Merge branch 'net-next' of git://git.kernel.org/pub/scm/linux/kernel/git/vxy/lksctp-dev

Add missing linux/vmalloc.h include to net/sctp/probe.c

Signed-off-by: David S. Miller <davem@davemloft.net>

include/net/sctp/sctp.h

@@ -547,7 +547,7 @@ for (pos = chunk->subh.fwdtsn_hdr->skip;\
 #define WORD_ROUND(s) (((s)+3)&~3)
 
 /* Make a new instance of type.  */
-#define t_new(type, flags)	(type *)kmalloc(sizeof(type), flags)
+#define t_new(type, flags)	(type *)kzalloc(sizeof(type), flags)
 
 /* Compare two timevals.  */
 #define tv_lt(s, t) \

include/net/sctp/sm.h

@@ -437,7 +437,7 @@ sctp_vtag_verify_either(const struct sctp_chunk *chunk,
 	 */
         if ((!sctp_test_T_bit(chunk) &&
              (ntohl(chunk->sctp_hdr->vtag) == asoc->c.my_vtag)) ||
-	    (sctp_test_T_bit(chunk) &&
+	    (sctp_test_T_bit(chunk) && asoc->c.peer_vtag &&
 	     (ntohl(chunk->sctp_hdr->vtag) == asoc->c.peer_vtag))) {
                 return 1;
 	}

include/net/sctp/structs.h

@@ -643,17 +643,15 @@ struct sctp_pf {
 struct sctp_datamsg {
 	/* Chunks waiting to be submitted to lower layer. */
 	struct list_head chunks;
-	/* Chunks that have been transmitted. */
-	size_t msg_size;
 	/* Reference counting. */
 	atomic_t refcnt;
 	/* When is this message no longer interesting to the peer? */
 	unsigned long expires_at;
 	/* Did the messenge fail to send? */
 	int send_error;
-	char send_failed;
-	/* Control whether chunks from this message can be abandoned. */
-	char can_abandon;
+	u8 send_failed:1,
+	   can_abandon:1,   /* can chunks from this message can be abandoned. */
+	   can_delay;	    /* should this message be Nagle delayed */
 };
 
 struct sctp_datamsg *sctp_datamsg_from_user(struct sctp_association *,
@@ -757,7 +755,6 @@ struct sctp_chunk {
 #define SCTP_NEED_FRTX 0x1
 #define SCTP_DONT_FRTX 0x2
 	__u16	rtt_in_progress:1,	/* This chunk used for RTT calc? */
-		resent:1,		/* Has this chunk ever been resent. */
 		has_tsn:1,		/* Does this chunk have a TSN yet? */
 		has_ssn:1,		/* Does this chunk have a SSN yet? */
 		singleton:1,		/* Only chunk in the packet? */
@@ -879,7 +876,30 @@ struct sctp_transport {
 
 	/* Reference counting. */
 	atomic_t refcnt;
-	int	 dead;
+	int	 dead:1,
+		/* RTO-Pending : A flag used to track if one of the DATA
+		 *		chunks sent to this address is currently being
+		 *		used to compute a RTT. If this flag is 0,
+		 *		the next DATA chunk sent to this destination
+		 *		should be used to compute a RTT and this flag
+		 *		should be set. Every time the RTT
+		 *		calculation completes (i.e. the DATA chunk
+		 *		is SACK'd) clear this flag.
+		 */
+		 rto_pending:1,
+
+		/*
+		 * hb_sent : a flag that signals that we have a pending
+		 * heartbeat.
+		 */
+		hb_sent:1,
+
+		/* Is the Path MTU update pending on this tranport */
+		pmtu_pending:1,
+
+		/* Is this structure kfree()able? */
+		malloced:1;
+
 
 	/* This is the peer's IP address and port. */
 	union sctp_addr ipaddr;
@@ -909,22 +929,6 @@ struct sctp_transport {
 	/* SRTT	       : The current smoothed round trip time.	*/
 	__u32 srtt;
 
-	/* RTO-Pending : A flag used to track if one of the DATA
-	 *		chunks sent to this address is currently being
-	 *		used to compute a RTT. If this flag is 0,
-	 *		the next DATA chunk sent to this destination
-	 *		should be used to compute a RTT and this flag
-	 *		should be set. Every time the RTT
-	 *		calculation completes (i.e. the DATA chunk
-	 *		is SACK'd) clear this flag.
-	 * hb_sent : a flag that signals that we have a pending heartbeat.
-	 */
-	__u8 rto_pending;
-	__u8 hb_sent;
-
-	/* Flag to track the current fast recovery state */
-	__u8 fast_recovery;
-
 	/*
 	 * These are the congestion stats.
 	 */
@@ -944,9 +948,6 @@ struct sctp_transport {
 
 	__u32 burst_limited;	/* Holds old cwnd when max.burst is applied */
 
-	/* TSN marking the fast recovery exit point */
-	__u32 fast_recovery_exit;
-
 	/* Destination */
 	struct dst_entry *dst;
 	/* Source address. */
@@ -977,9 +978,6 @@ struct sctp_transport {
 	 */
 	__u16 pathmaxrxt;
 
-	/* is the Path MTU update pending on this tranport */
-	__u8 pmtu_pending;
-
 	/* PMTU	      : The current known path MTU.  */
 	__u32 pathmtu;
 
@@ -1023,8 +1021,6 @@ struct sctp_transport {
 	/* This is the list of transports that have chunks to send.  */
 	struct list_head send_ready;
 
