1 files changed, 246 insertions, 126 deletions

diff --git a/net/tipc/msg.c b/net/tipc/msg.c
index f48e5857210f..76284fc538eb 100644
--- a/net/tipc/msg.c
+++ b/net/tipc/msg.c
@@ -39,31 +39,37 @@
 #include "msg.h"
 #include "addr.h"
 #include "name_table.h"
+#include "crypto.h"
 
+#define BUF_ALIGN(x) ALIGN(x, 4)
 #define MAX_FORWARD_SIZE 1024
+#ifdef CONFIG_TIPC_CRYPTO
+#define BUF_HEADROOM ALIGN(((LL_MAX_HEADER + 48) + EHDR_MAX_SIZE), 16)
+#define BUF_OVERHEAD (BUF_HEADROOM + TIPC_AES_GCM_TAG_SIZE)
+#else
 #define BUF_HEADROOM (LL_MAX_HEADER + 48)
-#define BUF_TAILROOM 16
+#define BUF_OVERHEAD BUF_HEADROOM
+#endif
 
-static unsigned int align(unsigned int i)
-{
-	return (i + 3) & ~3u;
-}
+const int one_page_mtu = PAGE_SIZE - SKB_DATA_ALIGN(BUF_OVERHEAD) -
+			 SKB_DATA_ALIGN(sizeof(struct skb_shared_info));
 
 /**
  * tipc_buf_acquire - creates a TIPC message buffer
  * @size: message size (including TIPC header)
+ * @gfp: memory allocation flags
  *
- * Returns a new buffer with data pointers set to the specified size.
+ * Return: a new buffer with data pointers set to the specified size.
  *
- * NOTE: Headroom is reserved to allow prepending of a data link header.
- *       There may also be unrequested tailroom present at the buffer's end.
+ * NOTE:
+ * Headroom is reserved to allow prepending of a data link header.
+ * There may also be unrequested tailroom present at the buffer's end.
  */
 struct sk_buff *tipc_buf_acquire(u32 size, gfp_t gfp)
 {
 	struct sk_buff *skb;
-	unsigned int buf_size = (BUF_HEADROOM + size + 3) & ~3u;
 
-	skb = alloc_skb_fclone(buf_size, gfp);
+	skb = alloc_skb_fclone(BUF_OVERHEAD + size, gfp);
 	if (skb) {
 		skb_reserve(skb, BUF_HEADROOM);
 		skb_put(skb, size);
@@ -105,10 +111,6 @@ struct sk_buff *tipc_msg_create(uint user, uint type,
 	msg_set_origport(msg, oport);
 	msg_set_destport(msg, dport);
 	msg_set_errcode(msg, errcode);
-	if (hdr_sz > SHORT_H_SIZE) {
-		msg_set_orignode(msg, onode);
-		msg_set_destnode(msg, dnode);
-	}
 	return buf;
 }
 
@@ -140,24 +142,25 @@ int tipc_buf_append(struct sk_buff **headbuf, struct sk_buff **buf)
 	if (fragid == FIRST_FRAGMENT) {
 		if (unlikely(head))
 			goto err;
-		if (unlikely(skb_unclone(frag, GFP_ATOMIC)))
+		if (skb_has_frag_list(frag) && __skb_linearize(frag))
 			goto err;
-		head = *headbuf = frag;
 		*buf = NULL;
+		frag = skb_unshare(frag, GFP_ATOMIC);
+		if (unlikely(!frag))
+			goto err;
+		head = *headbuf = frag;
 		TIPC_SKB_CB(head)->tail = NULL;
-		if (skb_is_nonlinear(head)) {
-			skb_walk_frags(head, tail) {
-				TIPC_SKB_CB(head)->tail = tail;
-			}
-		} else {
-			skb_frag_list_init(head);
-		}
 		return 0;
 	}
 
 	if (!head)
 		goto err;
 
