[SCTP] Partial Reliability Extension support.

include/linux/sctp.h

 /* SCTP kernel reference Implementation
- * (C) Copyright IBM Corp. 2001, 2003
+ * (C) Copyright IBM Corp. 2001, 2004
  * Copyright (c) 1999-2000 Cisco, Inc.
  * Copyright (c) 1999-2001 Motorola, Inc.
  * Copyright (c) 2001 Intel Corp.
@@ -93,6 +93,9 @@ typedef enum {
         SCTP_CID_ECN_CWR		= 13,
         SCTP_CID_SHUTDOWN_COMPLETE	= 14,
 
+	/* PR-SCTP Sec 3.2 */
+	SCTP_CID_FWD_TSN		= 0xC0,
+
 	/* Use hex, as defined in ADDIP sec. 3.1 */
 	SCTP_CID_ASCONF			= 0xC1,
 	SCTP_CID_ASCONF_ACK		= 0x80,
@@ -168,6 +171,9 @@ typedef enum {
 	SCTP_PARAM_SUPPORTED_ADDRESS_TYPES	= __constant_htons(12),
 	SCTP_PARAM_ECN_CAPABLE			= __constant_htons(0x8000),
 
+	/* PR-SCTP Sec 3.1 */
+	SCTP_PARAM_FWD_TSN_SUPPORT	= __constant_htons(0xc000),
+
 	/* Add-IP Extension. Section 3.2 */
 	SCTP_PARAM_ADD_IP		= __constant_htons(0xc001),
 	SCTP_PARAM_DEL_IP		= __constant_htons(0xc002),
@@ -472,9 +478,67 @@ typedef struct sctp_cwr_chunk {
 	sctp_cwrhdr_t cwr_hdr;
 } __attribute__((packed)) sctp_cwr_chunk_t;
 
-/*
- * ADDIP Section 3.1 New Chunk Types
+/* PR-SCTP
+ * 3.2 Forward Cumulative TSN Chunk Definition (FORWARD TSN)
+ *
+ * Forward Cumulative TSN chunk has the following format:
+ *
+ *        0                   1                   2                   3
+ *        0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+ *      +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
+ *      |   Type = 192  |  Flags = 0x00 |        Length = Variable      |
+ *      +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
+ *      |                      New Cumulative TSN                       |
+ *      +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
+ *      |         Stream-1              |       Stream Sequence-1       |
+ *      +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
+ *      \                                                               /
+ *      /                                                               \
+ *      +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
+ *      |         Stream-N              |       Stream Sequence-N       |
+ *      +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
+ *
+ *      Chunk Flags:
+ *
+ *        Set to all zeros on transmit and ignored on receipt.
+ *
+ *      New Cumulative TSN: 32 bit u_int
+ *
+ *       This indicates the new cumulative TSN to the data receiver. Upon
+ *       the reception of this value, the data receiver MUST consider
+ *       any missing TSNs earlier than or equal to this value as received
+ *       and stop reporting them as gaps in any subsequent SACKs.
+ *
+ *      Stream-N: 16 bit u_int
+ *
+ *       This field holds a stream number that was skipped by this
+ *       FWD-TSN.
+ *
+ *      Stream Sequence-N: 16 bit u_int
+ *       This field holds the sequence number associated with the stream
+ *       that was skipped. The stream sequence field holds the largest stream
+ *       sequence number in this stream being skipped.  The receiver of
+ *       the FWD-TSN's can use the Stream-N and Stream Sequence-N fields
+ *       to enable delivery of any stranded TSN's that remain on the stream
+ *       re-ordering queues. This field MUST NOT report TSN's corresponding
+ *       to DATA chunk that are marked as unordered. For ordered DATA
+ *       chunks this field MUST be filled in.
  */
+struct sctp_fwdtsn_skip {
+	__u16 stream;
+	__u16 ssn;
+} __attribute__((packed));
+
+struct sctp_fwdtsn_hdr {
+	__u32 new_cum_tsn;
+	struct sctp_fwdtsn_skip skip[0];
+} __attribute((packed));
+
+struct sctp_fwdtsn_chunk {
+	struct sctp_chunkhdr chunk_hdr;
+	struct sctp_fwdtsn_hdr fwdtsn_hdr;
+} __attribute((packed));
+
 
 /* ADDIP
  * Section 3.1.1 Address Configuration Change Chunk (ASCONF)

include/linux/sysctl.h

@@ -599,8 +599,7 @@ enum {
 	NET_SCTP_PRESERVE_ENABLE         = 11,
 	NET_SCTP_MAX_BURST               = 12,
 	NET_SCTP_ADDIP_ENABLE		 = 13,
-	NET_SCTP_RMEM			 = 14,
-	NET_SCTP_WMEM			 = 15,
+	NET_SCTP_PRSCTP_ENABLE		 = 14,
 };
 
 /* /proc/sys/net/bridge */

include/net/sctp/command.h

 /* SCTP kernel reference Implementation
- * (C) Copyright IBM Corp. 2001, 2003
+ * (C) Copyright IBM Corp. 2001, 2004
  * Copyright (C) 1999-2001 Cisco, Motorola
  *
  * This file is part of the SCTP kernel reference Implementation
@@ -29,6 +29,7 @@
  * La Monte H.P. Yarroll <piggy@acm.org>
  * Karl Knutson <karl@athena.chicago.il.us>
  * Ardelle Fan <ardelle.fan@intel.com>
+ * Sridhar Samudrala <sri@us.ibm.com>
  *
  * Any bugs reported given to us we will try to fix... any fixes shared will
  * be incorporated into the next SCTP release.
@@ -90,6 +91,8 @@ typedef enum {
 	SCTP_CMD_RENEGE,         /* Renege data on an association. */
 	SCTP_CMD_SETUP_T4,	 /* ADDIP, setup T4 RTO timer parms. */
 	SCTP_CMD_PROCESS_OPERR,  /* Process an ERROR chunk. */
+	SCTP_CMD_REPORT_FWDTSN,	 /* Report new cumulative TSN Ack. */
+	SCTP_CMD_PROCESS_FWDTSN, /* Skips were reported, so process further. */
 	SCTP_CMD_LAST
 } sctp_verb_t;
 

include/net/sctp/constants.h

 /* SCTP kernel reference Implementation
- * (C) Copyright IBM Corp. 2001, 2003
+ * (C) Copyright IBM Corp. 2001, 2004
  * Copyright (c) 1999-2000 Cisco, Inc.
  * Copyright (c) 1999-2001 Motorola, Inc.
  * Copyright (c) 2001 Intel Corp.
@@ -68,6 +68,8 @@ enum { SCTP_DEFAULT_INSTREAMS = SCTP_MAX_STREAM };
 
 #define SCTP_NUM_ADDIP_CHUNK_TYPES	2
 
+#define SCTP_NUM_PRSCTP_CHUNK_TYPES	1
+
 /* These are the different flavours of event.  */
 typedef enum {
 

include/net/sctp/sctp.h

@@ -475,6 +475,14 @@ for (err = (sctp_errhdr_t *)((void *)chunk_hdr + \
      err = (sctp_errhdr_t *)((void *)err + \
 	    WORD_ROUND(ntohs(err->length))))
 
+#define sctp_walk_fwdtsn(pos, chunk)\
+_sctp_walk_fwdtsn((pos), (chunk), ntohs((chunk)->chunk_hdr->length) - sizeof(struct sctp_fwdtsn_chunk))
+
+#define _sctp_walk_fwdtsn(pos, chunk, end)\
+for (pos = chunk->subh.fwdtsn_hdr->skip;\
+     (void *)pos <= (void *)chunk->subh.fwdtsn_hdr->skip + end - sizeof(struct sctp_fwdtsn_skip);\
+     pos++)
+
 /* Round an int up to the next multiple of 4.  */
 #define WORD_ROUND(s) (((s)+3)&~3)
 

include/net/sctp/sm.h

 /* SCTP kernel reference Implementation
- * (C) Copyright IBM Corp. 2001, 2003
+ * (C) Copyright IBM Corp. 2001, 2004
  * Copyright (c) 1999-2000 Cisco, Inc.
  * Copyright (c) 1999-2001 Motorola, Inc.
  * Copyright (c) 2001 Intel Corp.
@@ -141,6 +141,9 @@ sctp_state_fn_t sctp_sf_cookie_echoed_err;
 sctp_state_fn_t sctp_sf_do_5_2_6_stale;
 sctp_state_fn_t sctp_sf_do_asconf;
 sctp_state_fn_t sctp_sf_do_asconf_ack;
+sctp_state_fn_t sctp_sf_do_9_2_reshutack;
+sctp_state_fn_t sctp_sf_eat_fwd_tsn;
+sctp_state_fn_t sctp_sf_eat_fwd_tsn_fast;
 
 /* Prototypes for primitive event state functions.  */
 sctp_state_fn_t sctp_sf_do_prm_asoc;
@@ -170,25 +173,6 @@ sctp_state_fn_t sctp_sf_do_6_3_3_rtx;
 sctp_state_fn_t sctp_sf_do_6_2_sack;
 sctp_state_fn_t sctp_sf_autoclose_timer_expire;
 