-	int malloced; /* Is this structure kfree()able? */
-
 	/* State information saved for SFR_CACC algorithm. The key
 	 * idea in SFR_CACC is to maintain state at the sender on a
 	 * per-destination basis when a changeover happens.
@@ -1066,7 +1062,7 @@ void sctp_transport_route(struct sctp_transport *, union sctp_addr *,
 			  struct sctp_sock *);
 void sctp_transport_pmtu(struct sctp_transport *);
 void sctp_transport_free(struct sctp_transport *);
-void sctp_transport_reset_timers(struct sctp_transport *, int);
+void sctp_transport_reset_timers(struct sctp_transport *);
 void sctp_transport_hold(struct sctp_transport *);
 void sctp_transport_put(struct sctp_transport *);
 void sctp_transport_update_rto(struct sctp_transport *, __u32);
@@ -1720,6 +1716,12 @@ struct sctp_association {
 	/* Highest TSN that is acknowledged by incoming SACKs. */
 	__u32 highest_sacked;
 
+	/* TSN marking the fast recovery exit point */
+	__u32 fast_recovery_exit;
+
+	/* Flag to track the current fast recovery state */
+	__u8 fast_recovery;
+
 	/* The number of unacknowledged data chunks.  Reported through
 	 * the SCTP_STATUS sockopt.
 	 */

net/sctp/Kconfig

@@ -37,6 +37,18 @@ menuconfig IP_SCTP
 
 if IP_SCTP
 
+config NET_SCTPPROBE
+	tristate "SCTP: Association probing"
+        depends on PROC_FS && KPROBES
+        ---help---
+        This module allows for capturing the changes to SCTP association
+        state in response to incoming packets. It is used for debugging
+        SCTP congestion control algorithms. If you don't understand
+        what was just said, you don't need it: say N.
+
+        To compile this code as a module, choose M here: the
+        module will be called sctp_probe.
+
 config SCTP_DBG_MSG
 	bool "SCTP: Debug messages"
 	help

net/sctp/Makefile

@@ -3,6 +3,7 @@
 #
 
 obj-$(CONFIG_IP_SCTP) += sctp.o
+obj-$(CONFIG_NET_SCTPPROBE) += sctp_probe.o
 
 sctp-y := sm_statetable.o sm_statefuns.o sm_sideeffect.o \
 	  protocol.o endpointola.o associola.o \
@@ -11,6 +12,8 @@ sctp-y := sm_statetable.o sm_statefuns.o sm_sideeffect.o \
 	  tsnmap.o bind_addr.o socket.o primitive.o \
 	  output.o input.o debug.o ssnmap.o auth.o
 
+sctp_probe-y := probe.o
+
 sctp-$(CONFIG_SCTP_DBG_OBJCNT) += objcnt.o
 sctp-$(CONFIG_PROC_FS) += proc.o
 sctp-$(CONFIG_SYSCTL) += sysctl.o

net/sctp/associola.c

@@ -87,9 +87,6 @@ static struct sctp_association *sctp_association_init(struct sctp_association *a
 	/* Retrieve the SCTP per socket area.  */
 	sp = sctp_sk((struct sock *)sk);
 
-	/* Init all variables to a known value.  */
-	memset(asoc, 0, sizeof(struct sctp_association));
-
 	/* Discarding const is appropriate here.  */
 	asoc->ep = (struct sctp_endpoint *)ep;
 	sctp_endpoint_hold(asoc->ep);
@@ -762,7 +759,8 @@ struct sctp_transport *sctp_assoc_add_peer(struct sctp_association *asoc,
 		asoc->peer.retran_path = peer;
 	}
 
-	if (asoc->peer.active_path == asoc->peer.retran_path) {
+	if (asoc->peer.active_path == asoc->peer.retran_path &&
+	    peer->state != SCTP_UNCONFIRMED) {
 		asoc->peer.retran_path = peer;
 	}
 
@@ -1320,12 +1318,13 @@ void sctp_assoc_update_retran_path(struct sctp_association *asoc)
 			/* Keep track of the next transport in case
 			 * we don't find any active transport.
 			 */
-			if (!next)
+			if (t->state != SCTP_UNCONFIRMED && !next)
 				next = t;
 		}
 	}
 
-	asoc->peer.retran_path = t;
+	if (t)
+		asoc->peer.retran_path = t;
 
 	SCTP_DEBUG_PRINTK_IPADDR("sctp_assoc_update_retran_path:association"
 				 " %p addr: ",
@@ -1485,7 +1484,7 @@ void sctp_assoc_rwnd_decrease(struct sctp_association *asoc, unsigned len)
 	if (asoc->rwnd >= len) {
 		asoc->rwnd -= len;
 		if (over) {
-			asoc->rwnd_press = asoc->rwnd;
+			asoc->rwnd_press += asoc->rwnd;
 			asoc->rwnd = 0;
 		}
 	} else {

net/sctp/chunk.c

@@ -58,9 +58,9 @@ static void sctp_datamsg_init(struct sctp_datamsg *msg)
 	msg->send_failed = 0;
 	msg->send_error = 0;
 	msg->can_abandon = 0;
+	msg->can_delay = 1;
 	msg->expires_at = 0;
 	INIT_LIST_HEAD(&msg->chunks);
-	msg->msg_size = 0;
 }
 
 /* Allocate and initialize datamsg. */
@@ -157,7 +157,6 @@ static void sctp_datamsg_assign(struct sctp_datamsg *msg, struct sctp_chunk *chu
 {
 	sctp_datamsg_hold(msg);
 	chunk->msg = msg;
-	msg->msg_size += chunk->skb->len;
 }
 
