Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net

Pull networking fixes from David Miller:
 "This pull request is dedicated to the upcoming snowpocalypse parts 2
  and 3 in the Pacific Northwest:

   1) Drop profiles are broken because some drivers use dev_kfree_skb*
      instead of dev_consume_skb*, from Yang Wei.

   2) Fix IWLWIFI kconfig deps, from Luca Coelho.

   3) Fix percpu maps updating in bpftool, from Paolo Abeni.

   4) Missing station release in batman-adv, from Felix Fietkau.

   5) Fix some networking compat ioctl bugs, from Johannes Berg.

   6) ucc_geth must reset the BQL queue state when stopping the device,
      from Mathias Thore.

   7) Several XDP bug fixes in virtio_net from Toshiaki Makita.

   8) TSO packets must be sent always on queue 0 in stmmac, from Jose
      Abreu.

   9) Fix socket refcounting bug in RDS, from Eric Dumazet.

  10) Handle sparse cpu allocations in bpf selftests, from Martynas
      Pumputis.

  11) Make sure mgmt frames have enough tailroom in mac80211, from Felix
      Feitkau.

  12) Use safe list walking in sctp_sendmsg() asoc list traversal, from
      Greg Kroah-Hartman.

  13) Make DCCP's ccid_hc_[rt]x_parse_options always check for NULL
      ccid, from Eric Dumazet.

  14) Need to reload WoL password into bcmsysport device after deep
      sleeps, from Florian Fainelli.

  15) Remove filter from mask before freeing in cls_flower, from Petr
      Machata.

  16) Missing release and use after free in error paths of s390 qeth
      code, from Julian Wiedmann.

  17) Fix lockdep false positive in dsa code, from Marc Zyngier.

  18) Fix counting of ATU violations in mv88e6xxx, from Andrew Lunn.

  19) Fix EQ firmware assert in qed driver, from Manish Chopra.

  20) Don't default Caivum PTP to Y in kconfig, from Bjorn Helgaas"

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (116 commits)
  net: dsa: b53: Fix for failure when irq is not defined in dt
  sit: check if IPv6 enabled before calling ip6_err_gen_icmpv6_unreach()
  geneve: should not call rt6_lookup() when ipv6 was disabled
  net: Don't default Cavium PTP driver to 'y'
  net: broadcom: replace dev_kfree_skb_irq by dev_consume_skb_irq for drop profiles
  net: via-velocity: replace dev_kfree_skb_irq by dev_consume_skb_irq for drop profiles
  net: tehuti: replace dev_kfree_skb_irq by dev_consume_skb_irq for drop profiles
  net: sun: replace dev_kfree_skb_irq by dev_consume_skb_irq for drop profiles
  net: fsl_ucc_hdlc: replace dev_kfree_skb_irq by dev_consume_skb_irq for drop profiles
  net: fec_mpc52xx: replace dev_kfree_skb_irq by dev_consume_skb_irq for drop profiles
  net: smsc: epic100: replace dev_kfree_skb_irq by dev_consume_skb_irq for drop profiles
  net: dscc4: replace dev_kfree_skb_irq by dev_consume_skb_irq for drop profiles
  net: tulip: de2104x: replace dev_kfree_skb_irq by dev_consume_skb_irq for drop profiles
  net: defxx: replace dev_kfree_skb_irq by dev_consume_skb_irq for drop profiles
  net/mlx5e: Don't overwrite pedit action when multiple pedit used
  net/mlx5e: Update hw flows when encap source mac changed
  qed*: Advance drivers version to 8.37.0.20
  qed: Change verbosity for coalescing message.
  qede: Fix system crash on configuring channels.
  qed: Consider TX tcs while deriving the max num_queues for PF.
  ...

13 files changed, 158 insertions, 67 deletions

diff --git a/net/smc/af_smc.c b/net/smc/af_smc.c
index c4e56602e0c6..b04a813fc865 100644
--- a/net/smc/af_smc.c
+++ b/net/smc/af_smc.c
@@ -1505,6 +1505,11 @@ static int smc_recvmsg(struct socket *sock, struct msghdr *msg, size_t len,
 
 	smc = smc_sk(sk);
 	lock_sock(sk);
+	if (sk->sk_state == SMC_CLOSED && (sk->sk_shutdown & RCV_SHUTDOWN)) {
+		/* socket was connected before, no more data to read */
+		rc = 0;
+		goto out;
+	}
 	if ((sk->sk_state == SMC_INIT) ||
 	    (sk->sk_state == SMC_LISTEN) ||
 	    (sk->sk_state == SMC_CLOSED))
@@ -1840,7 +1845,11 @@ static ssize_t smc_splice_read(struct socket *sock, loff_t *ppos,
 