-
-/* These are state functions which are either obsolete or not in use yet.
- * If any of these functions needs to be revived, it should be renamed with
- * the "sctp_sf_xxx" prefix, and be moved to the above prototype groups.
- */
-
-/* Prototypes for chunk state functions.  Not in use. */
-sctp_state_fn_t sctp_sf_do_9_2_reshutack;
-sctp_state_fn_t sctp_sf_do_9_2_reshut;
-sctp_state_fn_t sctp_sf_do_9_2_shutack;
-
-/* Prototypes for timeout event state functions.  Not in use. */
-sctp_state_fn_t sctp_do_4_2_reinit;
-sctp_state_fn_t sctp_do_4_3_reecho;
-sctp_state_fn_t sctp_do_9_2_reshut;
-sctp_state_fn_t sctp_do_9_2_reshutack;
-sctp_state_fn_t sctp_do_8_3_hb_err;
-sctp_state_fn_t sctp_heartoff;
-
 /* Prototypes for utility support functions.  */
 __u8 sctp_get_chunk_type(struct sctp_chunk *chunk);
 const sctp_sm_table_entry_t *sctp_sm_lookup_event(sctp_event_t,
@@ -277,6 +261,9 @@ struct sctp_chunk *sctp_process_asconf(struct sctp_association *asoc,
 				       struct sctp_chunk *asconf);
 int sctp_process_asconf_ack(struct sctp_association *asoc,
 			    struct sctp_chunk *asconf_ack);
+struct sctp_chunk *sctp_make_fwdtsn(const struct sctp_association *asoc,
+				    __u32 new_cum_tsn, size_t nstreams,
+				    struct sctp_fwdtsn_skip *skiplist);
 
 void sctp_chunk_assign_tsn(struct sctp_chunk *);
 void sctp_chunk_assign_ssn(struct sctp_chunk *);

include/net/sctp/structs.h

@@ -193,6 +193,9 @@ extern struct sctp_globals {
 	
 	/* Flag to indicate if addip is enabled. */
 	int addip_enable;
+
+	/* Flag to indicate if PR-SCTP is enabled. */
+	int prsctp_enable;
 } sctp_globals;
 
 #define sctp_rto_initial		(sctp_globals.rto_initial)
@@ -221,6 +224,7 @@ extern struct sctp_globals {
 #define sctp_local_addr_list		(sctp_globals.local_addr_list)
 #define sctp_local_addr_lock		(sctp_globals.local_addr_lock)
 #define sctp_addip_enable		(sctp_globals.addip_enable)
+#define sctp_prsctp_enable		(sctp_globals.prsctp_enable)
 
 /* SCTP Socket type: UDP or TCP style. */
 typedef enum {
@@ -317,6 +321,8 @@ struct sctp_cookie {
 	/* This holds the originating address of the INIT packet.  */
 	union sctp_addr peer_addr;
 
+	__u8 prsctp_capable;
+
 	/* This is a shim for my peer's INIT packet, followed by
 	 * a copy of the raw address list of the association.
 	 * The length of the raw address list is saved in the
@@ -413,6 +419,13 @@ static inline __u16 sctp_ssn_next(struct sctp_stream *stream, __u16 id)
 	return stream->ssn[id]++;
 }
 
+/* Skip over this ssn and all below. */
+static inline void sctp_ssn_skip(struct sctp_stream *stream, __u16 id, 
+				 __u16 ssn)
+{
+	stream->ssn[id] = ssn+1;
+}
+              
 /*
  * Pointers to address related SCTP functions.
  * (i.e. things that depend on the address family.)
@@ -514,8 +527,8 @@ struct sctp_datamsg {
 	/* Did the messenge fail to send? */
 	int send_error;
 	char send_failed;
-	/* Control whether fragments from this message can expire. */
-	char can_expire;
+	/* Control whether chunks from this message can be abandoned. */
+	char can_abandon;
 };
 
 struct sctp_datamsg *sctp_datamsg_from_user(struct sctp_association *,
@@ -527,8 +540,8 @@ void sctp_datamsg_hold(struct sctp_datamsg *);
 void sctp_datamsg_free(struct sctp_datamsg *);
 void sctp_datamsg_track(struct sctp_chunk *);
 void sctp_datamsg_assign(struct sctp_datamsg *, struct sctp_chunk *);
-void sctp_datamsg_fail(struct sctp_chunk *, int error);
-int sctp_datamsg_expires(struct sctp_chunk *);
+void sctp_chunk_fail(struct sctp_chunk *, int error);
+int sctp_chunk_abandoned(struct sctp_chunk *);
 
 
 /* RFC2960 1.4 Key Terms
@@ -583,6 +596,7 @@ struct sctp_chunk {
 		struct sctp_cwrhdr *ecn_cwr_hdr;
 		struct sctp_errhdr *err_hdr;
 		struct sctp_addiphdr *addip_hdr;
+		struct sctp_fwdtsn_hdr *fwdtsn_hdr;
 	} subh;
 
 	__u8 *chunk_end;
@@ -890,7 +904,7 @@ struct sctp_transport {
 		/* A flag which indicates the occurrence of a changeover */
 		char changeover_active;
 
-		/* A glag which indicates whether the change of primary is
+		/* A flag which indicates whether the change of primary is
 		 * the first switch to this destination address during an
 		 * active switch.
 		 */
@@ -993,6 +1007,11 @@ struct sctp_outq {
 	 */
 	struct list_head retransmit;
 
+	/* Put chunks on this list to save them for FWD TSN processing as
+	 * they were abandoned.
+	 */
+	struct list_head abandoned;
+
 	/* How many unackd bytes do we have in-flight?	*/
 	__u32 outstanding_bytes;
 
@@ -1356,16 +1375,25 @@ struct sctp_association {
 		struct sctp_tsnmap tsn_map;
 		__u8 _map[sctp_tsnmap_storage_size(SCTP_TSN_MAP_SIZE)];
 
-		/* Do we need to sack the peer? */
-		__u8	sack_needed;
+		/* Ack State   : This flag indicates if the next received
+		 *             : packet is to be responded to with a
+		 *             : SACK. This is initializedto 0.  When a packet
+		 *             : is received it is incremented. If this value
+		 *             : reaches 2 or more, a SACK is sent and the
+		 *             : value is reset to 0. Note: This is used only
+		 *             : when no DATA chunks are received out of
+		 *             : order.  When DATA chunks are out of order,
+		 *             : SACK's are not delayed (see Section 6).
+		 */
+		__u8    sack_needed;     /* Do we need to sack the peer? */
+
 		/* These are capabilities which our peer advertised.  */
 		__u8	ecn_capable;	 /* Can peer do ECN? */
 		__u8	ipv4_address;	 /* Peer understands IPv4 addresses? */
 		__u8	ipv6_address;	 /* Peer understands IPv6 addresses? */
 		__u8	hostname_address;/* Peer understands DNS addresses? */
-
-		/* Does peer support ADDIP? */
-		__u8    asconf_capable;
+		__u8    asconf_capable;  /* Does peer support ADDIP? */
+		__u8    prsctp_capable;  /* Can peer do PR-SCTP? */
 
 		/* This mask is used to disable sending the ASCONF chunk
 		 * with specified parameter to peer.
@@ -1458,6 +1486,9 @@ struct sctp_association {
 
 	__u32 ctsn_ack_point;
 
+	/* PR-SCTP Advanced.Peer.Ack.Point */
+	__u32 adv_peer_ack_point;
+
 	/* Highest TSN that is acknowledged by incoming SACKs. */
 	__u32 highest_sacked;
 
@@ -1498,19 +1529,7 @@ struct sctp_association {
 	/* The message size at which SCTP fragmentation will occur. */
 	__u32 frag_point;
 
-	/* Ack State   : This flag indicates if the next received
-	 *	       : packet is to be responded to with a
-	 *	       : SACK. This is initializedto 0.	 When a packet
-	 *	       : is received it is incremented. If this value
-	 *	       : reaches 2 or more, a SACK is sent and the
-	 *	       : value is reset to 0. Note: This is used only
-	 *	       : when no DATA chunks are received out of
-	 *	       : order.	 When DATA chunks are out of order,
-	 *	       : SACK's are not delayed (see Section 6).
-	 */
-	/* Do we need to send an ack?
-	 * When counters[SctpCounterAckState] is above 1 we do!
-	 */
+	/* Currently only one counter is used to count INIT errors. */
 	int counters[SCTP_NUMBER_COUNTERS];
 
