sctp: Add GSO support

SCTP has this pecualiarity that its packets cannot be just segmented to
(P)MTU. Its chunks must be contained in IP segments, padding respected.
So we can't just generate a big skb, set gso_size to the fragmentation
point and deliver it to IP layer.

This patch takes a different approach. SCTP will now build a skb as it
would be if it was received using GRO. That is, there will be a cover
skb with protocol headers and children ones containing the actual
segments, already segmented to a way that respects SCTP RFCs.

With that, we can tell skb_segment() to just split based on frag_list,
trusting its sizes are already in accordance.

This way SCTP can benefit from GSO and instead of passing several
packets through the stack, it can pass a single large packet.

v2:
- Added support for receiving GSO frames, as requested by Dave Miller.
- Clear skb->cb if packet is GSO (otherwise it's not used by SCTP)
- Added heuristics similar to what we have in TCP for not generating
  single GSO packets that fills cwnd.
v3:
- consider sctphdr size in skb_gso_transport_seglen()
- rebased due to 5c7cdf339af5 ("gso: Remove arbitrary checks for
  unsupported GSO")

Signed-off-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Tested-by: Xin Long <lucien.xin@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

1 files changed, 249 insertions, 114 deletions

diff --git a/net/sctp/output.c b/net/sctp/output.c
index 9844fe573029..60499a69179d 100644
--- a/net/sctp/output.c
+++ b/net/sctp/output.c
@@ -84,18 +84,42 @@ static void sctp_packet_reset(struct sctp_packet *packet)
 struct sctp_packet *sctp_packet_config(struct sctp_packet *packet,
 				       __u32 vtag, int ecn_capable)
 {
-	struct sctp_chunk *chunk = NULL;
+	struct sctp_transport *tp = packet->transport;
+	struct sctp_association *asoc = tp->asoc;
 
 	pr_debug("%s: packet:%p vtag:0x%x\n", __func__, packet, vtag);
 
 	packet->vtag = vtag;
 
+	if (asoc && tp->dst) {
+		struct sock *sk = asoc->base.sk;
+
+		rcu_read_lock();
+		if (__sk_dst_get(sk) != tp->dst) {
+			dst_hold(tp->dst);
+			sk_setup_caps(sk, tp->dst);
+		}
+
+		if (sk_can_gso(sk)) {
+			struct net_device *dev = tp->dst->dev;
+
+			packet->max_size = dev->gso_max_size;
+		} else {
+			packet->max_size = asoc->pathmtu;
+		}
+		rcu_read_unlock();
+
+	} else {
+		packet->max_size = tp->pathmtu;
+	}
+
 	if (ecn_capable && sctp_packet_empty(packet)) {
-		chunk = sctp_get_ecne_prepend(packet->transport->asoc);
+		struct sctp_chunk *chunk;
 
 		/* If there a is a prepend chunk stick it on the list before
 		 * any other chunks get appended.
 		 */
+		chunk = sctp_get_ecne_prepend(asoc);
 		if (chunk)
 			sctp_packet_append_chunk(packet, chunk);
 	}
@@ -381,12 +405,15 @@ int sctp_packet_transmit(struct sctp_packet *packet, gfp_t gfp)
 	struct sctp_transport *tp = packet->transport;
 	struct sctp_association *asoc = tp->asoc;
 	struct sctphdr *sh;
-	struct sk_buff *nskb;
+	struct sk_buff *nskb = NULL, *head = NULL;
 	struct sctp_chunk *chunk, *tmp;
 	struct sock *sk;
 	int err = 0;
 	int padding;		/* How much padding do we need?  */
+	int pkt_size;
 	__u8 has_data = 0;
+	int gso = 0;
+	int pktcount = 0;
 	struct dst_entry *dst;
 	unsigned char *auth = NULL;	/* pointer to auth in skb data */
 
@@ -400,18 +427,37 @@ int sctp_packet_transmit(struct sctp_packet *packet, gfp_t gfp)
 	chunk = list_entry(packet->chunk_list.next, struct sctp_chunk, list);
 	sk = chunk->skb->sk;
 