 
@@ -247,6 +246,7 @@ struct sctp_datamsg *sctp_datamsg_from_user(struct sctp_association *asoc,
 	if (msg_len >= first_len) {
 		msg_len -= first_len;
 		whole = 1;
+		msg->can_delay = 0;
 	}
 
 	/* How many full sized?  How many bytes leftover? */

net/sctp/endpointola.c

@@ -70,8 +70,6 @@ static struct sctp_endpoint *sctp_endpoint_init(struct sctp_endpoint *ep,
 	struct sctp_shared_key *null_key;
 	int err;
 
-	memset(ep, 0, sizeof(struct sctp_endpoint));
-
 	ep->digest = kzalloc(SCTP_SIGNATURE_SIZE, gfp);
 	if (!ep->digest)
 		return NULL;

net/sctp/output.c

@@ -429,24 +429,17 @@ int sctp_packet_transmit(struct sctp_packet *packet)
 	list_for_each_entry_safe(chunk, tmp, &packet->chunk_list, list) {
 		list_del_init(&chunk->list);
 		if (sctp_chunk_is_data(chunk)) {
+			/* 6.3.1 C4) When data is in flight and when allowed
+			 * by rule C5, a new RTT measurement MUST be made each
+			 * round trip.  Furthermore, new RTT measurements
+			 * SHOULD be made no more than once per round-trip
+			 * for a given destination transport address.
+			 */
 
-			if (!chunk->resent) {
-
-				/* 6.3.1 C4) When data is in flight and when allowed
-				 * by rule C5, a new RTT measurement MUST be made each
-				 * round trip.  Furthermore, new RTT measurements
-				 * SHOULD be made no more than once per round-trip
-				 * for a given destination transport address.
-				 */
-
-				if (!tp->rto_pending) {
-					chunk->rtt_in_progress = 1;
-					tp->rto_pending = 1;
-				}
+			if (!tp->rto_pending) {
+				chunk->rtt_in_progress = 1;
+				tp->rto_pending = 1;
 			}
-
-			chunk->resent = 1;
-
 			has_data = 1;
 		}
 
@@ -681,7 +674,7 @@ static sctp_xmit_t sctp_packet_can_append_data(struct sctp_packet *packet,
 		 * Don't delay large message writes that may have been
 		 * fragmeneted into small peices.
 		 */
-		if ((len < max) && (chunk->msg->msg_size < max)) {
+		if ((len < max) && chunk->msg->can_delay) {
 			retval = SCTP_XMIT_NAGLE_DELAY;
 			goto finish;
 		}

net/sctp/outqueue.c

@@ -62,7 +62,7 @@ static void sctp_check_transmitted(struct sctp_outq *q,
 				   struct list_head *transmitted_queue,
 				   struct sctp_transport *transport,
 				   struct sctp_sackhdr *sack,
-				   __u32 highest_new_tsn);
+				   __u32 *highest_new_tsn);
 
 static void sctp_mark_missing(struct sctp_outq *q,
 			      struct list_head *transmitted_queue,
@@ -308,7 +308,7 @@ int sctp_outq_tail(struct sctp_outq *q, struct sctp_chunk *chunk)
 	/* If it is data, queue it up, otherwise, send it
 	 * immediately.
 	 */
-	if (SCTP_CID_DATA == chunk->chunk_hdr->type) {
+	if (sctp_chunk_is_data(chunk)) {
 		/* Is it OK to queue data chunks?  */
 		/* From 9. Termination of Association
 		 *
@@ -598,11 +598,23 @@ static int sctp_outq_flush_rtx(struct sctp_outq *q, struct sctp_packet *pkt,
 		if (fast_rtx && !chunk->fast_retransmit)
 			continue;
 
+redo:
 		/* Attempt to append this chunk to the packet. */
 		status = sctp_packet_append_chunk(pkt, chunk);
 
 		switch (status) {
 		case SCTP_XMIT_PMTU_FULL:
+			if (!pkt->has_data && !pkt->has_cookie_echo) {
+				/* If this packet did not contain DATA then
+				 * retransmission did not happen, so do it
+				 * again.  We'll ignore the error here since
+				 * control chunks are already freed so there
+				 * is nothing we can do.
+				 */
+				sctp_packet_transmit(pkt);
+				goto redo;
+			}
+
 			/* Send this packet.  */
 			error = sctp_packet_transmit(pkt);
 
@@ -647,14 +659,6 @@ static int sctp_outq_flush_rtx(struct sctp_outq *q, struct sctp_packet *pkt,
 			if (chunk->fast_retransmit == SCTP_NEED_FRTX)
 				chunk->fast_retransmit = SCTP_DONT_FRTX;
 
-			/* Force start T3-rtx timer when fast retransmitting
-			 * the earliest outstanding TSN
-			 */
-			if (!timer && fast_rtx &&
-			    ntohl(chunk->subh.data_hdr->tsn) ==
-					     asoc->ctsn_ack_point + 1)
-				timer = 2;
-
 			q->empty = 0;
 			break;
 		}
@@ -854,6 +858,12 @@ static int sctp_outq_flush(struct sctp_outq *q, int rtx_timeout)
 			if (status  != SCTP_XMIT_OK) {
 				/* put the chunk back */
 				list_add(&chunk->list, &q->control_chunk_list);
+			} else if (chunk->chunk_hdr->type == SCTP_CID_FWD_TSN) {
+				/* PR-SCTP C5) If a FORWARD TSN is sent, the
+				 * sender MUST assure that at least one T3-rtx
+				 * timer is running.
+				 */
+				sctp_transport_reset_timers(transport);
 			}
 			break;
 
