0;
 
 	/* Set the transport's RTO.initial value */
 	peer->rto = asoc->rto_initial;
diff --git a/net/sctp/output.c b/net/sctp/output.c
index 9e8e0ea844be..b210d2077e28 100644
--- a/net/sctp/output.c
+++ b/net/sctp/output.c
@@ -615,7 +615,6 @@ static sctp_xmit_t sctp_packet_can_append_data(struct sctp_packet *packet,
 	sctp_xmit_t retval = SCTP_XMIT_OK;
 	size_t datasize, rwnd, inflight, flight_size;
 	struct sctp_transport *transport = packet->transport;
-	__u32 max_burst_bytes;
 	struct sctp_association *asoc = transport->asoc;
 	struct sctp_outq *q = &asoc->outqueue;
 
@@ -648,28 +647,6 @@ static sctp_xmit_t sctp_packet_can_append_data(struct sctp_packet *packet,
 		}
 	}
 
-	/* sctpimpguide-05 2.14.2
-	 * D) When the time comes for the sender to
-	 * transmit new DATA chunks, the protocol parameter Max.Burst MUST
-	 * first be applied to limit how many new DATA chunks may be sent.
-	 * The limit is applied by adjusting cwnd as follows:
-	 * 	if ((flightsize + Max.Burst * MTU) < cwnd)
-	 *		cwnd = flightsize + Max.Burst * MTU
-	 */
-	max_burst_bytes = asoc->max_burst * asoc->pathmtu;
-	if ((flight_size + max_burst_bytes) < transport->cwnd) {
-		transport->cwnd = flight_size + max_burst_bytes;
-		SCTP_DEBUG_PRINTK("%s: cwnd limited by max_burst: "
-				  "transport: %p, cwnd: %d, "
-				  "ssthresh: %d, flight_size: %d, "
-				  "pba: %d\n",
-				  __func__, transport,
-				  transport->cwnd,
-				  transport->ssthresh,
-				  transport->flight_size,
-				  transport->partial_bytes_acked);
-	}
-
 	/* RFC 2960 6.1  Transmission of DATA Chunks
 	 *
 	 * B) At any given time, the sender MUST NOT transmit new data
diff --git a/net/sctp/outqueue.c b/net/sctp/outqueue.c
index 5732661c87d3..2f2377369e2b 100644
--- a/net/sctp/outqueue.c
+++ b/net/sctp/outqueue.c
@@ -931,6 +931,14 @@ static int sctp_outq_flush(struct sctp_outq *q, int rtx_timeout)
 				goto sctp_flush_out;
 		}
 
+		/* Apply Max.Burst limitation to the current transport in
+		 * case it will be used for new data.  We are going to
+		 * rest it before we return, but we want to apply the limit
+		 * to the currently queued data.
+		 */
+		if (transport)
+			sctp_transport_burst_limited(transport);
+
 		/* Finally, transmit new packets.  */
 		while ((chunk = sctp_outq_dequeue_data(q)) != NULL) {
 			/* RFC 2960 6.5 Every DATA chunk MUST carry a valid
@@ -976,6 +984,10 @@ static int sctp_outq_flush(struct sctp_outq *q, int rtx_timeout)
 				packet = &transport->packet;
 				sctp_packet_config(packet, vtag,
 						   asoc->peer.ecn_capable);
+				/* We've switched transports, so apply the
+				 * Burst limit to the new transport.
+				 */
+				sctp_transport_burst_limited(transport);
 			}
 
 			SCTP_DEBUG_PRINTK("sctp_outq_flush(%p, %p[%s]), ",
@@ -1070,6 +1082,9 @@ sctp_flush_out:
 		packet = &t->packet;
 		if (!sctp_packet_empty(packet))
 			error = sctp_packet_transmit(packet);
+
+		/* Clear the burst limited state, if any */
+		sctp_transport_burst_reset(t);
 	}
 
 	return error;
diff --git a/net/sctp/transport.c b/net/sctp/transport.c
index 2df29cbdaf5a..9a6cb22d129f 100644
--- a/net/sctp/transport.c
+++ b/net/sctp/transport.c
@@ -576,6 +576,43 @@ void sctp_transport_lower_cwnd(struct sctp_transport *transport,
 			  transport->cwnd, transport->ssthresh);
 }
 
+/* Apply Max.Burst limit to the congestion window:
+ * sctpimpguide-05 2.14.2
+ * D) When the time comes for the sender to
+ * transmit new DATA chunks, the protocol parameter Max.Burst MUST
+ * first be applied to limit how many new DATA chunks may be sent.
+ * The limit is applied by adjusting cwnd as follows:
+ * 	if ((flightsize+ Max.Burst * MTU) < cwnd)
+ * 		cwnd = flightsize + Max.Burst * MTU
+ */
+
+void sctp_transport_burst_limited(struct sctp_transport *t)
+{
+	struct sctp_association *asoc = t->asoc;
+	u32 old_cwnd = t->cwnd;
+	u32 max_burst_bytes;
+
+	if (t->burst_limited)
+		return;
+
+	max_burst_bytes = t->flight_size + (asoc->max_burst * asoc->pathmtu);
+	if (max_burst_bytes < old_cwnd) {
+		t->cwnd = max_burst_bytes;
+		t->burst_limited = old_cwnd;
+	}
+}
+
+/* Restore the old cwnd congestion window, after the burst had it's
+ * desired effect.
+ */
+void sctp_transport_burst_reset(struct sctp_transport *t)
+{
+	if (t->burst_limited) {
+		t->cwnd = t->burst_limited;
+		t->burst_limited = 0;
+	}
+}
+
 /* What is the next timeout value for this transport? */
 unsigned long sctp_transport_timeout(struct sctp_transport *t)
 {
@@ -598,6 +635,7 @@ void sctp_transport_reset(struct sctp_transport *t)
 	 * (see Section 6.2.1)
 	 */
 	t->cwnd = min(4*asoc->pathmtu, max_t(__u32, 2*asoc->pathmtu, 4380));
+	t->burst_limited = 0;
 	t->ssthresh = asoc->peer.i.a_rwnd;
 	t->last_rto = t->rto = asoc->rto_initial;
 	t->rtt = 0;