+	/* Either the input skb ownership is transferred to headskb
+	 * or the input skb is freed, clear the reference to avoid
+	 * bad access on error path.
+	 */
+	*buf = NULL;
 	if (skb_try_coalesce(head, frag, &headstolen, &delta)) {
 		kfree_skb_partial(frag, headstolen);
 	} else {
@@ -173,7 +176,7 @@ int tipc_buf_append(struct sk_buff **headbuf, struct sk_buff **buf)
 	}
 
 	if (fragid == LAST_FRAGMENT) {
-		TIPC_SKB_CB(head)->validated = false;
+		TIPC_SKB_CB(head)->validated = 0;
 		if (unlikely(!tipc_msg_validate(&head)))
 			goto err;
 		*buf = head;
@@ -181,7 +184,6 @@ int tipc_buf_append(struct sk_buff **headbuf, struct sk_buff **buf)
 		*headbuf = NULL;
 		return 1;
 	}
-	*buf = NULL;
 	return 0;
 err:
 	kfree_skb(*buf);
@@ -190,6 +192,56 @@ err:
 	return 0;
 }
 
+/**
+ * tipc_msg_append(): Append data to tail of an existing buffer queue
+ * @_hdr: header to be used
+ * @m: the data to be appended
+ * @mss: max allowable size of buffer
+ * @dlen: size of data to be appended
+ * @txq: queue to append to
+ *
+ * Return: the number of 1k blocks appended or errno value
+ */
+int tipc_msg_append(struct tipc_msg *_hdr, struct msghdr *m, int dlen,
+		    int mss, struct sk_buff_head *txq)
+{
+	struct sk_buff *skb;
+	int accounted, total, curr;
+	int mlen, cpy, rem = dlen;
+	struct tipc_msg *hdr;
+
+	skb = skb_peek_tail(txq);
+	accounted = skb ? msg_blocks(buf_msg(skb)) : 0;
+	total = accounted;
+
+	do {
+		if (!skb || skb->len >= mss) {
+			skb = tipc_buf_acquire(mss, GFP_KERNEL);
+			if (unlikely(!skb))
+				return -ENOMEM;
+			skb_orphan(skb);
+			skb_trim(skb, MIN_H_SIZE);
+			hdr = buf_msg(skb);
+			skb_copy_to_linear_data(skb, _hdr, MIN_H_SIZE);
+			msg_set_hdr_sz(hdr, MIN_H_SIZE);
+			msg_set_size(hdr, MIN_H_SIZE);
+			__skb_queue_tail(txq, skb);
+			total += 1;
+		}
+		hdr = buf_msg(skb);
+		curr = msg_blocks(hdr);
+		mlen = msg_size(hdr);
+		cpy = min_t(size_t, rem, mss - mlen);
+		if (cpy != copy_from_iter(skb->data + mlen, cpy, &m->msg_iter))
+			return -EFAULT;
+		msg_set_size(hdr, mlen + cpy);
+		skb_put(skb, cpy);
+		rem -= cpy;
+		total += msg_blocks(hdr) - curr;
+	} while (rem > 0);
+	return total - accounted;
+}
+
 /* tipc_msg_validate - validate basic format of received message
  *
  * This routine ensures a TIPC message has an acceptable header, and at least
@@ -218,6 +270,7 @@ bool tipc_msg_validate(struct sk_buff **_skb)
 
 	if (unlikely(TIPC_SKB_CB(skb)->validated))
 		return true;
+
 	if (unlikely(!pskb_may_pull(skb, MIN_H_SIZE)))
 		return false;
 
@@ -239,14 +292,74 @@ bool tipc_msg_validate(struct sk_buff **_skb)
 	if (unlikely(skb->len < msz))
 		return false;
 
-	TIPC_SKB_CB(skb)->validated = true;
+	TIPC_SKB_CB(skb)->validated = 1;
 	return true;
 }
 