-	/* Allocate the new skb.  */
-	nskb = alloc_skb(packet->size + MAX_HEADER, gfp);
-	if (!nskb)
+	/* Allocate the head skb, or main one if not in GSO */
+	if (packet->size > tp->pathmtu && !packet->ipfragok) {
+		if (sk_can_gso(sk)) {
+			gso = 1;
+			pkt_size = packet->overhead;
+		} else {
+			/* If this happens, we trash this packet and try
+			 * to build a new one, hopefully correct this
+			 * time. Application may notice this error.
+			 */
+			pr_err_once("Trying to GSO but underlying device doesn't support it.");
+			goto nomem;
+		}
+	} else {
+		pkt_size = packet->size;
+	}
+	head = alloc_skb(pkt_size + MAX_HEADER, gfp);
+	if (!head)
 		goto nomem;
+	if (gso) {
+		NAPI_GRO_CB(head)->last = head;
+		skb_shinfo(head)->gso_type = sk->sk_gso_type;
+	}
 
 	/* Make sure the outbound skb has enough header room reserved. */
-	skb_reserve(nskb, packet->overhead + MAX_HEADER);
+	skb_reserve(head, packet->overhead + MAX_HEADER);
 
 	/* Set the owning socket so that we know where to get the
 	 * destination IP address.
 	 */
-	sctp_packet_set_owner_w(nskb, sk);
+	sctp_packet_set_owner_w(head, sk);
 
 	if (!sctp_transport_dst_check(tp)) {
 		sctp_transport_route(tp, NULL, sctp_sk(sk));
@@ -422,11 +468,11 @@ int sctp_packet_transmit(struct sctp_packet *packet, gfp_t gfp)
 	dst = dst_clone(tp->dst);
 	if (!dst)
 		goto no_route;
-	skb_dst_set(nskb, dst);
+	skb_dst_set(head, dst);
 
 	/* Build the SCTP header.  */
-	sh = (struct sctphdr *)skb_push(nskb, sizeof(struct sctphdr));
-	skb_reset_transport_header(nskb);
+	sh = (struct sctphdr *)skb_push(head, sizeof(struct sctphdr));
+	skb_reset_transport_header(head);
 	sh->source = htons(packet->source_port);
 	sh->dest   = htons(packet->destination_port);
 
@@ -441,90 +487,133 @@ int sctp_packet_transmit(struct sctp_packet *packet, gfp_t gfp)
 	sh->vtag     = htonl(packet->vtag);
 	sh->checksum = 0;
 
-	/**
-	 * 6.10 Bundling
-	 *
-	 *    An endpoint bundles chunks by simply including multiple
-	 *    chunks in one outbound SCTP packet.  ...
-	 */
-
-	/**
-	 * 3.2  Chunk Field Descriptions
-	 *
-	 * The total length of a chunk (including Type, Length and
-	 * Value fields) MUST be a multiple of 4 bytes.  If the length
-	 * of the chunk is not a multiple of 4 bytes, the sender MUST
-	 * pad the chunk with all zero bytes and this padding is not
-	 * included in the chunk length field.  The sender should
-	 * never pad with more than 3 bytes.
-	 *
-	 * [This whole comment explains WORD_ROUND() below.]
-	 */
-
 	pr_debug("***sctp_transmit_packet***\n");
 
-	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.
-			 */
+	do {
+		/* Set up convenience variables... */
+		chunk = list_entry(packet->chunk_list.next, struct sctp_chunk, list);
+		pktcount++;
 
-			if (!chunk->resent && !tp->rto_pending) {
-				chunk->rtt_in_progress = 1;
-				tp->rto_pending = 1;
+		/* Calculate packet size, so it fits in PMTU. Leave
+		 * other chunks for the next packets.
+		 */
+		if (gso) {
+			pkt_size = packet->overhead;
+			list_for_each_entry(chunk, &packet->chunk_list, list) {
+				int padded = WORD_ROUND(chunk->skb->len);
+
+				if (pkt_size + padded > tp->pathmtu)
+					break;
+				pkt_size += padded;
 			}
 
-			has_data = 1;
-		}
+			/* Allocate a new skb. */
+			nskb = alloc_skb(pkt_size + MAX_HEADER, gfp);
+			if (!nskb)
+				goto nomem;
 
-		padding = WORD_ROUND(chunk->skb->len) - chunk->skb->len;
-		if (padding)
-			memset(skb_put(chunk->skb, padding), 0, padding);
+			/* Make sure the outbound skb has enough header
+			 * room reserved.
+			 */
+			skb_reserve(nskb, packet->overhead + MAX_HEADER);
+		} else {
+			nskb = head;
+		}
 