@@ -906,8 +916,7 @@ static int sctp_outq_flush(struct sctp_outq *q, int rtx_timeout)
 						    rtx_timeout, &start_timer);
 
 			if (start_timer)
-				sctp_transport_reset_timers(transport,
-							    start_timer-1);
+				sctp_transport_reset_timers(transport);
 
 			/* This can happen on COOKIE-ECHO resend.  Only
 			 * one chunk can get bundled with a COOKIE-ECHO.
@@ -1040,7 +1049,7 @@ static int sctp_outq_flush(struct sctp_outq *q, int rtx_timeout)
 			list_add_tail(&chunk->transmitted_list,
 				      &transport->transmitted);
 
-			sctp_transport_reset_timers(transport, 0);
+			sctp_transport_reset_timers(transport);
 
 			q->empty = 0;
 
@@ -1100,32 +1109,6 @@ static void sctp_sack_update_unack_data(struct sctp_association *assoc,
 	assoc->unack_data = unack_data;
 }
 
-/* Return the highest new tsn that is acknowledged by the given SACK chunk. */
-static __u32 sctp_highest_new_tsn(struct sctp_sackhdr *sack,
-				  struct sctp_association *asoc)
-{
-	struct sctp_transport *transport;
-	struct sctp_chunk *chunk;
-	__u32 highest_new_tsn, tsn;
-	struct list_head *transport_list = &asoc->peer.transport_addr_list;
-
-	highest_new_tsn = ntohl(sack->cum_tsn_ack);
-
-	list_for_each_entry(transport, transport_list, transports) {
-		list_for_each_entry(chunk, &transport->transmitted,
-				transmitted_list) {
-			tsn = ntohl(chunk->subh.data_hdr->tsn);
-
-			if (!chunk->tsn_gap_acked &&
-			    TSN_lt(highest_new_tsn, tsn) &&
-			    sctp_acked(sack, tsn))
-				highest_new_tsn = tsn;
-		}
-	}
-
-	return highest_new_tsn;
-}
-
 /* This is where we REALLY process a SACK.
  *
  * Process the SACK against the outqueue.  Mostly, this just frees
@@ -1145,6 +1128,7 @@ int sctp_outq_sack(struct sctp_outq *q, struct sctp_sackhdr *sack)
 	struct sctp_transport *primary = asoc->peer.primary_path;
 	int count_of_newacks = 0;
 	int gap_ack_blocks;
+	u8 accum_moved = 0;
 
 	/* Grab the association's destination address list. */
 	transport_list = &asoc->peer.transport_addr_list;
@@ -1193,18 +1177,15 @@ int sctp_outq_sack(struct sctp_outq *q, struct sctp_sackhdr *sack)
 	if (gap_ack_blocks)
 		highest_tsn += ntohs(frags[gap_ack_blocks - 1].gab.end);
 
-	if (TSN_lt(asoc->highest_sacked, highest_tsn)) {
-		highest_new_tsn = highest_tsn;
+	if (TSN_lt(asoc->highest_sacked, highest_tsn))
 		asoc->highest_sacked = highest_tsn;
-	} else {
-		highest_new_tsn = sctp_highest_new_tsn(sack, asoc);
-	}
 
+	highest_new_tsn = sack_ctsn;
 
 	/* Run through the retransmit queue.  Credit bytes received
 	 * and free those chunks that we can.
 	 */
-	sctp_check_transmitted(q, &q->retransmit, NULL, sack, highest_new_tsn);
+	sctp_check_transmitted(q, &q->retransmit, NULL, sack, &highest_new_tsn);
 
 	/* Run through the transmitted queue.
 	 * Credit bytes received and free those chunks which we can.
@@ -1213,7 +1194,7 @@ int sctp_outq_sack(struct sctp_outq *q, struct sctp_sackhdr *sack)
 	 */
 	list_for_each_entry(transport, transport_list, transports) {
 		sctp_check_transmitted(q, &transport->transmitted,
-				       transport, sack, highest_new_tsn);
+				       transport, sack, &highest_new_tsn);
 		/*
 		 * SFR-CACC algorithm:
 		 * C) Let count_of_newacks be the number of
@@ -1223,16 +1204,22 @@ int sctp_outq_sack(struct sctp_outq *q, struct sctp_sackhdr *sack)
 			count_of_newacks ++;
 	}
 
+	/* Move the Cumulative TSN Ack Point if appropriate.  */
+	if (TSN_lt(asoc->ctsn_ack_point, sack_ctsn)) {
+		asoc->ctsn_ack_point = sack_ctsn;
+		accum_moved = 1;
+	}
+
 	if (gap_ack_blocks) {
+
+		if (asoc->fast_recovery && accum_moved)
+			highest_new_tsn = highest_tsn;
+
 		list_for_each_entry(transport, transport_list, transports)
 			sctp_mark_missing(q, &transport->transmitted, transport,
 					  highest_new_tsn, count_of_newacks);
 	}
 
-	/* Move the Cumulative TSN Ack Point if appropriate.  */
-	if (TSN_lt(asoc->ctsn_ack_point, sack_ctsn))
-		asoc->ctsn_ack_point = sack_ctsn;
-
 	/* Update unack_data field in the assoc. */
 	sctp_sack_update_unack_data(asoc, sack);
 