 	/* Default send parameters. */

include/net/sctp/tsnmap.h

 /* SCTP kernel reference Implementation
+ * (C) Copyright IBM Corp. 2001, 2004
  * Copyright (c) 1999-2000 Cisco, Inc.
  * Copyright (c) 1999-2001 Motorola, Inc.
- * Copyright (c) 2001-2003 International Business Machines, Corp.
  * Copyright (c) 2001 Intel Corp.
  *
  * This file is part of the SCTP kernel reference Implementation
@@ -37,6 +37,7 @@
  *   Jon Grimm             <jgrimm@us.ibm.com>
  *   La Monte H.P. Yarroll <piggy@acm.org>
  *   Karl Knutson          <karl@athena.chicago.il.us>
+ *   Sridhar Samudrala     <sri@us.ibm.com>
  *
  * Any bugs reported given to us we will try to fix... any fixes shared will
  * be incorporated into the next SCTP release.
@@ -145,6 +146,9 @@ int sctp_tsnmap_check(const struct sctp_tsnmap *, __u32 tsn);
 /* Mark this TSN as seen.  */
 void sctp_tsnmap_mark(struct sctp_tsnmap *, __u32 tsn);
 
+/* Mark this TSN and all lower as seen. */
+void sctp_tsnmap_skip(struct sctp_tsnmap *map, __u32 tsn);
+
 /* Retrieve the Cumulative TSN ACK Point.  */
 static inline __u32 sctp_tsnmap_get_ctsn(const struct sctp_tsnmap *map)
 {

include/net/sctp/ulpqueue.h

 /* SCTP kernel reference Implementation
+ * (C) Copyright IBM Corp. 2001, 2004
  * Copyright (c) 1999-2000 Cisco, Inc.
  * Copyright (c) 1999-2001 Motorola, Inc.
- * Copyright (c) 2001-2003 International Business Machines, Corp.
  * Copyright (c) 2001 Intel Corp.
  * Copyright (c) 2001 Nokia, Inc.
  * Copyright (c) 2001 La Monte H.P. Yarroll
@@ -38,6 +38,7 @@
  * Written or modified by:
  *   Jon Grimm             <jgrimm@us.ibm.com>
  *   La Monte H.P. Yarroll <piggy@acm.org>
+ *   Sridhar Samudrala     <sri@us.ibm.com>
  *
  * Any bugs reported given to us we will try to fix... any fixes shared will
  * be incorporated into the next SCTP release.
@@ -79,6 +80,9 @@ void sctp_ulpq_abort_pd(struct sctp_ulpq *, int);
 /* Clear the partial data delivery condition on this socket. */
 int sctp_clear_pd(struct sock *sk);
 
+/* Skip over an SSN. */
+void sctp_ulpq_skip(struct sctp_ulpq *ulpq, __u16 sid, __u16 ssn);
+
 #endif /* __sctp_ulpqueue_h__ */
 
 

net/sctp/associola.c

@@ -210,6 +210,7 @@ struct sctp_association *sctp_association_init(struct sctp_association *asoc,
 	asoc->next_tsn = asoc->c.initial_tsn;
 
 	asoc->ctsn_ack_point = asoc->next_tsn - 1;
+	asoc->adv_peer_ack_point = asoc->ctsn_ack_point;
 	asoc->highest_sacked = asoc->ctsn_ack_point;
 	asoc->last_cwr_tsn = asoc->ctsn_ack_point;
 	asoc->unack_data = 0;
@@ -959,6 +960,7 @@ void sctp_assoc_update(struct sctp_association *asoc,
 
 		asoc->next_tsn = new->next_tsn;
 		asoc->ctsn_ack_point = new->ctsn_ack_point;
+		asoc->adv_peer_ack_point = new->adv_peer_ack_point;
 
 		/* Reinitialize SSN for both local streams
 		 * and peer's streams.
@@ -967,6 +969,7 @@ void sctp_assoc_update(struct sctp_association *asoc,
 
 	} else {
 		asoc->ctsn_ack_point = asoc->next_tsn - 1;
+		asoc->adv_peer_ack_point = asoc->ctsn_ack_point;
 		if (!asoc->ssnmap) {
 			/* Move the ssnmap. */
 			asoc->ssnmap = new->ssnmap;

net/sctp/chunk.c

 /* SCTP kernel reference Implementation
- * Copyright (c) 2003 International Business Machines Corp.
+ * (C) Copyright IBM Corp. 2003, 2004
  *
  * This file is part of the SCTP kernel reference Implementation
  *
@@ -31,6 +31,7 @@
  *
  * Written or modified by:
  *    Jon Grimm             <jgrimm@us.ibm.com>
+ *    Sridhar Samudrala     <sri@us.ibm.com>
  *
  * Any bugs reported given to us we will try to fix... any fixes shared will
  * be incorporated into the next SCTP release.
@@ -55,7 +56,8 @@ void sctp_datamsg_init(struct sctp_datamsg *msg)
 	atomic_set(&msg->refcnt, 1);
 	msg->send_failed = 0;
 	msg->send_error = 0;
-	msg->can_expire = 0;
+	msg->can_abandon = 0;
+	msg->expires_at = 0;
 	INIT_LIST_HEAD(&msg->chunks);
 }
 
@@ -182,14 +184,12 @@ struct sctp_datamsg *sctp_datamsg_from_user(struct sctp_association *asoc,
 	 * have the same expiration.
 	 */
 	if (sinfo->sinfo_timetolive) {
-		struct timeval tv;
-		__u32 ttl = sinfo->sinfo_timetolive;
-
 		/* sinfo_timetolive is in milliseconds */
-		tv.tv_sec = ttl / 1000;
-		tv.tv_usec = ttl % 1000 * 1000;
-		msg->expires_at = jiffies + timeval_to_jiffies(&tv);
-		msg->can_expire = 1;
+		msg->expires_at = jiffies +
+				    MSECS_TO_JIFFIES(sinfo->sinfo_timetolive);
+		msg->can_abandon = 1;
+		SCTP_DEBUG_PRINTK("%s: msg:%p expires_at: %ld jiffies:%ld\n",
+				  __FUNCTION__, msg, msg->expires_at, jiffies);
 	}
 
 	max = asoc->frag_point;
@@ -288,14 +288,11 @@ struct sctp_datamsg *sctp_datamsg_from_user(struct sctp_association *asoc,
 }
 
 /* Check whether this message has expired. */
-int sctp_datamsg_expires(struct sctp_chunk *chunk)
+int sctp_chunk_abandoned(struct sctp_chunk *chunk)
 {
 	struct sctp_datamsg *msg = chunk->msg;
 
-	/* FIXME: When PR-SCTP is supported we can make this
-	 * check more lenient.
-	 */
-	if (!msg->can_expire)
+	if (!msg->can_abandon)
 		return 0;
 
 	if (time_after(jiffies, msg->expires_at))
@@ -305,7 +302,7 @@ int sctp_datamsg_expires(struct sctp_chunk *chunk)
 }
 
 /* This chunk (and consequently entire message) has failed in its sending. */
-void sctp_datamsg_fail(struct sctp_chunk *chunk, int error)
+void sctp_chunk_fail(struct sctp_chunk *chunk, int error)
 {
 	chunk->msg->send_failed = 1;
 	chunk->msg->send_error = error;

net/sctp/debug.c

 /* SCTP kernel reference Implementation
+ * (C) Copyright IBM Corp. 2001, 2004
  * Copyright (c) 1999-2000 Cisco, Inc.
  * Copyright (c) 1999-2001 Motorola, Inc.
  * Copyright (c) 2001 Intel Corp.
- * Copyright (c) 2001 International Business Machines Corp.
  * 
  * This file is part of the SCTP kernel reference Implementation
  * 
@@ -43,6 +43,7 @@
  *    Xingang Guo           <xingang.guo@intel.com>
  *    Jon Grimm             <jgrimm@us.ibm.com>
  *    Daisy Chang	    <daisyc@us.ibm.com>
+ *    Sridhar Samudrala	    <sri@us.ibm.com>
  * 
  * Any bugs reported given to us we will try to fix... any fixes shared will
  * be incorporated into the next SCTP release.
@@ -88,6 +89,9 @@ const char *sctp_cname(const sctp_subtype_t cid)
 	case SCTP_CID_ASCONF_ACK:
 		return "ASCONF_ACK";
 
+	case SCTP_CID_FWD_TSN:
+		return "FWD_TSN";
+
 	default:
 		return "unknown chunk";
 	};

net/sctp/output.c

@@ -278,6 +278,7 @@ sctp_xmit_t sctp_packet_append_chunk(struct sctp_packet *packet,
 	/* It is OK to send this chunk.  */
 	__skb_queue_tail(&packet->chunks, (struct sk_buff *)chunk);
 	packet->size += chunk_len;
+	chunk->transport = packet->transport;
 finish:
 	return retval;
 }
@@ -637,7 +638,8 @@ static sctp_xmit_t sctp_packet_append_data(struct sctp_packet *packet,
 
 	asoc->peer.rwnd = rwnd;
 	/* Has been accepted for transmission. */
-	chunk->msg->can_expire = 0;
+	if (!asoc->peer.prsctp_capable)
+		chunk->msg->can_abandon = 0;
 
 finish:
 	return retval;

net/sctp/outqueue.c

 /* SCTP kernel reference Implementation
+ * (C) Copyright IBM Corp. 2001, 2004
  * Copyright (c) 1999-2000 Cisco, Inc.
  * Copyright (c) 1999-2001 Motorola, Inc.
  * Copyright (c) 2001-2003 Intel Corp.
- * Copyright (c) 2001-2003 International Business Machines Corp.
  *
  * This file is part of the SCTP kernel reference Implementation
  *
@@ -69,6 +69,8 @@ static void sctp_mark_missing(struct sctp_outq *q,
 			      __u32 highest_new_tsn,
 			      int count_of_newacks);
 