-		/* if this is the auth chunk that we are adding,
-		 * store pointer where it will be added and put
-		 * the auth into the packet.
+		/**
+		 * 3.2  Chunk Field Descriptions
+		 *
+		 * The total length of a chunk (including Type, Length and
+		 * Value fields) MUST be a multiple of 4 bytes.  If the length
+		 * of the chunk is not a multiple of 4 bytes, the sender MUST
+		 * pad the chunk with all zero bytes and this padding is not
+		 * included in the chunk length field.  The sender should
+		 * never pad with more than 3 bytes.
+		 *
+		 * [This whole comment explains WORD_ROUND() below.]
 		 */
-		if (chunk == packet->auth)
-			auth = skb_tail_pointer(nskb);
 
-		memcpy(skb_put(nskb, chunk->skb->len),
+		pkt_size -= packet->overhead;
+		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 && !tp->rto_pending) {
+					chunk->rtt_in_progress = 1;
+					tp->rto_pending = 1;
+				}
+
+				has_data = 1;
+			}
+
+			padding = WORD_ROUND(chunk->skb->len) - chunk->skb->len;
+			if (padding)
+				memset(skb_put(chunk->skb, padding), 0, padding);
+
+			/* if this is the auth chunk that we are adding,
+			 * store pointer where it will be added and put
+			 * the auth into the packet.
+			 */
+			if (chunk == packet->auth)
+				auth = skb_tail_pointer(nskb);
+
+			memcpy(skb_put(nskb, chunk->skb->len),
 			       chunk->skb->data, chunk->skb->len);
 
-		pr_debug("*** Chunk:%p[%s] %s 0x%x, length:%d, chunk->skb->len:%d, "
-			 "rtt_in_progress:%d\n", chunk,
-			 sctp_cname(SCTP_ST_CHUNK(chunk->chunk_hdr->type)),
-			 chunk->has_tsn ? "TSN" : "No TSN",
-			 chunk->has_tsn ? ntohl(chunk->subh.data_hdr->tsn) : 0,
-			 ntohs(chunk->chunk_hdr->length), chunk->skb->len,
-			 chunk->rtt_in_progress);
-
-		/*
-		 * If this is a control chunk, this is our last
-		 * reference. Free data chunks after they've been
-		 * acknowledged or have failed.
-		 */
-		if (!sctp_chunk_is_data(chunk))
-			sctp_chunk_free(chunk);
-	}
+			pr_debug("*** Chunk:%p[%s] %s 0x%x, length:%d, chunk->skb->len:%d, rtt_in_progress:%d\n",
+				 chunk,
+				 sctp_cname(SCTP_ST_CHUNK(chunk->chunk_hdr->type)),
+				 chunk->has_tsn ? "TSN" : "No TSN",
+				 chunk->has_tsn ? ntohl(chunk->subh.data_hdr->tsn) : 0,
+				 ntohs(chunk->chunk_hdr->length), chunk->skb->len,
+				 chunk->rtt_in_progress);
+
+			/* If this is a control chunk, this is our last
+			 * reference. Free data chunks after they've been
+			 * acknowledged or have failed.
+			 * Re-queue auth chunks if needed.
+			 */
+			pkt_size -= WORD_ROUND(chunk->skb->len);
 