@@ -1315,7 +1302,7 @@ static void sctp_check_transmitted(struct sctp_outq *q,
 				   struct list_head *transmitted_queue,
 				   struct sctp_transport *transport,
 				   struct sctp_sackhdr *sack,
-				   __u32 highest_new_tsn_in_sack)
+				   __u32 *highest_new_tsn_in_sack)
 {
 	struct list_head *lchunk;
 	struct sctp_chunk *tchunk;
@@ -1387,7 +1374,6 @@ static void sctp_check_transmitted(struct sctp_outq *q,
 				 * instance).
 				 */
 				if (!tchunk->tsn_gap_acked &&
-				    !tchunk->resent &&
 				    tchunk->rtt_in_progress) {
 					tchunk->rtt_in_progress = 0;
 					rtt = jiffies - tchunk->sent_at;
@@ -1404,6 +1390,7 @@ static void sctp_check_transmitted(struct sctp_outq *q,
 			 */
 			if (!tchunk->tsn_gap_acked) {
 				tchunk->tsn_gap_acked = 1;
+				*highest_new_tsn_in_sack = tsn;
 				bytes_acked += sctp_data_size(tchunk);
 				if (!tchunk->transport)
 					migrate_bytes += sctp_data_size(tchunk);
@@ -1677,7 +1664,8 @@ static void sctp_mark_missing(struct sctp_outq *q,
 	struct sctp_chunk *chunk;
 	__u32 tsn;
 	char do_fast_retransmit = 0;
-	struct sctp_transport *primary = q->asoc->peer.primary_path;
+	struct sctp_association *asoc = q->asoc;
+	struct sctp_transport *primary = asoc->peer.primary_path;
 
 	list_for_each_entry(chunk, transmitted_queue, transmitted_list) {
 

net/sctp/probe.c

@@ -0,0 +1,214 @@
+/*
+ * sctp_probe - Observe the SCTP flow with kprobes.
+ *
+ * The idea for this came from Werner Almesberger's umlsim
+ * Copyright (C) 2004, Stephen Hemminger <shemminger@osdl.org>
+ *
+ * Modified for SCTP from Stephen Hemminger's code
+ * Copyright (C) 2010, Wei Yongjun <yjwei@cn.fujitsu.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
+ */
+
+#include <linux/kernel.h>
+#include <linux/kprobes.h>
+#include <linux/socket.h>
+#include <linux/sctp.h>
+#include <linux/proc_fs.h>
+#include <linux/vmalloc.h>
+#include <linux/module.h>
+#include <linux/kfifo.h>
+#include <linux/time.h>
+#include <net/net_namespace.h>
+
+#include <net/sctp/sctp.h>
+#include <net/sctp/sm.h>
+
+MODULE_AUTHOR("Wei Yongjun <yjwei@cn.fujitsu.com>");
+MODULE_DESCRIPTION("SCTP snooper");
+MODULE_LICENSE("GPL");
+
+static int port __read_mostly = 0;
+MODULE_PARM_DESC(port, "Port to match (0=all)");
+module_param(port, int, 0);
+
+static int bufsize __read_mostly = 64 * 1024;
+MODULE_PARM_DESC(bufsize, "Log buffer size (default 64k)");
+module_param(bufsize, int, 0);
+
+static int full __read_mostly = 1;
+MODULE_PARM_DESC(full, "Full log (1=every ack packet received,  0=only cwnd changes)");
+module_param(full, int, 0);
+
+static const char procname[] = "sctpprobe";
+
+static struct {
+	struct kfifo	  fifo;
+	spinlock_t	  lock;
+	wait_queue_head_t wait;
+	struct timespec	  tstart;
+} sctpw;
+
+static void printl(const char *fmt, ...)
+{
+	va_list args;
+	int len;
+	char tbuf[256];
+
+	va_start(args, fmt);
+	len = vscnprintf(tbuf, sizeof(tbuf), fmt, args);
+	va_end(args);
+
+	kfifo_in_locked(&sctpw.fifo, tbuf, len, &sctpw.lock);
+	wake_up(&sctpw.wait);
+}
+
+static int sctpprobe_open(struct inode *inode, struct file *file)
+{
+	kfifo_reset(&sctpw.fifo);
+	getnstimeofday(&sctpw.tstart);
+
+	return 0;
+}
+
+static ssize_t sctpprobe_read(struct file *file, char __user *buf,
+			      size_t len, loff_t *ppos)
+{
+	int error = 0, cnt = 0;
+	unsigned char *tbuf;
+
+	if (!buf)
+		return -EINVAL;
+
+	if (len == 0)
+		return 0;
+
+	tbuf = vmalloc(len);
+	if (!tbuf)
+		return -ENOMEM;
+
+	error = wait_event_interruptible(sctpw.wait,
+					 kfifo_len(&sctpw.fifo) != 0);
+	if (error)
+		goto out_free;
+
+	cnt = kfifo_out_locked(&sctpw.fifo, tbuf, len, &sctpw.lock);
+	error = copy_to_user(buf, tbuf, cnt) ? -EFAULT : 0;
+
+out_free:
+	vfree(tbuf);
+
+	return error ? error : cnt;
+}
+
+static const struct file_operations sctpprobe_fops = {
+	.owner	= THIS_MODULE,
+	.open	= sctpprobe_open,
+	.read	= sctpprobe_read,
+};
+
+sctp_disposition_t jsctp_sf_eat_sack(const struct sctp_endpoint *ep,
+				     const struct sctp_association *asoc,
+				     const sctp_subtype_t type,
+				     void *arg,
+				     sctp_cmd_seq_t *commands)
+{
+	struct sctp_transport *sp;
+	static __u32 lcwnd = 0;
+	struct timespec now;
+
+	sp = asoc->peer.primary_path;
+
+	if ((full || sp->cwnd != lcwnd) &&
+	    (!port || asoc->peer.port == port ||
+	     ep->base.bind_addr.port == port)) {
+		lcwnd = sp->cwnd;
+
+		getnstimeofday(&now);
+		now = timespec_sub(now, sctpw.tstart);
+
+		printl("%lu.%06lu ", (unsigned long) now.tv_sec,
+		       (unsigned long) now.tv_nsec / NSEC_PER_USEC);
+
+		printl("%p %5d %5d %5d %8d %5d ", asoc,
+		       ep->base.bind_addr.port, asoc->peer.port,
+		       asoc->pathmtu, asoc->peer.rwnd, asoc->unack_data);
+
+		list_for_each_entry(sp, &asoc->peer.transport_addr_list,
+					transports) {
+			if (sp == asoc->peer.primary_path)
+				printl("*");
+
+			if (sp->ipaddr.sa.sa_family == AF_INET)
+				printl("%pI4 ", &sp->ipaddr.v4.sin_addr);
+			else
+				printl("%pI6 ", &sp->ipaddr.v6.sin6_addr);
+
+			printl("%2u %8u %8u %8u %8u %8u ",
+			       sp->state, sp->cwnd, sp->ssthresh,
+			       sp->flight_size, sp->partial_bytes_acked,
+			       sp->pathmtu);
+		}
+		printl("\n");
+	}
+
+	jprobe_return();
+	return 0;
+}
+
+static struct jprobe sctp_recv_probe = {
+	.kp	= {
+		.symbol_name = "sctp_sf_eat_sack_6_2",
+	},
+	.entry	= jsctp_sf_eat_sack,
+};
+
+static __init int sctpprobe_init(void)
+{
+	int ret = -ENOMEM;
+
+	init_waitqueue_head(&sctpw.wait);
+	spin_lock_init(&sctpw.lock);
+	if (kfifo_alloc(&sctpw.fifo, bufsize, GFP_KERNEL))
+		return ret;
+
+	if (!proc_net_fops_create(&init_net, procname, S_IRUSR,
+				  &sctpprobe_fops))
+		goto free_kfifo;
+
+	ret = register_jprobe(&sctp_recv_probe);
+	if (ret)
+		goto remove_proc;
+
+	pr_info("SCTP probe registered (port=%d)\n", port);
+
+	return 0;
+
+remove_proc:
+	proc_net_remove(&init_net, procname);
+free_kfifo:
+	kfifo_free(&sctpw.fifo);
+	return ret;
+}
+
+static __exit void sctpprobe_exit(void)
+{
+	kfifo_free(&sctpw.fifo);
+	proc_net_remove(&init_net, procname);
+	unregister_jprobe(&sctp_recv_probe);
+}
+
+module_init(sctpprobe_init);
+module_exit(sctpprobe_exit);