+static void sctp_generate_fwdtsn(struct sctp_outq *q, __u32 sack_ctsn);
+
 /* Add data to the front of the queue. */
 static inline void sctp_outq_head_data(struct sctp_outq *q,
 					struct sctp_chunk *ch)
@@ -222,6 +224,7 @@ void sctp_outq_init(struct sctp_association *asoc, struct sctp_outq *q)
 	skb_queue_head_init(&q->control);
 	INIT_LIST_HEAD(&q->retransmit);
 	INIT_LIST_HEAD(&q->sacked);
+	INIT_LIST_HEAD(&q->abandoned);
 
 	q->outstanding_bytes = 0;
 	q->empty = 1;
@@ -246,7 +249,7 @@ void sctp_outq_teardown(struct sctp_outq *q)
 			chunk = list_entry(lchunk, struct sctp_chunk,
 					   transmitted_list);
 			/* Mark as part of a failed message. */
-			sctp_datamsg_fail(chunk, q->error);
+			sctp_chunk_fail(chunk, q->error);
 			sctp_chunk_free(chunk);
 		}
 	}
@@ -256,7 +259,7 @@ void sctp_outq_teardown(struct sctp_outq *q)
 		list_del_init(lchunk);
 		chunk = list_entry(lchunk, struct sctp_chunk,
 				   transmitted_list);
-		sctp_datamsg_fail(chunk, q->error);
+		sctp_chunk_fail(chunk, q->error);
 		sctp_chunk_free(chunk);
 	}
 
@@ -265,7 +268,16 @@ void sctp_outq_teardown(struct sctp_outq *q)
 		list_del_init(lchunk);
 		chunk = list_entry(lchunk, struct sctp_chunk,
 				   transmitted_list);
-		sctp_datamsg_fail(chunk, q->error);
+		sctp_chunk_fail(chunk, q->error);
+		sctp_chunk_free(chunk);
+	}
+
+	/* Throw away any chunks that are in the abandoned queue. */
+	list_for_each_safe(lchunk, temp, &q->abandoned) {
+		list_del_init(lchunk);
+		chunk = list_entry(lchunk, struct sctp_chunk,
+				   transmitted_list);
+		sctp_chunk_fail(chunk, q->error);
 		sctp_chunk_free(chunk);
 	}
 
@@ -273,7 +285,7 @@ void sctp_outq_teardown(struct sctp_outq *q)
 	while ((chunk = sctp_outq_dequeue_data(q))) {
 
 		/* Mark as send failure. */
-		sctp_datamsg_fail(chunk, q->error);
+		sctp_chunk_fail(chunk, q->error);
 		sctp_chunk_free(chunk);
 	}
 
@@ -357,32 +369,30 @@ int sctp_outq_tail(struct sctp_outq *q, struct sctp_chunk *chunk)
 	return error;
 }
 
-/* Insert a chunk into the retransmit queue.  Chunks on the retransmit
- * queue are kept in order, based on the TSNs.
+/* Insert a chunk into the sorted list based on the TSNs.  The retransmit list
+ * and the abandoned list are in ascending order.
  */
-void sctp_retransmit_insert(struct list_head *tlchunk, struct sctp_outq *q)
+void sctp_insert_list(struct list_head *head, struct list_head *new)
 {
-	struct list_head *rlchunk;
-	struct sctp_chunk *tchunk, *rchunk;
-	__u32 ttsn, rtsn;
+	struct list_head *pos;
+	struct sctp_chunk *nchunk, *lchunk;
+	__u32 ntsn, ltsn;
 	int done = 0;
 
-	tchunk = list_entry(tlchunk, struct sctp_chunk, transmitted_list);
-	ttsn = ntohl(tchunk->subh.data_hdr->tsn);
+	nchunk = list_entry(new, struct sctp_chunk, transmitted_list);
+	ntsn = ntohl(nchunk->subh.data_hdr->tsn);
 
-	list_for_each(rlchunk, &q->retransmit) {
-		rchunk = list_entry(rlchunk, struct sctp_chunk,
-				    transmitted_list);
-		rtsn = ntohl(rchunk->subh.data_hdr->tsn);
-		if (TSN_lt(ttsn, rtsn)) {
-			list_add(tlchunk, rlchunk->prev);
+	list_for_each(pos, head) {
+		lchunk = list_entry(pos, struct sctp_chunk, transmitted_list);
+		ltsn = ntohl(lchunk->subh.data_hdr->tsn);
+		if (TSN_lt(ntsn, ltsn)) {
+			list_add(new, pos->prev);
 			done = 1;
 			break;
 		}
 	}
-	if (!done) {
-		list_add_tail(tlchunk, &q->retransmit);
-	}
+	if (!done)
+		list_add_tail(new, head); 
 }
 
 /* Mark all the eligible packets on a transport for retransmission.  */
@@ -398,6 +408,13 @@ void sctp_retransmit_mark(struct sctp_outq *q,
 		chunk = list_entry(lchunk, struct sctp_chunk,
 				   transmitted_list);
 
+		/* If the chunk is abandoned, move it to abandoned list. */
+		if (sctp_chunk_abandoned(chunk)) {
+			list_del_init(lchunk);
+			sctp_insert_list(&q->abandoned, lchunk);
+			continue;
+		}
+
 		/* If we are doing retransmission due to a fast retransmit,
 		 * only the chunk's that are marked for fast retransmit
 		 * should be added to the retransmit queue.  If we are doing
@@ -438,10 +455,10 @@ void sctp_retransmit_mark(struct sctp_outq *q,
 			}
 
 			/* Move the chunk to the retransmit queue. The chunks
-			 * on the retransmit queue is always kept in order.
+			 * on the retransmit queue are always kept in order.
 			 */
 			list_del_init(lchunk);
-			sctp_retransmit_insert(lchunk, q);
+			sctp_insert_list(&q->retransmit, lchunk);
 		}
 	}
 
@@ -484,6 +501,12 @@ void sctp_retransmit(struct sctp_outq *q, struct sctp_transport *transport,
 
 	sctp_retransmit_mark(q, transport, fast_retransmit);
 
+	/* PR-SCTP A5) Any time the T3-rtx timer expires, on any destination,
+	 * the sender SHOULD try to advance the "Advanced.Peer.Ack.Point" by
+	 * following the procedures outlined in C1 - C5.
+	 */
+	sctp_generate_fwdtsn(q, q->asoc->ctsn_ack_point);
+
 	error = sctp_outq_flush(q, /* rtx_timeout */ 1);
 
 	if (error)
@@ -717,6 +740,7 @@ int sctp_outq_flush(struct sctp_outq *q, int rtx_timeout)
 		case SCTP_CID_ECN_CWR:
 		case SCTP_CID_ASCONF:
 		case SCTP_CID_ASCONF_ACK:
+		case SCTP_CID_FWD_TSN:
 			sctp_packet_transmit_chunk(packet, chunk);
 			break;
 
@@ -795,15 +819,15 @@ int sctp_outq_flush(struct sctp_outq *q, int rtx_timeout)
 			if (chunk->sinfo.sinfo_stream >=
 			    asoc->c.sinit_num_ostreams) {
 
-				/* Mark as s failed send. */
-				sctp_datamsg_fail(chunk, SCTP_ERROR_INV_STRM);
+				/* Mark as failed send. */
+				sctp_chunk_fail(chunk, SCTP_ERROR_INV_STRM);
 				sctp_chunk_free(chunk);
 				continue;
 			}
 
 			/* Has this chunk expired? */
-			if (sctp_datamsg_expires(chunk)) {
-				sctp_datamsg_fail(chunk, 0);
+			if (sctp_chunk_abandoned(chunk)) {
+				sctp_chunk_fail(chunk, 0);
 				sctp_chunk_free(chunk);
 				continue;
 			}
@@ -980,7 +1004,7 @@ int sctp_outq_sack(struct sctp_outq *q, struct sctp_sackhdr *sack)
 {
 	struct sctp_association *asoc = q->asoc;
 	struct sctp_transport *transport;
-	struct sctp_chunk *tchunk;
+	struct sctp_chunk *tchunk = NULL;
 	struct list_head *lchunk, *transport_list, *pos, *temp;
 	sctp_sack_variable_t *frags = sack->variable;
 	__u32 sack_ctsn, ctsn, tsn;
@@ -1083,11 +1107,6 @@ int sctp_outq_sack(struct sctp_outq *q, struct sctp_sackhdr *sack)
 
 	ctsn = asoc->ctsn_ack_point;
 