-	/* SCTP-AUTH, Section 6.2
-	 *    The sender MUST calculate the MAC as described in RFC2104 [2]
-	 *    using the hash function H as described by the MAC Identifier and
-	 *    the shared association key K based on the endpoint pair shared key
-	 *    described by the shared key identifier.  The 'data' used for the
-	 *    computation of the AUTH-chunk is given by the AUTH chunk with its
-	 *    HMAC field set to zero (as shown in Figure 6) followed by all
-	 *    chunks that are placed after the AUTH chunk in the SCTP packet.
-	 */
-	if (auth)
-		sctp_auth_calculate_hmac(asoc, nskb,
-					 (struct sctp_auth_chunk *)auth,
-					 gfp);
+			if (chunk == packet->auth && !list_empty(&packet->chunk_list))
+				list_add(&chunk->list, &packet->chunk_list);
+			else if (!sctp_chunk_is_data(chunk))
+				sctp_chunk_free(chunk);
+
+			if (!pkt_size)
+				break;
+		}
+
+		/* SCTP-AUTH, Section 6.2
+		 *    The sender MUST calculate the MAC as described in RFC2104 [2]
+		 *    using the hash function H as described by the MAC Identifier and
+		 *    the shared association key K based on the endpoint pair shared key
+		 *    described by the shared key identifier.  The 'data' used for the
+		 *    computation of the AUTH-chunk is given by the AUTH chunk with its
+		 *    HMAC field set to zero (as shown in Figure 6) followed by all
+		 *    chunks that are placed after the AUTH chunk in the SCTP packet.
+		 */
+		if (auth)
+			sctp_auth_calculate_hmac(asoc, nskb,
+						 (struct sctp_auth_chunk *)auth,
+						 gfp);
+
+		if (!gso)
+			break;
+
+		if (skb_gro_receive(&head, nskb))
+			goto nomem;
+		nskb = NULL;
+		if (WARN_ON_ONCE(skb_shinfo(head)->gso_segs >=
+				 sk->sk_gso_max_segs))
+			goto nomem;
+	} while (!list_empty(&packet->chunk_list));
 
 	/* 2) Calculate the Adler-32 checksum of the whole packet,
 	 *    including the SCTP common header and all the
@@ -532,16 +621,18 @@ int sctp_packet_transmit(struct sctp_packet *packet, gfp_t gfp)
 	 *
 	 * Note: Adler-32 is no longer applicable, as has been replaced
 	 * by CRC32-C as described in <draft-ietf-tsvwg-sctpcsum-02.txt>.
+	 *
+	 * If it's a GSO packet, it's postponed to sctp_skb_segment.
 	 */
-	if (!sctp_checksum_disable) {
-		if (!(dst->dev->features & NETIF_F_SCTP_CRC) ||
-		    (dst_xfrm(dst) != NULL) || packet->ipfragok) {
-			sh->checksum = sctp_compute_cksum(nskb, 0);
+	if (!sctp_checksum_disable || gso) {
+		if (!gso && (!(dst->dev->features & NETIF_F_SCTP_CRC) ||
+			     dst_xfrm(dst) || packet->ipfragok)) {
+			sh->checksum = sctp_compute_cksum(head, 0);
 		} else {
 			/* no need to seed pseudo checksum for SCTP */
-			nskb->ip_summed = CHECKSUM_PARTIAL;
-			nskb->csum_start = skb_transport_header(nskb) - nskb->head;
-			nskb->csum_offset = offsetof(struct sctphdr, checksum);
+			head->ip_summed = CHECKSUM_PARTIAL;
+			head->csum_start = skb_transport_header(head) - head->head;
+			head->csum_offset = offsetof(struct sctphdr, checksum);
 		}
 	}
 
@@ -557,7 +648,7 @@ int sctp_packet_transmit(struct sctp_packet *packet, gfp_t gfp)
 	 * Note: The works for IPv6 layer checks this bit too later
 	 * in transmission.  See IP6_ECN_flow_xmit().
 	 */
-	tp->af_specific->ecn_capable(nskb->sk);
+	tp->af_specific->ecn_capable(sk);
 
 	/* Set up the IP options.  */
 	/* BUG: not implemented
@@ -566,7 +657,7 @@ int sctp_packet_transmit(struct sctp_packet *packet, gfp_t gfp)
 
 	/* Dump that on IP!  */
 	if (asoc) {
-		asoc->stats.opackets++;
+		asoc->stats.opackets += pktcount;
 		if (asoc->peer.last_sent_to != tp)
 			/* Considering the multiple CPU scenario, this is a
 			 * "correcter" place for last_sent_to.  --xguo
@@ -589,16 +680,36 @@ int sctp_packet_transmit(struct sctp_packet *packet, gfp_t gfp)
 		}
 	}
 
-	pr_debug("***sctp_transmit_packet*** skb->len:%d\n", nskb->len);
+	pr_debug("***sctp_transmit_packet*** skb->len:%d\n", head->len);
+
+	if (gso) {
+		/* Cleanup our debris for IP stacks */
+		memset(head->cb, 0, max(sizeof(struct inet_skb_parm),
+					sizeof(struct inet6_skb_parm)));
 