 	smc = smc_sk(sk);
 	lock_sock(sk);
-
+	if (sk->sk_state == SMC_CLOSED && (sk->sk_shutdown & RCV_SHUTDOWN)) {
+		/* socket was connected before, no more data to read */
+		rc = 0;
+		goto out;
+	}
 	if (sk->sk_state == SMC_INIT ||
 	    sk->sk_state == SMC_LISTEN ||
 	    sk->sk_state == SMC_CLOSED)
diff --git a/net/smc/smc_cdc.c b/net/smc/smc_cdc.c
index db83332ac1c8..a712c9f8699b 100644
--- a/net/smc/smc_cdc.c
+++ b/net/smc/smc_cdc.c
@@ -21,13 +21,6 @@
 
 /********************************** send *************************************/
 
-struct smc_cdc_tx_pend {
-	struct smc_connection	*conn;		/* socket connection */
-	union smc_host_cursor	cursor;	/* tx sndbuf cursor sent */
-	union smc_host_cursor	p_cursor;	/* rx RMBE cursor produced */
-	u16			ctrl_seq;	/* conn. tx sequence # */
-};
-
 /* handler for send/transmission completion of a CDC msg */
 static void smc_cdc_tx_handler(struct smc_wr_tx_pend_priv *pnd_snd,
 			       struct smc_link *link,
@@ -61,12 +54,14 @@ static void smc_cdc_tx_handler(struct smc_wr_tx_pend_priv *pnd_snd,
 
 int smc_cdc_get_free_slot(struct smc_connection *conn,
 			  struct smc_wr_buf **wr_buf,
+			  struct smc_rdma_wr **wr_rdma_buf,
 			  struct smc_cdc_tx_pend **pend)
 {
 	struct smc_link *link = &conn->lgr->lnk[SMC_SINGLE_LINK];
 	int rc;
 
 	rc = smc_wr_tx_get_free_slot(link, smc_cdc_tx_handler, wr_buf,
+				     wr_rdma_buf,
 				     (struct smc_wr_tx_pend_priv **)pend);
 	if (!conn->alert_token_local)
 		/* abnormal termination */
@@ -96,6 +91,7 @@ int smc_cdc_msg_send(struct smc_connection *conn,
 		     struct smc_wr_buf *wr_buf,
 		     struct smc_cdc_tx_pend *pend)
 {
+	union smc_host_cursor cfed;
 	struct smc_link *link;
 	int rc;
 
@@ -107,10 +103,10 @@ int smc_cdc_msg_send(struct smc_connection *conn,
 	conn->local_tx_ctrl.seqno = conn->tx_cdc_seq;
 	smc_host_msg_to_cdc((struct smc_cdc_msg *)wr_buf,
 			    &conn->local_tx_ctrl, conn);
+	smc_curs_copy(&cfed, &((struct smc_host_cdc_msg *)wr_buf)->cons, conn);
 	rc = smc_wr_tx_send(link, (struct smc_wr_tx_pend_priv *)pend);
 	if (!rc)
-		smc_curs_copy(&conn->rx_curs_confirmed,
-			      &conn->local_tx_ctrl.cons, conn);
+		smc_curs_copy(&conn->rx_curs_confirmed, &cfed, conn);
 
 	return rc;
 }
@@ -121,11 +117,14 @@ static int smcr_cdc_get_slot_and_msg_send(struct smc_connection *conn)
 	struct smc_wr_buf *wr_buf;
 	int rc;
 
-	rc = smc_cdc_get_free_slot(conn, &wr_buf, &pend);
+	rc = smc_cdc_get_free_slot(conn, &wr_buf, NULL, &pend);
 	if (rc)
 		return rc;
 
-	return smc_cdc_msg_send(conn, wr_buf, pend);
+	spin_lock_bh(&conn->send_lock);
+	rc = smc_cdc_msg_send(conn, wr_buf, pend);
+	spin_unlock_bh(&conn->send_lock);
+	return rc;
 }
 
 int smc_cdc_get_slot_and_msg_send(struct smc_connection *conn)
diff --git a/net/smc/smc_cdc.h b/net/smc/smc_cdc.h
index b5bfe38c7f9b..271e2524dc8f 100644
--- a/net/smc/smc_cdc.h
+++ b/net/smc/smc_cdc.h
@@ -160,7 +160,9 @@ static inline void smcd_curs_copy(union smcd_cdc_cursor *tgt,
 #endif
 }
 