-	SCTP_DEBUG_PRINTK("%s: sack Cumulative TSN Ack is 0x%x.\n",
-			  __FUNCTION__, sack_ctsn);
-	SCTP_DEBUG_PRINTK("%s: Cumulative TSN Ack of association "
-			  "%p is 0x%x.\n", __FUNCTION__, asoc, ctsn);
-
 	/* Throw away stuff rotting on the sack queue.  */
 	list_for_each_safe(lchunk, temp, &q->sacked) {
 		tchunk = list_entry(lchunk, struct sctp_chunk,
@@ -1112,10 +1131,19 @@ int sctp_outq_sack(struct sctp_outq *q, struct sctp_sackhdr *sack)
 
 	asoc->peer.rwnd = sack_a_rwnd;
 
+	sctp_generate_fwdtsn(q, sack_ctsn);
+
+	SCTP_DEBUG_PRINTK("%s: sack Cumulative TSN Ack is 0x%x.\n",
+			  __FUNCTION__, sack_ctsn);
+	SCTP_DEBUG_PRINTK("%s: Cumulative TSN Ack of association, "
+			  "%p is 0x%x. Adv peer ack point: 0x%x\n",
+			  __FUNCTION__, asoc, ctsn, asoc->adv_peer_ack_point);
+
 	/* See if all chunks are acked.
 	 * Make sure the empty queue handler will get run later.
 	 */
-	q->empty = skb_queue_empty(&q->out) && list_empty(&q->retransmit);
+	q->empty = skb_queue_empty(&q->out) && skb_queue_empty(&q->control) &&
+			list_empty(&q->retransmit);
 	if (!q->empty)
 		goto finish;
 
@@ -1191,6 +1219,12 @@ static void sctp_check_transmitted(struct sctp_outq *q,
 		tchunk = list_entry(lchunk, struct sctp_chunk,
 				    transmitted_list);
 
+		if (sctp_chunk_abandoned(tchunk)) {
+			/* Move the chunk to abandoned list. */
+			sctp_insert_list(&q->abandoned, lchunk);
+			continue;
+		}
+
 		tsn = ntohl(tchunk->subh.data_hdr->tsn);
 		if (sctp_acked(sack, tsn)) {
 			/* If this queue is the retransmit queue, the
@@ -1572,3 +1606,123 @@ static int sctp_acked(struct sctp_sackhdr *sack, __u32 tsn)
 pass:
 	return 1;
 }
+
+static inline int sctp_get_skip_pos(struct sctp_fwdtsn_skip *skiplist,
+				    int nskips, __u16 stream)
+{
+	int i;
+
+	for (i = 0; i < nskips; i++) {
+		if (skiplist[i].stream == stream)
+			return i;
+	}
+	return i;
+}
+
+/* Create and add a fwdtsn chunk to the outq's control queue if needed. */
+static void sctp_generate_fwdtsn(struct sctp_outq *q, __u32 ctsn)
+{
+	struct sctp_association *asoc = q->asoc;
+	struct sctp_chunk *ftsn_chunk = NULL;
+	struct sctp_fwdtsn_skip ftsn_skip_arr[10];
+	int nskips = 0;
+	int skip_pos = 0;
+	__u32 tsn;
+	struct sctp_chunk *chunk;
+	struct list_head *lchunk, *temp;
+
+	/* PR-SCTP C1) Let SackCumAck be the Cumulative TSN ACK carried in the
+	 * received SACK.
+	 * 
+	 * If (Advanced.Peer.Ack.Point < SackCumAck), then update
+	 * Advanced.Peer.Ack.Point to be equal to SackCumAck.
+	 */
+	if (TSN_lt(asoc->adv_peer_ack_point, ctsn))
+		asoc->adv_peer_ack_point = ctsn;
+
+	/* PR-SCTP C2) Try to further advance the "Advanced.Peer.Ack.Point"
+	 * locally, that is, to move "Advanced.Peer.Ack.Point" up as long as
+	 * the chunk next in the out-queue space is marked as "abandoned" as
+	 * shown in the following example:
+	 *
+	 * Assuming that a SACK arrived with the Cumulative TSN ACK 102
+	 * and the Advanced.Peer.Ack.Point is updated to this value:
+	 * 
+	 *   out-queue at the end of  ==>   out-queue after Adv.Ack.Point
+	 *   normal SACK processing           local advancement
+	 *                ...                           ...
+	 *   Adv.Ack.Pt-> 102 acked                     102 acked
+	 *                103 abandoned                 103 abandoned
+	 *                104 abandoned     Adv.Ack.P-> 104 abandoned
+	 *                105                           105
+	 *                106 acked                     106 acked
+	 *                ...                           ...
+	 *
+	 * In this example, the data sender successfully advanced the
+	 * "Advanced.Peer.Ack.Point" from 102 to 104 locally.
+	 */
+	list_for_each_safe(lchunk, temp, &q->abandoned) {
+		chunk = list_entry(lchunk, struct sctp_chunk,
+					transmitted_list);
+		tsn = ntohl(chunk->subh.data_hdr->tsn);
+
+		/* Remove any chunks in the abandoned queue that are acked by
+		 * the ctsn.
+		 */ 
+		if (TSN_lte(tsn, ctsn)) {
+			list_del_init(lchunk);
+			if (!chunk->tsn_gap_acked) {
+			chunk->transport->flight_size -=
+						 sctp_data_size(chunk);
+			q->outstanding_bytes -= sctp_data_size(chunk);
+			}
+			sctp_chunk_free(chunk);
+		} else {
+			if (TSN_lte(tsn, asoc->adv_peer_ack_point+1)) {
+				asoc->adv_peer_ack_point = tsn;
+				if (chunk->chunk_hdr->flags &
+					 SCTP_DATA_UNORDERED)
+					continue;
+				skip_pos = sctp_get_skip_pos(&ftsn_skip_arr[0],
+						nskips,
+						chunk->subh.data_hdr->stream);
+				ftsn_skip_arr[skip_pos].stream =
+					chunk->subh.data_hdr->stream;
+				ftsn_skip_arr[skip_pos].ssn =
+					 chunk->subh.data_hdr->ssn;
+				if (skip_pos == nskips)
+					nskips++;
+				if (nskips == 10)
+					break;
+			} else
+				break;
+		}
+	}
+
+	/* PR-SCTP C3) If, after step C1 and C2, the "Advanced.Peer.Ack.Point"
+	 * is greater than the Cumulative TSN ACK carried in the received
+	 * SACK, the data sender MUST send the data receiver a FORWARD TSN
+	 * chunk containing the latest value of the
+	 * "Advanced.Peer.Ack.Point".
+	 *
+	 * C4) For each "abandoned" TSN the sender of the FORWARD TSN SHOULD
+	 * list each stream and sequence number in the forwarded TSN. This
+	 * information will enable the receiver to easily find any
+	 * stranded TSN's waiting on stream reorder queues. Each stream
+	 * SHOULD only be reported once; this means that if multiple
+	 * abandoned messages occur in the same stream then only the
+	 * highest abandoned stream sequence number is reported. If the
+	 * total size of the FORWARD TSN does NOT fit in a single MTU then
+	 * the sender of the FORWARD TSN SHOULD lower the
+	 * Advanced.Peer.Ack.Point to the last TSN that will fit in a
+	 * single MTU.
+	 */
+	if (asoc->adv_peer_ack_point > ctsn)
+		ftsn_chunk = sctp_make_fwdtsn(asoc, asoc->adv_peer_ack_point,
+					      nskips, &ftsn_skip_arr[0]); 
+
+	if (ftsn_chunk) {
+		__skb_queue_tail(&q->control, (struct sk_buff *)ftsn_chunk);
+		SCTP_INC_STATS(SctpOutCtrlChunks);
+	}
+}

net/sctp/protocol.c

@@ -1127,6 +1127,9 @@ __init int sctp_init(void)
 	/* Disable ADDIP by default. */
 	sctp_addip_enable = 0;
 
+	/* Enable PR-SCTP by default. */
+	sctp_prsctp_enable = 1;
+
 	sctp_sysctl_register();
 
 	INIT_LIST_HEAD(&sctp_address_families);

net/sctp/sm_make_chunk.c

@@ -6,10 +6,6 @@
  *
  * This file is part of the SCTP kernel reference Implementation
  *
- * This file includes part of the implementation of the add-IP extension,
- * based on <draft-ietf-tsvwg-addip-sctp-02.txt> June 29, 2001,
- * for the SCTP kernel reference Implementation.
- *
  * These functions work with the state functions in sctp_sm_statefuns.c
  * to implement the state operations.  These functions implement the
  * steps which require modifying existing data structures.
@@ -89,11 +85,13 @@ int sctp_chunk_iif(const struct sctp_chunk *chunk)
  * Note 2: The ECN capable field is reserved for future use of
  * Explicit Congestion Notification.
  */
-static const sctp_ecn_capable_param_t ecap_param = {
-	{
-		SCTP_PARAM_ECN_CAPABLE,
-		__constant_htons(sizeof(sctp_ecn_capable_param_t)),
-	}
+static const struct sctp_paramhdr ecap_param = {
+	SCTP_PARAM_ECN_CAPABLE,
+	__constant_htons(sizeof(struct sctp_paramhdr)),
+};
+static const struct sctp_paramhdr prsctp_param = {
+	SCTP_PARAM_FWD_TSN_SUPPORT,
+	__constant_htons(sizeof(struct sctp_paramhdr)),
 };
 