net/sctp/protocol.c

@@ -474,13 +474,17 @@ static struct dst_entry *sctp_v4_get_dst(struct sctp_association *asoc,
 
 	memset(&fl, 0x0, sizeof(struct flowi));
 	fl.fl4_dst  = daddr->v4.sin_addr.s_addr;
+	fl.fl_ip_dport = daddr->v4.sin_port;
 	fl.proto = IPPROTO_SCTP;
 	if (asoc) {
 		fl.fl4_tos = RT_CONN_FLAGS(asoc->base.sk);
 		fl.oif = asoc->base.sk->sk_bound_dev_if;
+		fl.fl_ip_sport = htons(asoc->base.bind_addr.port);
 	}
-	if (saddr)
+	if (saddr) {
 		fl.fl4_src = saddr->v4.sin_addr.s_addr;
+		fl.fl_ip_sport = saddr->v4.sin_port;
+	}
 
 	SCTP_DEBUG_PRINTK("%s: DST:%pI4, SRC:%pI4 - ",
 			  __func__, &fl.fl4_dst, &fl.fl4_src);
@@ -528,6 +532,7 @@ static struct dst_entry *sctp_v4_get_dst(struct sctp_association *asoc,
 		if ((laddr->state == SCTP_ADDR_SRC) &&
 		    (AF_INET == laddr->a.sa.sa_family)) {
 			fl.fl4_src = laddr->a.v4.sin_addr.s_addr;
+			fl.fl_ip_sport = laddr->a.v4.sin_port;
 			if (!ip_route_output_key(&init_net, &rt, &fl)) {
 				dst = &rt->u.dst;
 				goto out_unlock;

net/sctp/sm_make_chunk.c

@@ -445,10 +445,17 @@ struct sctp_chunk *sctp_make_init_ack(const struct sctp_association *asoc,
 	if (!retval)
 		goto nomem_chunk;
 
-	/* Per the advice in RFC 2960 6.4, send this reply to
-	 * the source of the INIT packet.
+	/* RFC 2960 6.4 Multi-homed SCTP Endpoints
+	 *
+	 * An endpoint SHOULD transmit reply chunks (e.g., SACK,
+	 * HEARTBEAT ACK, * etc.) to the same destination transport
+	 * address from which it received the DATA or control chunk
+	 * to which it is replying.
+	 *
+	 * [INIT ACK back to where the INIT came from.]
 	 */
 	retval->transport = chunk->transport;
+
 	retval->subh.init_hdr =
 		sctp_addto_chunk(retval, sizeof(initack), &initack);
 	retval->param_hdr.v = sctp_addto_chunk(retval, addrs_len, addrs.v);
@@ -487,18 +494,6 @@ struct sctp_chunk *sctp_make_init_ack(const struct sctp_association *asoc,
 	/* We need to remove the const qualifier at this point.  */
 	retval->asoc = (struct sctp_association *) asoc;
 