-/* calculate cursor difference between old and new, where old <= new */
+/* calculate cursor difference between old and new, where old <= new and
+ * difference cannot exceed size
+ */
 static inline int smc_curs_diff(unsigned int size,
 				union smc_host_cursor *old,
 				union smc_host_cursor *new)
@@ -185,6 +187,28 @@ static inline int smc_curs_comp(unsigned int size,
 	return smc_curs_diff(size, old, new);
 }
 
+/* calculate cursor difference between old and new, where old <= new and
+ * difference may exceed size
+ */
+static inline int smc_curs_diff_large(unsigned int size,
+				      union smc_host_cursor *old,
+				      union smc_host_cursor *new)
+{
+	if (old->wrap < new->wrap)
+		return min_t(int,
+			     (size - old->count) + new->count +
+			     (new->wrap - old->wrap - 1) * size,
+			     size);
+
+	if (old->wrap > new->wrap) /* wrap has switched from 0xffff to 0x0000 */
+		return min_t(int,
+			     (size - old->count) + new->count +
+			     (new->wrap + 0xffff - old->wrap) * size,
+			     size);
+
+	return max_t(int, 0, (new->count - old->count));
+}
+
 static inline void smc_host_cursor_to_cdc(union smc_cdc_cursor *peer,
 					  union smc_host_cursor *local,
 					  struct smc_connection *conn)
@@ -270,10 +294,16 @@ static inline void smc_cdc_msg_to_host(struct smc_host_cdc_msg *local,
 		smcr_cdc_msg_to_host(local, peer, conn);
 }
 
-struct smc_cdc_tx_pend;
+struct smc_cdc_tx_pend {
+	struct smc_connection	*conn;		/* socket connection */
+	union smc_host_cursor	cursor;		/* tx sndbuf cursor sent */
+	union smc_host_cursor	p_cursor;	/* rx RMBE cursor produced */
+	u16			ctrl_seq;	/* conn. tx sequence # */
+};
 
 int smc_cdc_get_free_slot(struct smc_connection *conn,
 			  struct smc_wr_buf **wr_buf,
+			  struct smc_rdma_wr **wr_rdma_buf,
 			  struct smc_cdc_tx_pend **pend);
 void smc_cdc_tx_dismiss_slots(struct smc_connection *conn);
 int smc_cdc_msg_send(struct smc_connection *conn, struct smc_wr_buf *wr_buf,
diff --git a/net/smc/smc_clc.c b/net/smc/smc_clc.c
index 776e9dfc915d..d53fd588d1f5 100644
--- a/net/smc/smc_clc.c
+++ b/net/smc/smc_clc.c
@@ -378,7 +378,7 @@ int smc_clc_send_decline(struct smc_sock *smc, u32 peer_diag_info)
 	vec.iov_len = sizeof(struct smc_clc_msg_decline);
 	len = kernel_sendmsg(smc->clcsock, &msg, &vec, 1,
 			     sizeof(struct smc_clc_msg_decline));
-	if (len < sizeof(struct smc_clc_msg_decline))
+	if (len < 0 || len < sizeof(struct smc_clc_msg_decline))
 		len = -EPROTO;
 	return len > 0 ? 0 : len;
 }
diff --git a/net/smc/smc_close.c b/net/smc/smc_close.c
index ea2b87f29469..e39cadda1bf5 100644
--- a/net/smc/smc_close.c
+++ b/net/smc/smc_close.c
@@ -345,14 +345,7 @@ static void smc_close_passive_work(struct work_struct *work)
 
 	switch (sk->sk_state) {
 	case SMC_INIT:
-		if (atomic_read(&conn->bytes_to_rcv) ||
-		    (rxflags->peer_done_writing &&
-		     !smc_cdc_rxed_any_close(conn))) {
-			sk->sk_state = SMC_APPCLOSEWAIT1;
-		} else {
-			sk->sk_state = SMC_CLOSED;
-			sock_put(sk); /* passive closing */
-		}
+		sk->sk_state = SMC_APPCLOSEWAIT1;
 		break;
 	case SMC_ACTIVE:
 		sk->sk_state = SMC_APPCLOSEWAIT1;
diff --git a/net/smc/smc_core.c b/net/smc/smc_core.c
index 35c1cdc93e1c..aa1c551cee81 100644
--- a/net/smc/smc_core.c
+++ b/net/smc/smc_core.c
@@ -128,6 +128,8 @@ static void smc_lgr_unregister_conn(struct smc_connection *conn)
 {
 	struct smc_link_group *lgr = conn->lgr;
 