 /* A helper to initialize to initialize an op error inside a
@@ -196,6 +194,8 @@ struct sctp_chunk *sctp_make_init(const struct sctp_association *asoc,
 
 	chunksize = sizeof(init) + addrs_len + SCTP_SAT_LEN(num_types);
 	chunksize += sizeof(ecap_param);
+	if (sctp_prsctp_enable)
+		chunksize += sizeof(prsctp_param);
 	chunksize += vparam_len;
 
 	/* RFC 2960 3.3.2 Initiation (INIT) (1)
@@ -232,6 +232,8 @@ struct sctp_chunk *sctp_make_init(const struct sctp_association *asoc,
 	sctp_addto_chunk(retval, num_types * sizeof(__u16), &types);
 
 	sctp_addto_chunk(retval, sizeof(ecap_param), &ecap_param);
+	if (sctp_prsctp_enable)
+		sctp_addto_chunk(retval, sizeof(prsctp_param), &prsctp_param);
 nodata:
 	if (addrs.v)
 		kfree(addrs.v);
@@ -278,6 +280,10 @@ struct sctp_chunk *sctp_make_init_ack(const struct sctp_association *asoc,
 	if (asoc->peer.ecn_capable)
 		chunksize += sizeof(ecap_param);
 
+        /* Tell peer that we'll do PR-SCTP only if peer advertised.  */
+	if (asoc->peer.prsctp_capable)
+		chunksize += sizeof(prsctp_param);
+
 	/* Now allocate and fill out the chunk.  */
 	retval = sctp_make_chunk(asoc, SCTP_CID_INIT_ACK, 0, chunksize);
 	if (!retval)
@@ -293,6 +299,8 @@ struct sctp_chunk *sctp_make_init_ack(const struct sctp_association *asoc,
 	sctp_addto_chunk(retval, cookie_len, cookie);
 	if (asoc->peer.ecn_capable)
 		sctp_addto_chunk(retval, sizeof(ecap_param), &ecap_param);
+	if (asoc->peer.prsctp_capable)
+		sctp_addto_chunk(retval, sizeof(prsctp_param), &prsctp_param);
 
 	/* We need to remove the const qualifier at this point.  */
 	retval->asoc = (struct sctp_association *) asoc;
@@ -1286,6 +1294,9 @@ sctp_cookie_param_t *sctp_pack_cookie(const struct sctp_endpoint *ep,
 	/* Save the raw address list length in the cookie. */
 	cookie->c.raw_addr_list_len = addrs_len;
 
+	/* Remember PR-SCTP capability. */
+	cookie->c.prsctp_capable = asoc->peer.prsctp_capable;
+
 	/* Set an expiration time for the cookie.  */
 	do_gettimeofday(&cookie->c.expiration);
 	TIMEVAL_ADD(asoc->cookie_life, cookie->c.expiration);
@@ -1442,6 +1453,8 @@ struct sctp_association *sctp_unpack_cookie(
 	retval->next_tsn = retval->c.initial_tsn;
 	retval->ctsn_ack_point = retval->next_tsn - 1;
 	retval->addip_serial = retval->c.initial_tsn;
+	retval->adv_peer_ack_point = retval->ctsn_ack_point;
+	retval->peer.prsctp_capable = retval->c.prsctp_capable;
 
 	/* The INIT stuff will be done by the side effects.  */
 	return retval;
@@ -1653,6 +1666,10 @@ static int sctp_verify_param(const struct sctp_association *asoc,
 	case SCTP_PARAM_HOST_NAME_ADDRESS:
 		/* Tell the peer, we won't support this param.  */
 		return sctp_process_hn_param(asoc, param, chunk, err_chunk);
+	case SCTP_PARAM_FWD_TSN_SUPPORT:
+		if (sctp_prsctp_enable)
+			break;
+		/* Fall Through */ 
 	default:
 		SCTP_DEBUG_PRINTK("Unrecognized param: %d for chunk %d.\n",
 				ntohs(param.p->type), cid);
@@ -1968,6 +1985,12 @@ int sctp_process_param(struct sctp_association *asoc, union sctp_params param,
 		asoc->peer.ecn_capable = 1;
 		break;
 
+	case SCTP_PARAM_FWD_TSN_SUPPORT:
+		if (sctp_prsctp_enable) {
+			asoc->peer.prsctp_capable = 1;
+			break;
+		}
+		/* Fall Through */ 
 	default:
 		/* Any unrecognized parameters should have been caught
 		 * and handled by sctp_verify_param() which should be
@@ -2622,3 +2645,38 @@ int sctp_process_asconf_ack(struct sctp_association *asoc,
 
 	return retval;
 }
+
+/* Make a FWD TSN chunk. */ 
+struct sctp_chunk *sctp_make_fwdtsn(const struct sctp_association *asoc,
+				    __u32 new_cum_tsn, size_t nstreams,
+				    struct sctp_fwdtsn_skip *skiplist)
+{
+	struct sctp_chunk *retval = NULL;
+	struct sctp_fwdtsn_chunk *ftsn_chunk;
+	struct sctp_fwdtsn_hdr ftsn_hdr; 
+	struct sctp_fwdtsn_skip skip;
+	size_t hint;
+	int i;
+
+	hint = (nstreams + 1) * sizeof(__u32);
+
+	/* Maybe set the T-bit if we have no association.  */
+	retval = sctp_make_chunk(asoc, SCTP_CID_FWD_TSN, 0, hint);
+
+	if (!retval)
+		return NULL;
+
+	ftsn_chunk = (struct sctp_fwdtsn_chunk *)retval->subh.fwdtsn_hdr;
+
+	ftsn_hdr.new_cum_tsn = htonl(new_cum_tsn);
+	retval->subh.fwdtsn_hdr =
+		sctp_addto_chunk(retval, sizeof(ftsn_hdr), &ftsn_hdr);
+
+	for (i = 0; i < nstreams; i++) {
+		skip.stream = skiplist[i].stream;
+		skip.ssn = skiplist[i].ssn;
+		sctp_addto_chunk(retval, sizeof(skip), &skip);
+	}
+
+	return retval;
+}

net/sctp/sm_sideeffect.c

@@ -579,7 +579,7 @@ static void sctp_cmd_transport_reset(sctp_cmd_seq_t *cmds,
 /* Helper function to process the process SACK command.  */
 static int sctp_cmd_process_sack(sctp_cmd_seq_t *cmds,
 				 struct sctp_association *asoc,
-				 sctp_sackhdr_t *sackh)
+				 struct sctp_sackhdr *sackh)
 {
 	int err;
 
@@ -729,6 +729,19 @@ static void sctp_cmd_process_operr(sctp_cmd_seq_t *cmds,
 	}
 }
 
+/* Process variable FWDTSN chunk information. */
+static void sctp_cmd_process_fwdtsn(struct sctp_ulpq *ulpq, 
+				    struct sctp_chunk *chunk)
+{
+	struct sctp_fwdtsn_skip *skip;
+	/* Walk through all the skipped SSNs */
+	sctp_walk_fwdtsn(skip, chunk) {
+		sctp_ulpq_skip(ulpq, ntohs(skip->stream), ntohs(skip->ssn));
+	}
+
+	return;
+}
+
 /* These three macros allow us to pull the debugging code out of the
  * main flow of sctp_do_sm() to keep attention focused on the real
  * functionality there.
@@ -903,7 +916,7 @@ int sctp_cmd_interpreter(sctp_event_t event_type, sctp_subtype_t subtype,
 	struct timer_list *timer;
 	unsigned long timeout;
 	struct sctp_transport *t;
-	sctp_sackhdr_t sackh;
+	struct sctp_sackhdr sackh;
 	int local_cork = 0;
 
 	if (SCTP_EVENT_T_TIMEOUT != event_type)
@@ -962,6 +975,18 @@ int sctp_cmd_interpreter(sctp_event_t event_type, sctp_subtype_t subtype,
 			sctp_tsnmap_mark(&asoc->peer.tsn_map, cmd->obj.u32);
 			break;
 
+		case SCTP_CMD_REPORT_FWDTSN:
+			/* Move the Cumulattive TSN Ack ahead. */
+			sctp_tsnmap_skip(&asoc->peer.tsn_map, cmd->obj.u32);
+
+			/* Abort any in progress partial delivery. */
+			sctp_ulpq_abort_pd(&asoc->ulpq, GFP_ATOMIC);
+			break;
+
+		case SCTP_CMD_PROCESS_FWDTSN:
+			sctp_cmd_process_fwdtsn(&asoc->ulpq, cmd->obj.ptr);
+                        break;
+
 		case SCTP_CMD_GEN_SACK:
 			/* Generate a Selective ACK.
 			 * The argument tells us whether to just count

net/sctp/sm_statefuns.c

@@ -3204,6 +3204,143 @@ sctp_disposition_t sctp_sf_do_asconf_ack(const struct sctp_endpoint *ep,
 	return SCTP_DISPOSITION_DISCARD;
 }
 