 /**
+ * tipc_msg_fragment - build a fragment skb list for TIPC message
+ *
+ * @skb: TIPC message skb
+ * @hdr: internal msg header to be put on the top of the fragments
+ * @pktmax: max size of a fragment incl. the header
+ * @frags: returned fragment skb list
+ *
+ * Return: 0 if the fragmentation is successful, otherwise: -EINVAL
+ * or -ENOMEM
+ */
+int tipc_msg_fragment(struct sk_buff *skb, const struct tipc_msg *hdr,
+		      int pktmax, struct sk_buff_head *frags)
+{
+	int pktno, nof_fragms, dsz, dmax, eat;
+	struct tipc_msg *_hdr;
+	struct sk_buff *_skb;
+	u8 *data;
+
+	/* Non-linear buffer? */
+	if (skb_linearize(skb))
+		return -ENOMEM;
+
+	data = (u8 *)skb->data;
+	dsz = msg_size(buf_msg(skb));
+	dmax = pktmax - INT_H_SIZE;
+	if (dsz <= dmax || !dmax)
+		return -EINVAL;
+
+	nof_fragms = dsz / dmax + 1;
+	for (pktno = 1; pktno <= nof_fragms; pktno++) {
+		if (pktno < nof_fragms)
+			eat = dmax;
+		else
+			eat = dsz % dmax;
+		/* Allocate a new fragment */
+		_skb = tipc_buf_acquire(INT_H_SIZE + eat, GFP_ATOMIC);
+		if (!_skb)
+			goto error;
+		skb_orphan(_skb);
+		__skb_queue_tail(frags, _skb);
+		/* Copy header & data to the fragment */
+		skb_copy_to_linear_data(_skb, hdr, INT_H_SIZE);
+		skb_copy_to_linear_data_offset(_skb, INT_H_SIZE, data, eat);
+		data += eat;
+		/* Update the fragment's header */
+		_hdr = buf_msg(_skb);
+		msg_set_fragm_no(_hdr, pktno);
+		msg_set_nof_fragms(_hdr, nof_fragms);
+		msg_set_size(_hdr, INT_H_SIZE + eat);
+	}
+	return 0;
+
+error:
+	__skb_queue_purge(frags);
+	__skb_queue_head_init(frags);
+	return -ENOMEM;
+}
+
+/**
  * tipc_msg_build - create buffer chain containing specified header and data
  * @mhdr: Message header, to be prepended to data
  * @m: User message
+ * @offset: buffer offset for fragmented messages (FIXME)
  * @dsz: Total length of user data
  * @pktmax: Max packet size that can be used
  * @list: Buffer or chain of buffers to be returned to caller
@@ -254,7 +367,7 @@ bool tipc_msg_validate(struct sk_buff **_skb)
  * Note that the recursive call we are making here is safe, since it can
  * logically go only one further level down.
  *
- * Returns message data size or errno: -ENOMEM, -EFAULT
+ * Return: message data size or errno: -ENOMEM, -EFAULT
  */
 int tipc_msg_build(struct tipc_msg *mhdr, struct msghdr *m, int offset,
 		   int dsz, int pktmax, struct sk_buff_head *list)
@@ -280,7 +393,8 @@ int tipc_msg_build(struct tipc_msg *mhdr, struct msghdr *m, int offset,
 		if (unlikely(!skb)) {
 			if (pktmax != MAX_MSG_SIZE)
 				return -ENOMEM;
-			rc = tipc_msg_build(mhdr, m, offset, dsz, FB_MTU, list);
+			rc = tipc_msg_build(mhdr, m, offset, dsz,
+					    one_page_mtu, list);
 			if (rc != dsz)
 				return rc;
 			if (tipc_msg_assemble(list))
@@ -360,48 +474,98 @@ error:
 }
 
 /**
- * tipc_msg_bundle(): Append contents of a buffer to tail of an existing one
- * @skb: the buffer to append to ("bundle")
- * @msg:  message to be appended
- * @mtu:  max allowable size for the bundle buffer
- * Consumes buffer if successful
- * Returns true if bundling could be performed, otherwise false
+ * tipc_msg_bundle - Append contents of a buffer to tail of an existing one
+ * @bskb: the bundle buffer to append to
+ * @msg: message to be appended
+ * @max: max allowable size for the bundle buffer
+ *
+ * Return: "true" if bundling has been performed, otherwise "false"
  */
-bool tipc_msg_bundle(struct sk_buff *skb, struct tipc_msg *msg, u32 mtu)
+static bool tipc_msg_bundle(struct sk_buff *bskb, struct tipc_msg *msg,
+			    u32 max)
 {
-	struct tipc_msg *bmsg;
-	unsigned int bsz;
-	unsigned int msz = msg_size(msg);
-	u32 start, pad;
-	u32 max = mtu - INT_H_SIZE;
+	struct tipc_msg *bmsg = buf_msg(bskb);
+	u32 msz, bsz, offset, pad;
 