+	if (!lgr)
+		return;
 	write_lock_bh(&lgr->conns_lock);
 	if (conn->alert_token_local) {
 		__smc_lgr_unregister_conn(conn);
@@ -300,13 +302,13 @@ static void smc_buf_unuse(struct smc_connection *conn,
 		conn->sndbuf_desc->used = 0;
 	if (conn->rmb_desc) {
 		if (!conn->rmb_desc->regerr) {
-			conn->rmb_desc->used = 0;
 			if (!lgr->is_smcd) {
 				/* unregister rmb with peer */
 				smc_llc_do_delete_rkey(
 						&lgr->lnk[SMC_SINGLE_LINK],
 						conn->rmb_desc);
 			}
+			conn->rmb_desc->used = 0;
 		} else {
 			/* buf registration failed, reuse not possible */
 			write_lock_bh(&lgr->rmbs_lock);
@@ -628,6 +630,8 @@ int smc_conn_create(struct smc_sock *smc, bool is_smcd, int srv_first_contact,
 			local_contact = SMC_REUSE_CONTACT;
 			conn->lgr = lgr;
 			smc_lgr_register_conn(conn); /* add smc conn to lgr */
+			if (delayed_work_pending(&lgr->free_work))
+				cancel_delayed_work(&lgr->free_work);
 			write_unlock_bh(&lgr->conns_lock);
 			break;
 		}
diff --git a/net/smc/smc_core.h b/net/smc/smc_core.h
index b00287989a3d..8806d2afa6ed 100644
--- a/net/smc/smc_core.h
+++ b/net/smc/smc_core.h
@@ -52,6 +52,24 @@ enum smc_wr_reg_state {
 	FAILED		/* ib_wr_reg_mr response: failure */
 };
 
+struct smc_rdma_sge {				/* sges for RDMA writes */
+	struct ib_sge		wr_tx_rdma_sge[SMC_IB_MAX_SEND_SGE];
+};
+
+#define SMC_MAX_RDMA_WRITES	2		/* max. # of RDMA writes per
+						 * message send
+						 */
+
+struct smc_rdma_sges {				/* sges per message send */
+	struct smc_rdma_sge	tx_rdma_sge[SMC_MAX_RDMA_WRITES];
+};
+
+struct smc_rdma_wr {				/* work requests per message
+						 * send
+						 */
+	struct ib_rdma_wr	wr_tx_rdma[SMC_MAX_RDMA_WRITES];
+};
+
 struct smc_link {
 	struct smc_ib_device	*smcibdev;	/* ib-device */
 	u8			ibport;		/* port - values 1 | 2 */
@@ -64,6 +82,8 @@ struct smc_link {
 	struct smc_wr_buf	*wr_tx_bufs;	/* WR send payload buffers */
 	struct ib_send_wr	*wr_tx_ibs;	/* WR send meta data */
 	struct ib_sge		*wr_tx_sges;	/* WR send gather meta data */
+	struct smc_rdma_sges	*wr_tx_rdma_sges;/*RDMA WRITE gather meta data*/
+	struct smc_rdma_wr	*wr_tx_rdmas;	/* WR RDMA WRITE */
 	struct smc_wr_tx_pend	*wr_tx_pends;	/* WR send waiting for CQE */
 	/* above four vectors have wr_tx_cnt elements and use the same index */
 	dma_addr_t		wr_tx_dma_addr;	/* DMA address of wr_tx_bufs */
diff --git a/net/smc/smc_ib.c b/net/smc/smc_ib.c
index e519ef29c0ff..76487a16934e 100644
--- a/net/smc/smc_ib.c
+++ b/net/smc/smc_ib.c
@@ -289,8 +289,8 @@ int smc_ib_create_protection_domain(struct smc_link *lnk)
 
 static void smc_ib_qp_event_handler(struct ib_event *ibevent, void *priv)
 {
-	struct smc_ib_device *smcibdev =
-		(struct smc_ib_device *)ibevent->device;
+	struct smc_link *lnk = (struct smc_link *)priv;
+	struct smc_ib_device *smcibdev = lnk->smcibdev;
 	u8 port_idx;
 
 	switch (ibevent->event) {
@@ -298,7 +298,7 @@ static void smc_ib_qp_event_handler(struct ib_event *ibevent, void *priv)
 	case IB_EVENT_GID_CHANGE:
 	case IB_EVENT_PORT_ERR:
 	case IB_EVENT_QP_ACCESS_ERR:
-		port_idx = ibevent->element.port_num - 1;
+		port_idx = ibevent->element.qp->port - 1;
 		set_bit(port_idx, &smcibdev->port_event_mask);
 		schedule_work(&smcibdev->port_event_work);
 		break;
diff --git a/net/smc/smc_llc.c b/net/smc/smc_llc.c
index a6d3623d06f4..4fd60c522802 100644
--- a/net/smc/smc_llc.c
+++ b/net/smc/smc_llc.c
@@ -166,7 +166,8 @@ static int smc_llc_add_pending_send(struct smc_link *link,
 {
 	int rc;
 