+/*
+ * PR-SCTP Section 3.6 Receiver Side Implementation of PR-SCTP
+ *
+ * When a FORWARD TSN chunk arrives, the data receiver MUST first update
+ * its cumulative TSN point to the value carried in the FORWARD TSN
+ * chunk, and then MUST further advance its cumulative TSN point locally
+ * if possible.
+ * After the above processing, the data receiver MUST stop reporting any
+ * missing TSNs earlier than or equal to the new cumulative TSN point.
+ *
+ * Verification Tag:  8.5 Verification Tag [Normal verification]
+ *
+ * The return value is the disposition of the chunk.
+ */
+sctp_disposition_t sctp_sf_eat_fwd_tsn(const struct sctp_endpoint *ep,
+				       const struct sctp_association *asoc,
+				       const sctp_subtype_t type,
+				       void *arg,
+				       sctp_cmd_seq_t *commands)
+{
+	struct sctp_chunk *chunk = arg;
+	struct sctp_fwdtsn_hdr *fwdtsn_hdr;
+	__u16 len;
+	__u32 tsn;
+
+	/* RFC 2960 8.5 Verification Tag
+	 *
+	 * When receiving an SCTP packet, the endpoint MUST ensure
+	 * that the value in the Verification Tag field of the
+	 * received SCTP packet matches its own Tag.
+	 */
+	if (ntohl(chunk->sctp_hdr->vtag) != asoc->c.my_vtag) {
+		sctp_add_cmd_sf(commands, SCTP_CMD_REPORT_BAD_TAG,
+				SCTP_NULL());
+		return sctp_sf_pdiscard(ep, asoc, type, arg, commands);
+	}
+
+	fwdtsn_hdr = (struct sctp_fwdtsn_hdr *)chunk->skb->data;
+	chunk->subh.fwdtsn_hdr = fwdtsn_hdr;
+	len = ntohs(chunk->chunk_hdr->length);
+	len -= sizeof(struct sctp_chunkhdr);
+	skb_pull(chunk->skb, len);
+
+	tsn = ntohl(fwdtsn_hdr->new_cum_tsn);
+	SCTP_DEBUG_PRINTK("%s: TSN 0x%x.\n", __FUNCTION__, tsn);
+
+	/* The TSN is too high--silently discard the chunk and count on it
+	 * getting retransmitted later.
+	 */
+	if (sctp_tsnmap_check(&asoc->peer.tsn_map, tsn) < 0)
+		goto discard_noforce;
+
+	sctp_add_cmd_sf(commands, SCTP_CMD_REPORT_FWDTSN, SCTP_U32(tsn));
+	if (len > sizeof(struct sctp_fwdtsn_hdr))
+		sctp_add_cmd_sf(commands, SCTP_CMD_PROCESS_FWDTSN, 
+				SCTP_CHUNK(chunk));
+	
+	/* Count this as receiving DATA. */
+	if (asoc->autoclose) {
+		sctp_add_cmd_sf(commands, SCTP_CMD_TIMER_RESTART,
+				SCTP_TO(SCTP_EVENT_TIMEOUT_AUTOCLOSE));
+	}
+	
+	/* FIXME: For now send a SACK, but DATA processing may
+	 * send another. 
+	 */
+	sctp_add_cmd_sf(commands, SCTP_CMD_GEN_SACK, SCTP_NOFORCE());
+	/* Start the SACK timer.  */
+	sctp_add_cmd_sf(commands, SCTP_CMD_TIMER_RESTART,
+			SCTP_TO(SCTP_EVENT_TIMEOUT_SACK));
+
+	return SCTP_DISPOSITION_CONSUME;
+
+discard_noforce:
+	return SCTP_DISPOSITION_DISCARD;
+}
+
+sctp_disposition_t sctp_sf_eat_fwd_tsn_fast(
+	const struct sctp_endpoint *ep,
+	const struct sctp_association *asoc,
+	const sctp_subtype_t type,
+	void *arg,
+	sctp_cmd_seq_t *commands)
+{
+	struct sctp_chunk *chunk = arg;
+	struct sctp_fwdtsn_hdr *fwdtsn_hdr;
+	__u16 len;
+	__u32 tsn;
+
+	/* RFC 2960 8.5 Verification Tag
+	 *
+	 * When receiving an SCTP packet, the endpoint MUST ensure
+	 * that the value in the Verification Tag field of the
+	 * received SCTP packet matches its own Tag.
+	 */
+	if (ntohl(chunk->sctp_hdr->vtag) != asoc->c.my_vtag) {
+		sctp_add_cmd_sf(commands, SCTP_CMD_REPORT_BAD_TAG,
+				SCTP_NULL());
+		return sctp_sf_pdiscard(ep, asoc, type, arg, commands);
+	}
+
+	fwdtsn_hdr = (struct sctp_fwdtsn_hdr *)chunk->skb->data;
+	chunk->subh.fwdtsn_hdr = fwdtsn_hdr;
+	len = ntohs(chunk->chunk_hdr->length);
+	len -= sizeof(struct sctp_chunkhdr);
+	skb_pull(chunk->skb, len);
+
+	tsn = ntohl(fwdtsn_hdr->new_cum_tsn);
+	SCTP_DEBUG_PRINTK("%s: TSN 0x%x.\n", __FUNCTION__, tsn);
+
+	/* The TSN is too high--silently discard the chunk and count on it
+	 * getting retransmitted later.
+	 */
+	if (sctp_tsnmap_check(&asoc->peer.tsn_map, tsn) < 0)
+		goto gen_shutdown;
+
+	sctp_add_cmd_sf(commands, SCTP_CMD_REPORT_FWDTSN, SCTP_U32(tsn));
+	if (len > sizeof(struct sctp_fwdtsn_hdr))
+		sctp_add_cmd_sf(commands, SCTP_CMD_PROCESS_FWDTSN, 
+				SCTP_CHUNK(chunk));
+	
+	/* Go a head and force a SACK, since we are shutting down. */
+gen_shutdown:
+	/* Implementor's Guide.
+	 *
+	 * While in SHUTDOWN-SENT state, the SHUTDOWN sender MUST immediately
+	 * respond to each received packet containing one or more DATA chunk(s)
+	 * with a SACK, a SHUTDOWN chunk, and restart the T2-shutdown timer
+	 */
+	sctp_add_cmd_sf(commands, SCTP_CMD_GEN_SHUTDOWN, SCTP_NULL());
+	sctp_add_cmd_sf(commands, SCTP_CMD_GEN_SACK, SCTP_FORCE());
+	sctp_add_cmd_sf(commands, SCTP_CMD_TIMER_RESTART,
+			SCTP_TO(SCTP_EVENT_TIMEOUT_T2_SHUTDOWN));
+
+        return SCTP_DISPOSITION_CONSUME;
+}
+
 /*
  * Process an unknown chunk.
  *

net/sctp/sm_statetable.c

 /* SCTP kernel reference Implementation
- * (C) Copyright IBM Corp. 2001, 2003
+ * (C) Copyright IBM Corp. 2001, 2004
  * Copyright (c) 1999-2000 Cisco, Inc.
  * Copyright (c) 1999-2001 Motorola, Inc.
  * Copyright (c) 2001 Intel Corp.
@@ -40,6 +40,7 @@
  *    Hui Huang		    <hui.huang@nokia.com>
  *    Daisy Chang	    <daisyc@us.ibm.com>
  *    Ardelle Fan	    <ardelle.fan@intel.com>
+ *    Sridhar Samudrala	    <sri@us.ibm.com>
  *
  * Any bugs reported given to us we will try to fix... any fixes shared will
  * be incorporated into the next SCTP release.
@@ -50,7 +51,7 @@
 #include <net/sctp/sm.h>
 
 static const sctp_sm_table_entry_t bug = {
-	.fn = sctp_sf_bug, 
+	.fn = sctp_sf_bug,
 	.name = "sctp_sf_bug"
 };
 
@@ -73,7 +74,7 @@ const sctp_sm_table_entry_t *sctp_sm_lookup_event(sctp_event_t event_type,
 		return sctp_chunk_event_lookup(event_subtype.chunk, state);
 		break;
 	case SCTP_EVENT_T_TIMEOUT:
-		DO_LOOKUP(SCTP_EVENT_TIMEOUT_MAX, timeout, 
+		DO_LOOKUP(SCTP_EVENT_TIMEOUT_MAX, timeout,
 			  timeout_event_table);
 		break;
 
@@ -486,6 +487,34 @@ const sctp_sm_table_entry_t addip_chunk_event_table[SCTP_NUM_ADDIP_CHUNK_TYPES][
 	TYPE_SCTP_ASCONF_ACK,
 }; /*state_fn_t addip_chunk_event_table[][] */
 