-	/* RFC 2960 6.4 Multi-homed SCTP Endpoints
-	 *
-	 * An endpoint SHOULD transmit reply chunks (e.g., SACK,
-	 * HEARTBEAT ACK, * etc.) to the same destination transport
-	 * address from which it received the DATA or control chunk
-	 * to which it is replying.
-	 *
-	 * [INIT ACK back to where the INIT came from.]
-	 */
-	if (chunk)
-		retval->transport = chunk->transport;
-
 nomem_chunk:
 	kfree(cookie);
 nomem_cookie:
@@ -1254,7 +1249,6 @@ struct sctp_chunk *sctp_chunkify(struct sk_buff *skb,
 	INIT_LIST_HEAD(&retval->list);
 	retval->skb		= skb;
 	retval->asoc		= (struct sctp_association *)asoc;
-	retval->resent  	= 0;
 	retval->has_tsn		= 0;
 	retval->has_ssn         = 0;
 	retval->rtt_in_progress	= 0;

net/sctp/sm_sideeffect.c

@@ -697,11 +697,15 @@ static void sctp_cmd_setup_t2(sctp_cmd_seq_t *cmds,
 {
 	struct sctp_transport *t;
 
-	t = sctp_assoc_choose_alter_transport(asoc,
+	if (chunk->transport)
+		t = chunk->transport;
+	else {
+		t = sctp_assoc_choose_alter_transport(asoc,
 					      asoc->shutdown_last_sent_to);
+		chunk->transport = t;
+	}
 	asoc->shutdown_last_sent_to = t;
 	asoc->timeouts[SCTP_EVENT_TIMEOUT_T2_SHUTDOWN] = t->rto;
-	chunk->transport = t;
 }
 
 /* Helper function to change the state of an association. */

net/sctp/socket.c

@@ -4384,7 +4384,7 @@ static int sctp_getsockopt_peer_addrs(struct sock *sk, int len,
 				transports) {
 		memcpy(&temp, &from->ipaddr, sizeof(temp));
 		sctp_get_pf_specific(sk->sk_family)->addr_v4map(sp, &temp);
-		addrlen = sctp_get_af_specific(sk->sk_family)->sockaddr_len;
+		addrlen = sctp_get_af_specific(temp.sa.sa_family)->sockaddr_len;
 		if (space_left < addrlen)
 			return -ENOMEM;
 		if (copy_to_user(to, &temp, addrlen))

net/sctp/transport.c

@@ -64,9 +64,6 @@ static struct sctp_transport *sctp_transport_init(struct sctp_transport *peer,
 	/* Copy in the address.  */
 	peer->ipaddr = *addr;
 	peer->af_specific = sctp_get_af_specific(addr->sa.sa_family);
-	peer->asoc = NULL;
-
-	peer->dst = NULL;
 	memset(&peer->saddr, 0, sizeof(union sctp_addr));
 
 	/* From 6.3.1 RTO Calculation:
@@ -76,34 +73,21 @@ static struct sctp_transport *sctp_transport_init(struct sctp_transport *peer,
 	 * parameter 'RTO.Initial'.
 	 */
 	peer->rto = msecs_to_jiffies(sctp_rto_initial);
-	peer->rtt = 0;
-	peer->rttvar = 0;
-	peer->srtt = 0;
-	peer->rto_pending = 0;
-	peer->hb_sent = 0;
-	peer->fast_recovery = 0;
 
 	peer->last_time_heard = jiffies;
 	peer->last_time_ecne_reduced = jiffies;
 
-	peer->init_sent_count = 0;
-
 	peer->param_flags = SPP_HB_DISABLE |
 			    SPP_PMTUD_ENABLE |
 			    SPP_SACKDELAY_ENABLE;
-	peer->hbinterval  = 0;
 
 	/* Initialize the default path max_retrans.  */
 	peer->pathmaxrxt  = sctp_max_retrans_path;
-	peer->error_count = 0;
 
 	INIT_LIST_HEAD(&peer->transmitted);
 	INIT_LIST_HEAD(&peer->send_ready);
 	INIT_LIST_HEAD(&peer->transports);
 
-	peer->T3_rtx_timer.expires = 0;
-	peer->hb_timer.expires = 0;
-
 	setup_timer(&peer->T3_rtx_timer, sctp_generate_t3_rtx_event,
 			(unsigned long)peer);
 	setup_timer(&peer->hb_timer, sctp_generate_heartbeat_event,
@@ -113,15 +97,6 @@ static struct sctp_transport *sctp_transport_init(struct sctp_transport *peer,
 	get_random_bytes(&peer->hb_nonce, sizeof(peer->hb_nonce));
 
 	atomic_set(&peer->refcnt, 1);
-	peer->dead = 0;
-
-	peer->malloced = 0;
-
-	/* Initialize the state information for SFR-CACC */
-	peer->cacc.changeover_active = 0;
-	peer->cacc.cycling_changeover = 0;
-	peer->cacc.next_tsn_at_change = 0;
-	peer->cacc.cacc_saw_newack = 0;
 
 	return peer;
 }
@@ -195,7 +170,7 @@ static void sctp_transport_destroy(struct sctp_transport *transport)
 /* Start T3_rtx timer if it is not already running and update the heartbeat
  * timer.  This routine is called every time a DATA chunk is sent.
  */
-void sctp_transport_reset_timers(struct sctp_transport *transport, int force)
+void sctp_transport_reset_timers(struct sctp_transport *transport)
 {
 	/* RFC 2960 6.3.2 Retransmission Timer Rules
 	 *
@@ -205,7 +180,7 @@ void sctp_transport_reset_timers(struct sctp_transport *transport, int force)
 	 * address.
 	 */
 