-	rc = smc_wr_tx_get_free_slot(link, smc_llc_tx_handler, wr_buf, pend);
+	rc = smc_wr_tx_get_free_slot(link, smc_llc_tx_handler, wr_buf, NULL,
+				     pend);
 	if (rc < 0)
 		return rc;
 	BUILD_BUG_ON_MSG(
diff --git a/net/smc/smc_pnet.c b/net/smc/smc_pnet.c
index 7cb3e4f07c10..632c3109dee5 100644
--- a/net/smc/smc_pnet.c
+++ b/net/smc/smc_pnet.c
@@ -27,7 +27,7 @@
 static struct nla_policy smc_pnet_policy[SMC_PNETID_MAX + 1] = {
 	[SMC_PNETID_NAME] = {
 		.type = NLA_NUL_STRING,
-		.len = SMC_MAX_PNETID_LEN - 1
+		.len = SMC_MAX_PNETID_LEN
 	},
 	[SMC_PNETID_ETHNAME] = {
 		.type = NLA_NUL_STRING,
diff --git a/net/smc/smc_tx.c b/net/smc/smc_tx.c
index d8366ed51757..f93f3580c100 100644
--- a/net/smc/smc_tx.c
+++ b/net/smc/smc_tx.c
@@ -165,12 +165,11 @@ int smc_tx_sendmsg(struct smc_sock *smc, struct msghdr *msg, size_t len)
 			conn->local_tx_ctrl.prod_flags.urg_data_pending = 1;
 
 		if (!atomic_read(&conn->sndbuf_space) || conn->urg_tx_pend) {
+			if (send_done)
+				return send_done;
 			rc = smc_tx_wait(smc, msg->msg_flags);
-			if (rc) {
-				if (send_done)
-					return send_done;
+			if (rc)
 				goto out_err;
-			}
 			continue;
 		}
 
@@ -267,27 +266,23 @@ int smcd_tx_ism_write(struct smc_connection *conn, void *data, size_t len,
 
 /* sndbuf consumer: actual data transfer of one target chunk with RDMA write */
 static int smc_tx_rdma_write(struct smc_connection *conn, int peer_rmbe_offset,
-			     int num_sges, struct ib_sge sges[])
+			     int num_sges, struct ib_rdma_wr *rdma_wr)
 {
 	struct smc_link_group *lgr = conn->lgr;
-	struct ib_rdma_wr rdma_wr;
 	struct smc_link *link;
 	int rc;
 
-	memset(&rdma_wr, 0, sizeof(rdma_wr));
 	link = &lgr->lnk[SMC_SINGLE_LINK];
-	rdma_wr.wr.wr_id = smc_wr_tx_get_next_wr_id(link);
-	rdma_wr.wr.sg_list = sges;
-	rdma_wr.wr.num_sge = num_sges;
-	rdma_wr.wr.opcode = IB_WR_RDMA_WRITE;
-	rdma_wr.remote_addr =
+	rdma_wr->wr.wr_id = smc_wr_tx_get_next_wr_id(link);
+	rdma_wr->wr.num_sge = num_sges;
+	rdma_wr->remote_addr =
 		lgr->rtokens[conn->rtoken_idx][SMC_SINGLE_LINK].dma_addr +
 		/* RMBE within RMB */
 		conn->tx_off +
 		/* offset within RMBE */
 		peer_rmbe_offset;
-	rdma_wr.rkey = lgr->rtokens[conn->rtoken_idx][SMC_SINGLE_LINK].rkey;
-	rc = ib_post_send(link->roce_qp, &rdma_wr.wr, NULL);
+	rdma_wr->rkey = lgr->rtokens[conn->rtoken_idx][SMC_SINGLE_LINK].rkey;
+	rc = ib_post_send(link->roce_qp, &rdma_wr->wr, NULL);
 	if (rc) {
 		conn->local_tx_ctrl.conn_state_flags.peer_conn_abort = 1;
 		smc_lgr_terminate(lgr);
@@ -314,24 +309,25 @@ static inline void smc_tx_advance_cursors(struct smc_connection *conn,
 /* SMC-R helper for smc_tx_rdma_writes() */
 static int smcr_tx_rdma_writes(struct smc_connection *conn, size_t len,
 			       size_t src_off, size_t src_len,
-			       size_t dst_off, size_t dst_len)
+			       size_t dst_off, size_t dst_len,
+			       struct smc_rdma_wr *wr_rdma_buf)
 {
 	dma_addr_t dma_addr =
 		sg_dma_address(conn->sndbuf_desc->sgt[SMC_SINGLE_LINK].sgl);
-	struct smc_link *link = &conn->lgr->lnk[SMC_SINGLE_LINK];
 	int src_len_sum = src_len, dst_len_sum = dst_len;
-	struct ib_sge sges[SMC_IB_MAX_SEND_SGE];
 	int sent_count = src_off;
 	int srcchunk, dstchunk;
 	int num_sges;
 	int rc;
 