+#define TYPE_SCTP_FWD_TSN { \
+	/* SCTP_STATE_EMPTY */ \
+	{.fn = sctp_sf_ootb, .name = "sctp_sf_ootb"}, \
+	/* SCTP_STATE_CLOSED */ \
+	{.fn = sctp_sf_tabort_8_4_8, .name = "sctp_sf_tabort_8_4_8"}, \
+	/* SCTP_STATE_COOKIE_WAIT */ \
+	{.fn = sctp_sf_discard_chunk, .name = "sctp_sf_discard_chunk"}, \
+	/* SCTP_STATE_COOKIE_ECHOED */ \
+	{.fn = sctp_sf_discard_chunk, .name = "sctp_sf_discard_chunk"}, \
+	/* SCTP_STATE_ESTABLISHED */ \
+	{.fn = sctp_sf_eat_fwd_tsn, .name = "sctp_sf_eat_fwd_tsn"}, \
+	/* SCTP_STATE_SHUTDOWN_PENDING */ \
+	{.fn = sctp_sf_eat_fwd_tsn, .name = "sctp_sf_eat_fwd_tsn"}, \
+	/* SCTP_STATE_SHUTDOWN_SENT */ \
+	{.fn = sctp_sf_eat_fwd_tsn_fast, .name = "sctp_sf_eat_fwd_tsn_fast"}, \
+	/* SCTP_STATE_SHUTDOWN_RECEIVED */ \
+	{.fn = sctp_sf_discard_chunk, .name = "sctp_sf_discard_chunk"}, \
+	/* SCTP_STATE_SHUTDOWN_ACK_SENT */ \
+	{.fn = sctp_sf_discard_chunk, .name = "sctp_sf_discard_chunk"}, \
+} /* TYPE_SCTP_FWD_TSN */
+
+/* The primary index for this table is the chunk type.
+ * The secondary index for this table is the state.
+ */
+const sctp_sm_table_entry_t prsctp_chunk_event_table[SCTP_NUM_PRSCTP_CHUNK_TYPES][SCTP_STATE_NUM_STATES] = {
+	TYPE_SCTP_FWD_TSN,
+}; /*state_fn_t prsctp_chunk_event_table[][] */
+
 static const sctp_sm_table_entry_t
 chunk_event_table_unknown[SCTP_STATE_NUM_STATES] = {
 	/* SCTP_STATE_EMPTY */
@@ -924,6 +953,11 @@ const sctp_sm_table_entry_t *sctp_chunk_event_lookup(sctp_cid_t cid,
 	if (cid >= 0 && cid <= SCTP_CID_BASE_MAX)
 		return &chunk_event_table[cid][state];
 
+	if (sctp_prsctp_enable) {
+		if (cid == SCTP_CID_FWD_TSN)
+			return &prsctp_chunk_event_table[0][state];
+	}
+
 	if (sctp_addip_enable) {
 		if (cid == SCTP_CID_ASCONF)
 			return &addip_chunk_event_table[0][state];

net/sctp/sysctl.c

 /* SCTP kernel reference Implementation
- * Copyright (c) 2002 International Business Machines Corp.
+ * (C) Copyright IBM Corp. 2002, 2004
  * Copyright (c) 2002 Intel Corp.
  *
  * This file is part of the SCTP kernel reference Implementation
@@ -35,6 +35,7 @@
  *    Jon Grimm             <jgrimm@us.ibm.com>
  *    Ardelle Fan           <ardelle.fan@intel.com>
  *    Ryan Layer            <rmlayer@us.ibm.com>
+ *    Sridhar Samudrala     <sri@us.ibm.com>
  *
  * Any bugs reported given to us we will try to fix... any fixes shared will
  * be incorporated into the next SCTP release.
@@ -170,6 +171,14 @@ static ctl_table sctp_table[] = {
 		.mode		= 0644,
 		.proc_handler	= &proc_dointvec
 	},
+	{
+		.ctl_name	= NET_SCTP_PRSCTP_ENABLE,
+		.procname	= "prsctp_enable",
+		.data		= &sctp_prsctp_enable,
+		.maxlen		= sizeof(int),
+		.mode		= 0644,
+		.proc_handler	= &proc_dointvec
+	},
 	{ .ctl_name = 0 }
 };
 

net/sctp/tsnmap.c

 /* SCTP kernel reference Implementation
+ * (C) Copyright IBM Corp. 2001, 2004
  * Copyright (c) 1999-2000 Cisco, Inc.
  * Copyright (c) 1999-2001 Motorola, Inc.
- * Copyright (c) 2001-2003 International Business Machines, Corp.
  * Copyright (c) 2001 Intel Corp.
  *
  * This file is part of the SCTP kernel reference Implementation
@@ -36,6 +36,7 @@
  *    La Monte H.P. Yarroll <piggy@acm.org>
  *    Jon Grimm             <jgrimm@us.ibm.com>
  *    Karl Knutson          <karl@athena.chicago.il.us>
+ *    Sridhar Samudrala     <sri@us.ibm.com>
  *
  * Any bugs reported given to us we will try to fix... any fixes shared will
  * be incorporated into the next SCTP release.
@@ -253,6 +254,40 @@ int sctp_tsnmap_next_gap_ack(const struct sctp_tsnmap *map,
 	return ended;
 }
 
+/* Mark this and any lower TSN as seen.  */
+void sctp_tsnmap_skip(struct sctp_tsnmap *map, __u32 tsn)
+{
+	__s32 gap;
+
+	/* Vacuously mark any TSN which precedes the map base or
+	 * exceeds the end of the map.
+	 */
+	if (TSN_lt(tsn, map->base_tsn))
+		return;
+	if (!TSN_lt(tsn, map->base_tsn + map->len + map->len))
+		return;
+
+	/* Bump the max.  */
+	if (TSN_lt(map->max_tsn_seen, tsn))
+		map->max_tsn_seen = tsn;
+
+	/* Assert: TSN is in range.  */
+	gap = tsn - map->base_tsn + 1;
+
+	/* Mark the TSNs as received.  */
+	if (gap <= map->len)
+		memset(map->tsn_map, 0x01, gap);
+	else {
+		memset(map->tsn_map, 0x01, map->len);
+		memset(map->overflow_map, 0x01, (gap - map->len));
+	}
+
+	/* Go fixup any internal TSN mapping variables including
+	 * cumulative_tsn_ack_point.
+	 */
+	sctp_tsnmap_update(map);
+}
+
 /********************************************************************
  * 2nd Level Abstractions
  ********************************************************************/

net/sctp/ulpqueue.c

@@ -680,6 +680,71 @@ static inline struct sctp_ulpevent *sctp_ulpq_order(struct sctp_ulpq *ulpq,
 	return event;
 }
 
+/* Helper function to gather skbs that have possibly become
+ * ordered by forward tsn skipping their dependencies.
+ */
+static inline void sctp_ulpq_reap_ordered(struct sctp_ulpq *ulpq)
+{
+	struct sk_buff *pos, *tmp;
+	struct sctp_ulpevent *cevent;
+	struct sctp_ulpevent *event = NULL;
+	struct sctp_stream *in;
+	struct sk_buff_head temp;
+	__u16 csid, cssn;
+
+	in  = &ulpq->asoc->ssnmap->in;
+
+	/* We are holding the chunks by stream, by SSN.  */
+	sctp_skb_for_each(pos, &ulpq->lobby, tmp) {
+		cevent = (struct sctp_ulpevent *) pos->cb;
+		csid = cevent->stream;
+		cssn = cevent->ssn;
+
+		if (cssn != sctp_ssn_peek(in, csid))
+			break;
+
+		/* Found it, so mark in the ssnmap. */	       
+		sctp_ssn_next(in, csid);
+
+		__skb_unlink(pos, pos->list);
+		if (!event) {						
+			/* Create a temporary list to collect chunks on.  */
+			event = sctp_skb2event(pos);
+			skb_queue_head_init(&temp);
+			__skb_queue_tail(&temp, sctp_event2skb(event));
+		} else {
+			/* Attach all gathered skbs to the event.  */
+			__skb_queue_tail(sctp_event2skb(event)->list, pos);
+		}
+	}
+
+	/* Send event to the ULP.  */
+	if (event)
+		sctp_ulpq_tail_event(ulpq, event);
+}
+
+/* Skip over an SSN. */
+void sctp_ulpq_skip(struct sctp_ulpq *ulpq, __u16 sid, __u16 ssn)
+{
+	struct sctp_stream *in;
+
+	/* Note: The stream ID must be verified before this routine.  */
+	in  = &ulpq->asoc->ssnmap->in;
+
+	/* Is this an old SSN?  If so ignore. */
+	if (SSN_lt(ssn, sctp_ssn_peek(in, sid)))
+		return;
+
+	/* Mark that we are no longer expecting this SSN or lower. */
+	sctp_ssn_skip(in, sid, ssn);
+
+	/* Go find any other chunks that were waiting for
+	 * ordering and deliver them if needed. 
+	 */
+	sctp_ulpq_reap_ordered(ulpq);
+	return;
+}
+
 /* Renege 'needed' bytes from the ordering queue. */
 static __u16 sctp_ulpq_renege_order(struct sctp_ulpq *ulpq, __u16 needed)
 {