-	if (likely(msg_user(msg) == MSG_FRAGMENTER))
-		return false;
-	if (!skb)
-		return false;
-	bmsg = buf_msg(skb);
+	msz = msg_size(msg);
 	bsz = msg_size(bmsg);
-	start = align(bsz);
-	pad = start - bsz;
+	offset = BUF_ALIGN(bsz);
+	pad = offset - bsz;
 
-	if (unlikely(msg_user(msg) == TUNNEL_PROTOCOL))
+	if (unlikely(skb_tailroom(bskb) < (pad + msz)))
 		return false;
-	if (unlikely(msg_user(msg) == BCAST_PROTOCOL))
+	if (unlikely(max < (offset + msz)))
 		return false;
-	if (unlikely(msg_user(bmsg) != MSG_BUNDLER))
+
+	skb_put(bskb, pad + msz);
+	skb_copy_to_linear_data_offset(bskb, offset, msg, msz);
+	msg_set_size(bmsg, offset + msz);
+	msg_set_msgcnt(bmsg, msg_msgcnt(bmsg) + 1);
+	return true;
+}
+
+/**
+ * tipc_msg_try_bundle - Try to bundle a new message to the last one
+ * @tskb: the last/target message to which the new one will be appended
+ * @skb: the new message skb pointer
+ * @mss: max message size (header inclusive)
+ * @dnode: destination node for the message
+ * @new_bundle: if this call made a new bundle or not
+ *
+ * Return: "true" if the new message skb is potential for bundling this time or
+ * later, in the case a bundling has been done this time, the skb is consumed
+ * (the skb pointer = NULL).
+ * Otherwise, "false" if the skb cannot be bundled at all.
+ */
+bool tipc_msg_try_bundle(struct sk_buff *tskb, struct sk_buff **skb, u32 mss,
+			 u32 dnode, bool *new_bundle)
+{
+	struct tipc_msg *msg, *inner, *outer;
+	u32 tsz;
+
+	/* First, check if the new buffer is suitable for bundling */
+	msg = buf_msg(*skb);
+	if (msg_user(msg) == MSG_FRAGMENTER)
 		return false;
-	if (unlikely(skb_tailroom(skb) < (pad + msz)))
+	if (msg_user(msg) == TUNNEL_PROTOCOL)
 		return false;
-	if (unlikely(max < (start + msz)))
+	if (msg_user(msg) == BCAST_PROTOCOL)
 		return false;
-	if ((msg_importance(msg) < TIPC_SYSTEM_IMPORTANCE) &&
-	    (msg_importance(bmsg) == TIPC_SYSTEM_IMPORTANCE))
+	if (mss <= INT_H_SIZE + msg_size(msg))
 		return false;
 