 	for (dstchunk = 0; dstchunk < 2; dstchunk++) {
+		struct ib_sge *sge =
+			wr_rdma_buf->wr_tx_rdma[dstchunk].wr.sg_list;
+
 		num_sges = 0;
 		for (srcchunk = 0; srcchunk < 2; srcchunk++) {
-			sges[srcchunk].addr = dma_addr + src_off;
-			sges[srcchunk].length = src_len;
-			sges[srcchunk].lkey = link->roce_pd->local_dma_lkey;
+			sge[srcchunk].addr = dma_addr + src_off;
+			sge[srcchunk].length = src_len;
 			num_sges++;
 
 			src_off += src_len;
@@ -344,7 +340,8 @@ static int smcr_tx_rdma_writes(struct smc_connection *conn, size_t len,
 			src_len = dst_len - src_len; /* remainder */
 			src_len_sum += src_len;
 		}
-		rc = smc_tx_rdma_write(conn, dst_off, num_sges, sges);
+		rc = smc_tx_rdma_write(conn, dst_off, num_sges,
+				       &wr_rdma_buf->wr_tx_rdma[dstchunk]);
 		if (rc)
 			return rc;
 		if (dst_len_sum == len)
@@ -403,7 +400,8 @@ static int smcd_tx_rdma_writes(struct smc_connection *conn, size_t len,
 /* sndbuf consumer: prepare all necessary (src&dst) chunks of data transmit;
  * usable snd_wnd as max transmit
  */
-static int smc_tx_rdma_writes(struct smc_connection *conn)
+static int smc_tx_rdma_writes(struct smc_connection *conn,
+			      struct smc_rdma_wr *wr_rdma_buf)
 {
 	size_t len, src_len, dst_off, dst_len; /* current chunk values */
 	union smc_host_cursor sent, prep, prod, cons;
@@ -464,7 +462,7 @@ static int smc_tx_rdma_writes(struct smc_connection *conn)
 					 dst_off, dst_len);
 	else
 		rc = smcr_tx_rdma_writes(conn, len, sent.count, src_len,
-					 dst_off, dst_len);
+					 dst_off, dst_len, wr_rdma_buf);
 	if (rc)
 		return rc;
 
@@ -485,31 +483,30 @@ static int smc_tx_rdma_writes(struct smc_connection *conn)
 static int smcr_tx_sndbuf_nonempty(struct smc_connection *conn)
 {
 	struct smc_cdc_producer_flags *pflags;
+	struct smc_rdma_wr *wr_rdma_buf;
 	struct smc_cdc_tx_pend *pend;
 	struct smc_wr_buf *wr_buf;
 	int rc;
 
-	spin_lock_bh(&conn->send_lock);
-	rc = smc_cdc_get_free_slot(conn, &wr_buf, &pend);
+	rc = smc_cdc_get_free_slot(conn, &wr_buf, &wr_rdma_buf, &pend);
 	if (rc < 0) {
 		if (rc == -EBUSY) {
 			struct smc_sock *smc =
 				container_of(conn, struct smc_sock, conn);
 
-			if (smc->sk.sk_err == ECONNABORTED) {
-				rc = sock_error(&smc->sk);
-				goto out_unlock;
-			}
+			if (smc->sk.sk_err == ECONNABORTED)
+				return sock_error(&smc->sk);
 			rc = 0;
 			if (conn->alert_token_local) /* connection healthy */
 				mod_delayed_work(system_wq, &conn->tx_work,
 						 SMC_TX_WORK_DELAY);
 		}
-		goto out_unlock;
+		return rc;
 	}
 
+	spin_lock_bh(&conn->send_lock);
 	if (!conn->local_tx_ctrl.prod_flags.urg_data_present) {
-		rc = smc_tx_rdma_writes(conn);
+		rc = smc_tx_rdma_writes(conn, wr_rdma_buf);
 		if (rc) {
 			smc_wr_tx_put_slot(&conn->lgr->lnk[SMC_SINGLE_LINK],
 					   (struct smc_wr_tx_pend_priv *)pend);
@@ -536,7 +533,7 @@ static int smcd_tx_sndbuf_nonempty(struct smc_connection *conn)
 
 	spin_lock_bh(&conn->send_lock);
 	if (!pflags->urg_data_present)
-		rc = smc_tx_rdma_writes(conn);
+		rc = smc_tx_rdma_writes(conn, NULL);
 	if (!rc)
 		rc = smcd_cdc_msg_send(conn);
 