-	nskb->ignore_df = packet->ipfragok;
-	tp->af_specific->sctp_xmit(nskb, tp);
+		skb_shinfo(head)->gso_segs = pktcount;
+		skb_shinfo(head)->gso_size = GSO_BY_FRAGS;
+
+		/* We have to refresh this in case we are xmiting to
+		 * more than one transport at a time
+		 */
+		rcu_read_lock();
+		if (__sk_dst_get(sk) != tp->dst) {
+			dst_hold(tp->dst);
+			sk_setup_caps(sk, tp->dst);
+		}
+		rcu_read_unlock();
+	}
+	head->ignore_df = packet->ipfragok;
+	tp->af_specific->sctp_xmit(head, tp);
 
 out:
 	sctp_packet_reset(packet);
 	return err;
 no_route:
-	kfree_skb(nskb);
+	kfree_skb(head);
+	if (nskb != head)
+		kfree_skb(nskb);
 
 	if (asoc)
 		IP_INC_STATS(sock_net(asoc->base.sk), IPSTATS_MIB_OUTNOROUTES);
@@ -751,39 +862,63 @@ static sctp_xmit_t sctp_packet_will_fit(struct sctp_packet *packet,
 					struct sctp_chunk *chunk,
 					u16 chunk_len)
 {
-	size_t psize;
-	size_t pmtu;
-	int too_big;
+	size_t psize, pmtu;
 	sctp_xmit_t retval = SCTP_XMIT_OK;
 
 	psize = packet->size;
-	pmtu  = ((packet->transport->asoc) ?
-		(packet->transport->asoc->pathmtu) :
-		(packet->transport->pathmtu));
-
-	too_big = (psize + chunk_len > pmtu);
+	if (packet->transport->asoc)
+		pmtu = packet->transport->asoc->pathmtu;
+	else
+		pmtu = packet->transport->pathmtu;
 
 	/* Decide if we need to fragment or resubmit later. */
-	if (too_big) {
-		/* It's OK to fragmet at IP level if any one of the following
+	if (psize + chunk_len > pmtu) {
+		/* It's OK to fragment at IP level if any one of the following
 		 * is true:
-		 * 	1. The packet is empty (meaning this chunk is greater
-		 * 	   the MTU)
-		 * 	2. The chunk we are adding is a control chunk
-		 * 	3. The packet doesn't have any data in it yet and data
-		 * 	requires authentication.
+		 *	1. The packet is empty (meaning this chunk is greater
+		 *	   the MTU)
+		 *	2. The packet doesn't have any data in it yet and data
+		 *	   requires authentication.
 		 */
-		if (sctp_packet_empty(packet) || !sctp_chunk_is_data(chunk) ||
+		if (sctp_packet_empty(packet) ||
 		    (!packet->has_data && chunk->auth)) {
 			/* We no longer do re-fragmentation.
 			 * Just fragment at the IP layer, if we
 			 * actually hit this condition
 			 */
 			packet->ipfragok = 1;
-		} else {
-			retval = SCTP_XMIT_PMTU_FULL;
+			goto out;
 		}
+
+		/* It is also okay to fragment if the chunk we are
+		 * adding is a control chunk, but only if current packet
+		 * is not a GSO one otherwise it causes fragmentation of
+		 * a large frame. So in this case we allow the
+		 * fragmentation by forcing it to be in a new packet.
+		 */
+		if (!sctp_chunk_is_data(chunk) && packet->has_data)
+			retval = SCTP_XMIT_PMTU_FULL;
+
+		if (psize + chunk_len > packet->max_size)
+			/* Hit GSO/PMTU limit, gotta flush */
+			retval = SCTP_XMIT_PMTU_FULL;
+
+		if (!packet->transport->burst_limited &&
+		    psize + chunk_len > (packet->transport->cwnd >> 1))
+			/* Do not allow a single GSO packet to use more
+			 * than half of cwnd.
+			 */
+			retval = SCTP_XMIT_PMTU_FULL;
+
+		if (packet->transport->burst_limited &&
+		    psize + chunk_len > (packet->transport->burst_limited >> 1))
+			/* Do not allow a single GSO packet to use more
+			 * than half of original cwnd.
+			 */
+			retval = SCTP_XMIT_PMTU_FULL;
+		/* Otherwise it will fit in the GSO packet */
 	}
 
+out:
 	return retval;
 }