-	if (force || !timer_pending(&transport->T3_rtx_timer))
+	if (!timer_pending(&transport->T3_rtx_timer))
 		if (!mod_timer(&transport->T3_rtx_timer,
 			       jiffies + transport->rto))
 			sctp_transport_hold(transport);
@@ -403,15 +378,16 @@ void sctp_transport_update_rto(struct sctp_transport *tp, __u32 rtt)
 void sctp_transport_raise_cwnd(struct sctp_transport *transport,
 			       __u32 sack_ctsn, __u32 bytes_acked)
 {
+	struct sctp_association *asoc = transport->asoc;
 	__u32 cwnd, ssthresh, flight_size, pba, pmtu;
 
 	cwnd = transport->cwnd;
 	flight_size = transport->flight_size;
 
 	/* See if we need to exit Fast Recovery first */
-	if (transport->fast_recovery &&
-	    TSN_lte(transport->fast_recovery_exit, sack_ctsn))
-		transport->fast_recovery = 0;
+	if (asoc->fast_recovery &&
+	    TSN_lte(asoc->fast_recovery_exit, sack_ctsn))
+		asoc->fast_recovery = 0;
 
 	/* The appropriate cwnd increase algorithm is performed if, and only
 	 * if the cumulative TSN whould advanced and the congestion window is
@@ -440,7 +416,7 @@ void sctp_transport_raise_cwnd(struct sctp_transport *transport,
 		 *    2) the destination's path MTU.  This upper bound protects
 		 *    against the ACK-Splitting attack outlined in [SAVAGE99].
 		 */
-		if (transport->fast_recovery)
+		if (asoc->fast_recovery)
 			return;
 
 		if (bytes_acked > pmtu)
@@ -491,6 +467,8 @@ void sctp_transport_raise_cwnd(struct sctp_transport *transport,
 void sctp_transport_lower_cwnd(struct sctp_transport *transport,
 			       sctp_lower_cwnd_t reason)
 {
+	struct sctp_association *asoc = transport->asoc;
+
 	switch (reason) {
 	case SCTP_LOWER_CWND_T3_RTX:
 		/* RFC 2960 Section 7.2.3, sctpimpguide
@@ -501,11 +479,11 @@ void sctp_transport_lower_cwnd(struct sctp_transport *transport,
 		 *      partial_bytes_acked = 0
 		 */
 		transport->ssthresh = max(transport->cwnd/2,
-					  4*transport->asoc->pathmtu);
-		transport->cwnd = transport->asoc->pathmtu;
+					  4*asoc->pathmtu);
+		transport->cwnd = asoc->pathmtu;
 
-		/* T3-rtx also clears fast recovery on the transport */
-		transport->fast_recovery = 0;
+		/* T3-rtx also clears fast recovery */
+		asoc->fast_recovery = 0;
 		break;
 
 	case SCTP_LOWER_CWND_FAST_RTX:
@@ -521,15 +499,15 @@ void sctp_transport_lower_cwnd(struct sctp_transport *transport,
 		 *      cwnd = ssthresh
 		 *      partial_bytes_acked = 0
 		 */
-		if (transport->fast_recovery)
+		if (asoc->fast_recovery)
 			return;
 
 		/* Mark Fast recovery */
-		transport->fast_recovery = 1;
-		transport->fast_recovery_exit = transport->asoc->next_tsn - 1;
+		asoc->fast_recovery = 1;
+		asoc->fast_recovery_exit = asoc->next_tsn - 1;
 
 		transport->ssthresh = max(transport->cwnd/2,
-					  4*transport->asoc->pathmtu);
+					  4*asoc->pathmtu);
 		transport->cwnd = transport->ssthresh;
 		break;
 
@@ -549,7 +527,7 @@ void sctp_transport_lower_cwnd(struct sctp_transport *transport,
 		if (time_after(jiffies, transport->last_time_ecne_reduced +
 					transport->rtt)) {
 			transport->ssthresh = max(transport->cwnd/2,
-						  4*transport->asoc->pathmtu);
+						  4*asoc->pathmtu);
 			transport->cwnd = transport->ssthresh;
 			transport->last_time_ecne_reduced = jiffies;
 		}
@@ -565,7 +543,7 @@ void sctp_transport_lower_cwnd(struct sctp_transport *transport,
 		 * interval.
 		 */
 		transport->cwnd = max(transport->cwnd/2,
-					 4*transport->asoc->pathmtu);
+					 4*asoc->pathmtu);
 		break;
 	}
 
@@ -650,7 +628,6 @@ void sctp_transport_reset(struct sctp_transport *t)
 	t->error_count = 0;
 	t->rto_pending = 0;
 	t->hb_sent = 0;
-	t->fast_recovery = 0;
 
 	/* Initialize the state information for SFR-CACC */
 	t->cacc.changeover_active = 0;