-	skb_put(skb, pad + msz);
-	skb_copy_to_linear_data_offset(skb, start, msg, msz);
-	msg_set_size(bmsg, start + msz);
-	msg_set_msgcnt(bmsg, msg_msgcnt(bmsg) + 1);
+	/* Ok, but the last/target buffer can be empty? */
+	if (unlikely(!tskb))
+		return true;
+
+	/* Is it a bundle already? Try to bundle the new message to it */
+	if (msg_user(buf_msg(tskb)) == MSG_BUNDLER) {
+		*new_bundle = false;
+		goto bundle;
+	}
+
+	/* Make a new bundle of the two messages if possible */
+	tsz = msg_size(buf_msg(tskb));
+	if (unlikely(mss < BUF_ALIGN(INT_H_SIZE + tsz) + msg_size(msg)))
+		return true;
+	if (unlikely(pskb_expand_head(tskb, INT_H_SIZE, mss - tsz - INT_H_SIZE,
+				      GFP_ATOMIC)))
+		return true;
+	inner = buf_msg(tskb);
+	skb_push(tskb, INT_H_SIZE);
+	outer = buf_msg(tskb);
+	tipc_msg_init(msg_prevnode(inner), outer, MSG_BUNDLER, 0, INT_H_SIZE,
+		      dnode);
+	msg_set_importance(outer, msg_importance(inner));
+	msg_set_size(outer, INT_H_SIZE + tsz);
+	msg_set_msgcnt(outer, 1);
+	*new_bundle = true;
+
+bundle:
+	if (likely(tipc_msg_bundle(tskb, msg, mss))) {
+		consume_skb(*skb);
+		*skb = NULL;
+	}
 	return true;
 }
 
@@ -410,9 +574,9 @@ bool tipc_msg_bundle(struct sk_buff *skb, struct tipc_msg *msg, u32 mtu)
  *  @skb: buffer to be extracted from.
  *  @iskb: extracted inner buffer, to be returned
  *  @pos: position in outer message of msg to be extracted.
- *        Returns position of next msg
+ *  Returns position of next msg.
  *  Consumes outer buffer when last packet extracted
- *  Returns true when when there is an extracted buffer, otherwise false
+ *  Return: true when there is an extracted buffer, otherwise false
  */
 bool tipc_msg_extract(struct sk_buff *skb, struct sk_buff **iskb, int *pos)
 {
@@ -441,7 +605,7 @@ bool tipc_msg_extract(struct sk_buff *skb, struct sk_buff **iskb, int *pos)
 	if (unlikely(!tipc_msg_validate(iskb)))
 		goto none;
 
-	*pos += align(imsz);
+	*pos += BUF_ALIGN(imsz);
 	return true;
 none:
 	kfree_skb(skb);
@@ -451,58 +615,12 @@ none:
 }
 
 /**
- * tipc_msg_make_bundle(): Create bundle buf and append message to its tail
- * @list: the buffer chain, where head is the buffer to replace/append
- * @skb: buffer to be created, appended to and returned in case of success
- * @msg: message to be appended
- * @mtu: max allowable size for the bundle buffer, inclusive header
- * @dnode: destination node for message. (Not always present in header)
- * Returns true if success, otherwise false
- */
-bool tipc_msg_make_bundle(struct sk_buff **skb,  struct tipc_msg *msg,
-			  u32 mtu, u32 dnode)
-{
-	struct sk_buff *_skb;
-	struct tipc_msg *bmsg;
-	u32 msz = msg_size(msg);
-	u32 max = mtu - INT_H_SIZE;
-
-	if (msg_user(msg) == MSG_FRAGMENTER)
-		return false;
-	if (msg_user(msg) == TUNNEL_PROTOCOL)
-		return false;
-	if (msg_user(msg) == BCAST_PROTOCOL)
-		return false;
-	if (msz > (max / 2))
-		return false;
-
-	_skb = tipc_buf_acquire(max, GFP_ATOMIC);
-	if (!_skb)
-		return false;
-
-	skb_trim(_skb, INT_H_SIZE);
-	bmsg = buf_msg(_skb);
-	tipc_msg_init(msg_prevnode(msg), bmsg, MSG_BUNDLER, 0,
-		      INT_H_SIZE, dnode);
-	if (msg_isdata(msg))
-		msg_set_importance(bmsg, TIPC_CRITICAL_IMPORTANCE);
-	else
-		msg_set_importance(bmsg, TIPC_SYSTEM_IMPORTANCE);
-	msg_set_seqno(bmsg, msg_seqno(msg));
-	msg_set_ack(bmsg, msg_ack(msg));
-	msg_set_bcast_ack(bmsg, msg_bcast_ack(msg));
-	tipc_msg_bundle(_skb, msg, mtu);
-	*skb = _skb;
-	return true;
-}
-
-/**
  * tipc_msg_reverse(): swap source and destination addresses and add error code
  * @own_node: originating node id for reversed message
  * @skb:  buffer containing message to be reversed; will be consumed
  * @err:  error code to be set in message, if any
  * Replaces consumed buffer with new one when successful
- * Returns true if success, otherwise false
+ * Return: true if success, otherwise false
  */
 bool tipc_msg_reverse(u32 own_node,  struct sk_buff **skb, int err)
 {
@@ -574,16 +692,20 @@ bool tipc_msg_skb_clone(struct sk_buff_head *msg, struct sk_buff_head *cpy)
 