@@ -598,7 +595,8 @@ void smc_tx_consumer_update(struct smc_connection *conn, bool force)
 	if (to_confirm > conn->rmbe_update_limit) {
 		smc_curs_copy(&prod, &conn->local_rx_ctrl.prod, conn);
 		sender_free = conn->rmb_desc->len -
-			      smc_curs_diff(conn->rmb_desc->len, &prod, &cfed);
+			      smc_curs_diff_large(conn->rmb_desc->len,
+						  &cfed, &prod);
 	}
 
 	if (conn->local_rx_ctrl.prod_flags.cons_curs_upd_req ||
diff --git a/net/smc/smc_wr.c b/net/smc/smc_wr.c
index c2694750a6a8..253aa75dc2b6 100644
--- a/net/smc/smc_wr.c
+++ b/net/smc/smc_wr.c
@@ -160,6 +160,7 @@ static inline int smc_wr_tx_get_free_slot_index(struct smc_link *link, u32 *idx)
  * @link:		Pointer to smc_link used to later send the message.
  * @handler:		Send completion handler function pointer.
  * @wr_buf:		Out value returns pointer to message buffer.
+ * @wr_rdma_buf:	Out value returns pointer to rdma work request.
  * @wr_pend_priv:	Out value returns pointer serving as handler context.
  *
  * Return: 0 on success, or -errno on error.
@@ -167,6 +168,7 @@ static inline int smc_wr_tx_get_free_slot_index(struct smc_link *link, u32 *idx)
 int smc_wr_tx_get_free_slot(struct smc_link *link,
 			    smc_wr_tx_handler handler,
 			    struct smc_wr_buf **wr_buf,
+			    struct smc_rdma_wr **wr_rdma_buf,
 			    struct smc_wr_tx_pend_priv **wr_pend_priv)
 {
 	struct smc_wr_tx_pend *wr_pend;
@@ -204,6 +206,8 @@ int smc_wr_tx_get_free_slot(struct smc_link *link,
 	wr_ib = &link->wr_tx_ibs[idx];
 	wr_ib->wr_id = wr_id;
 	*wr_buf = &link->wr_tx_bufs[idx];
+	if (wr_rdma_buf)
+		*wr_rdma_buf = &link->wr_tx_rdmas[idx];
 	*wr_pend_priv = &wr_pend->priv;
 	return 0;
 }
@@ -218,10 +222,10 @@ int smc_wr_tx_put_slot(struct smc_link *link,
 		u32 idx = pend->idx;
 