 /**
  * tipc_msg_lookup_dest(): try to find new destination for named message
+ * @net: pointer to associated network namespace
  * @skb: the buffer containing the message.
  * @err: error code to be used by caller if lookup fails
  * Does not consume buffer
- * Returns true if a destination is found, false otherwise
+ * Return: true if a destination is found, false otherwise
  */
 bool tipc_msg_lookup_dest(struct net *net, struct sk_buff *skb, int *err)
 {
 	struct tipc_msg *msg = buf_msg(skb);
-	u32 dport, dnode;
-	u32 onode = tipc_own_addr(net);
+	u32 scope = msg_lookup_scope(msg);
+	u32 self = tipc_own_addr(net);
+	u32 inst = msg_nameinst(msg);
+	struct tipc_socket_addr sk;
+	struct tipc_uaddr ua;
 
 	if (!msg_isdata(msg))
 		return false;
@@ -597,21 +719,18 @@ bool tipc_msg_lookup_dest(struct net *net, struct sk_buff *skb, int *err)
 	msg = buf_msg(skb);
 	if (msg_reroute_cnt(msg))
 		return false;
-	dnode = tipc_scope2node(net, msg_lookup_scope(msg));
-	dport = tipc_nametbl_translate(net, msg_nametype(msg),
-				       msg_nameinst(msg), &dnode);
-	if (!dport)
+	tipc_uaddr(&ua, TIPC_SERVICE_RANGE, scope,
+		   msg_nametype(msg), inst, inst);
+	sk.node = tipc_scope2node(net, scope);
+	if (!tipc_nametbl_lookup_anycast(net, &ua, &sk))
 		return false;
 	msg_incr_reroute_cnt(msg);
-	if (dnode != onode)
-		msg_set_prevnode(msg, onode);
-	msg_set_destnode(msg, dnode);
-	msg_set_destport(msg, dport);
+	if (sk.node != self)
+		msg_set_prevnode(msg, self);
+	msg_set_destnode(msg, sk.node);
+	msg_set_destport(msg, sk.ref);
 	*err = TIPC_OK;
 
-	if (!skb_cloned(skb))
-		return true;
-
 	return true;
 }
 
@@ -701,19 +820,19 @@ bool tipc_msg_pskb_copy(u32 dst, struct sk_buff_head *msg,
  * @seqno: sequence number of buffer to add
  * @skb: buffer to add
  */
-void __tipc_skb_queue_sorted(struct sk_buff_head *list, u16 seqno,
+bool __tipc_skb_queue_sorted(struct sk_buff_head *list, u16 seqno,
 			     struct sk_buff *skb)
 {
 	struct sk_buff *_skb, *tmp;
 
 	if (skb_queue_empty(list) || less(seqno, buf_seqno(skb_peek(list)))) {
 		__skb_queue_head(list, skb);
-		return;
+		return true;
 	}
 
 	if (more(seqno, buf_seqno(skb_peek_tail(list)))) {
 		__skb_queue_tail(list, skb);
-		return;
+		return true;
 	}
 
 	skb_queue_walk_safe(list, _skb, tmp) {
@@ -722,9 +841,10 @@ void __tipc_skb_queue_sorted(struct sk_buff_head *list, u16 seqno,
 		if (seqno == buf_seqno(_skb))
 			break;
 		__skb_queue_before(list, _skb, skb);
-		return;
+		return true;
 	}
 	kfree_skb(skb);
+	return false;
 }
 
 void tipc_skb_reject(struct net *net, int err, struct sk_buff *skb,