 		/* clear the full struct smc_wr_tx_pend including .priv */
-		memset(&link->wr_tx_pends[pend->idx], 0,
-		       sizeof(link->wr_tx_pends[pend->idx]));
-		memset(&link->wr_tx_bufs[pend->idx], 0,
-		       sizeof(link->wr_tx_bufs[pend->idx]));
+		memset(&link->wr_tx_pends[idx], 0,
+		       sizeof(link->wr_tx_pends[idx]));
+		memset(&link->wr_tx_bufs[idx], 0,
+		       sizeof(link->wr_tx_bufs[idx]));
 		test_and_clear_bit(idx, link->wr_tx_mask);
 		return 1;
 	}
@@ -465,12 +469,26 @@ static void smc_wr_init_sge(struct smc_link *lnk)
 			lnk->wr_tx_dma_addr + i * SMC_WR_BUF_SIZE;
 		lnk->wr_tx_sges[i].length = SMC_WR_TX_SIZE;
 		lnk->wr_tx_sges[i].lkey = lnk->roce_pd->local_dma_lkey;
+		lnk->wr_tx_rdma_sges[i].tx_rdma_sge[0].wr_tx_rdma_sge[0].lkey =
+			lnk->roce_pd->local_dma_lkey;
+		lnk->wr_tx_rdma_sges[i].tx_rdma_sge[0].wr_tx_rdma_sge[1].lkey =
+			lnk->roce_pd->local_dma_lkey;
+		lnk->wr_tx_rdma_sges[i].tx_rdma_sge[1].wr_tx_rdma_sge[0].lkey =
+			lnk->roce_pd->local_dma_lkey;
+		lnk->wr_tx_rdma_sges[i].tx_rdma_sge[1].wr_tx_rdma_sge[1].lkey =
+			lnk->roce_pd->local_dma_lkey;
 		lnk->wr_tx_ibs[i].next = NULL;
 		lnk->wr_tx_ibs[i].sg_list = &lnk->wr_tx_sges[i];
 		lnk->wr_tx_ibs[i].num_sge = 1;
 		lnk->wr_tx_ibs[i].opcode = IB_WR_SEND;
 		lnk->wr_tx_ibs[i].send_flags =
 			IB_SEND_SIGNALED | IB_SEND_SOLICITED;
+		lnk->wr_tx_rdmas[i].wr_tx_rdma[0].wr.opcode = IB_WR_RDMA_WRITE;
+		lnk->wr_tx_rdmas[i].wr_tx_rdma[1].wr.opcode = IB_WR_RDMA_WRITE;
+		lnk->wr_tx_rdmas[i].wr_tx_rdma[0].wr.sg_list =
+			lnk->wr_tx_rdma_sges[i].tx_rdma_sge[0].wr_tx_rdma_sge;
+		lnk->wr_tx_rdmas[i].wr_tx_rdma[1].wr.sg_list =
+			lnk->wr_tx_rdma_sges[i].tx_rdma_sge[1].wr_tx_rdma_sge;
 	}
 	for (i = 0; i < lnk->wr_rx_cnt; i++) {
 		lnk->wr_rx_sges[i].addr =
@@ -521,8 +539,12 @@ void smc_wr_free_link_mem(struct smc_link *lnk)
 	lnk->wr_tx_mask = NULL;
 	kfree(lnk->wr_tx_sges);
 	lnk->wr_tx_sges = NULL;
+	kfree(lnk->wr_tx_rdma_sges);
+	lnk->wr_tx_rdma_sges = NULL;
 	kfree(lnk->wr_rx_sges);
 	lnk->wr_rx_sges = NULL;
+	kfree(lnk->wr_tx_rdmas);
+	lnk->wr_tx_rdmas = NULL;
 	kfree(lnk->wr_rx_ibs);
 	lnk->wr_rx_ibs = NULL;
 	kfree(lnk->wr_tx_ibs);
@@ -552,10 +574,20 @@ int smc_wr_alloc_link_mem(struct smc_link *link)
 				  GFP_KERNEL);
 	if (!link->wr_rx_ibs)
 		goto no_mem_wr_tx_ibs;
+	link->wr_tx_rdmas = kcalloc(SMC_WR_BUF_CNT,
+				    sizeof(link->wr_tx_rdmas[0]),
+				    GFP_KERNEL);
+	if (!link->wr_tx_rdmas)
+		goto no_mem_wr_rx_ibs;
+	link->wr_tx_rdma_sges = kcalloc(SMC_WR_BUF_CNT,
+					sizeof(link->wr_tx_rdma_sges[0]),
+					GFP_KERNEL);
+	if (!link->wr_tx_rdma_sges)
+		goto no_mem_wr_tx_rdmas;
 	link->wr_tx_sges = kcalloc(SMC_WR_BUF_CNT, sizeof(link->wr_tx_sges[0]),
 				   GFP_KERNEL);
 	if (!link->wr_tx_sges)
-		goto no_mem_wr_rx_ibs;
+		goto no_mem_wr_tx_rdma_sges;
 	link->wr_rx_sges = kcalloc(SMC_WR_BUF_CNT * 3,
 				   sizeof(link->wr_rx_sges[0]),
 				   GFP_KERNEL);
@@ -579,6 +611,10 @@ no_mem_wr_rx_sges:
 	kfree(link->wr_rx_sges);
 no_mem_wr_tx_sges:
 	kfree(link->wr_tx_sges);
+no_mem_wr_tx_rdma_sges:
+	kfree(link->wr_tx_rdma_sges);
+no_mem_wr_tx_rdmas:
+	kfree(link->wr_tx_rdmas);
 no_mem_wr_rx_ibs:
 	kfree(link->wr_rx_ibs);
 no_mem_wr_tx_ibs:
diff --git a/net/smc/smc_wr.h b/net/smc/smc_wr.h
index 1d85bb14fd6f..09bf32fd3959 100644
--- a/net/smc/smc_wr.h
+++ b/net/smc/smc_wr.h
@@ -85,6 +85,7 @@ void smc_wr_add_dev(struct smc_ib_device *smcibdev);
 
 int smc_wr_tx_get_free_slot(struct smc_link *link, smc_wr_tx_handler handler,
 			    struct smc_wr_buf **wr_buf,
+			    struct smc_rdma_wr **wrs,
 			    struct smc_wr_tx_pend_priv **wr_pend_priv);
 int smc_wr_tx_put_slot(struct smc_link *link,
 		       struct smc_wr_tx_pend_priv *wr_pend_priv);