Merge branch 'net-smc-introduce-SMC-Dv2-support'

Karsten Graul says:

====================
net/smc: introduce SMC-Dv2 support

SMC-Dv2 support (see https://www.ibm.com/support/pages/node/6326337)
provides multi-subnet support for SMC-D, eliminating the current
same-subnet restriction. The new version detects if any of the virtual
ISM devices are on the same system and can therefore be used for an
SMC-Dv2 connection. Furthermore, SMC-Dv2 eliminates the need for
PNET IDs on s390.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

drivers/s390/net/ism.h

@@ -16,6 +16,7 @@
 #define ISM_DMB_WORD_OFFSET	1
 #define ISM_DMB_BIT_OFFSET	(ISM_DMB_WORD_OFFSET * 32)
 #define ISM_NR_DMBS		1920
+#define ISM_IDENT_MASK		0x00FFFF
 
 #define ISM_REG_SBA	0x1
 #define ISM_REG_IEQ	0x2
@@ -206,6 +207,12 @@ struct ism_dev {
 #define ISM_CREATE_REQ(dmb, idx, sf, offset)		\
 	((dmb) | (idx) << 24 | (sf) << 23 | (offset))
 
+struct ism_systemeid {
+	u8	seid_string[24];
+	u8	serial_number[4];
+	u8	type[4];
+};
+
 static inline void __ism_read_cmd(struct ism_dev *ism, void *data,
 				  unsigned long offset, unsigned long len)
 {

drivers/s390/net/ism_drv.c

@@ -13,6 +13,8 @@
 #include <linux/device.h>
 #include <linux/pci.h>
 #include <linux/err.h>
+#include <linux/ctype.h>
+#include <linux/processor.h>
 #include <net/smc.h>
 
 #include <asm/debug.h>
@@ -387,6 +389,42 @@ static int ism_move(struct smcd_dev *smcd, u64 dmb_tok, unsigned int idx,
 	return 0;
 }
 
+static struct ism_systemeid SYSTEM_EID = {
+	.seid_string = "IBM-SYSZ-IBMSEID00000000",
+	.serial_number = "0000",
+	.type = "0000",
+};
+
+static void ism_create_system_eid(void)
+{
+	struct cpuid id;
+	u16 ident_tail;
+	char tmp[5];
+
+	get_cpu_id(&id);
+	ident_tail = (u16)(id.ident & ISM_IDENT_MASK);
+	snprintf(tmp, 5, "%04X", ident_tail);
+	memcpy(&SYSTEM_EID.serial_number, tmp, 4);
+	snprintf(tmp, 5, "%04X", id.machine);
+	memcpy(&SYSTEM_EID.type, tmp, 4);
+}
+
+static void ism_get_system_eid(struct smcd_dev *smcd, u8 **eid)
+{
+	*eid = &SYSTEM_EID.seid_string[0];
+}
+
+static u16 ism_get_chid(struct smcd_dev *smcd)
+{
+	struct ism_dev *ismdev;
+
+	ismdev = (struct ism_dev *)smcd->priv;
+	if (!ismdev || !ismdev->pdev)
+		return 0;
+
+	return to_zpci(ismdev->pdev)->pchid;
+}
+
 static void ism_handle_event(struct ism_dev *ism)
 {
 	struct smcd_event *entry;
@@ -443,6 +481,8 @@ static const struct smcd_ops ism_ops = {
 	.reset_vlan_required = ism_reset_vlan_required,
 	.signal_event = ism_signal_ieq,
 	.move_data = ism_move,
+	.get_system_eid = ism_get_system_eid,
+	.get_chid = ism_get_chid,
 };
 
 static int ism_dev_init(struct ism_dev *ism)
@@ -471,6 +511,10 @@ static int ism_dev_init(struct ism_dev *ism)
 	if (ret)
 		goto unreg_ieq;
 
+	if (!ism_add_vlan_id(ism->smcd, ISM_RESERVED_VLANID))
+		/* hardware is V2 capable */
+		ism_create_system_eid();
+
 	ret = smcd_register_dev(ism->smcd);
 	if (ret)
 		goto unreg_ieq;
@@ -550,6 +594,9 @@ static void ism_dev_exit(struct ism_dev *ism)
 	struct pci_dev *pdev = ism->pdev;
 
 	smcd_unregister_dev(ism->smcd);
+	if (SYSTEM_EID.serial_number[0] != '0' ||
+	    SYSTEM_EID.type[0] != '0')
+		ism_del_vlan_id(ism->smcd, ISM_RESERVED_VLANID);
 	unregister_ieq(ism);
 	unregister_sba(ism);
 	free_irq(pci_irq_vector(pdev, 0), ism);

include/net/smc.h

@@ -37,6 +37,8 @@ struct smcd_dmb {
 #define ISM_EVENT_GID	1
 #define ISM_EVENT_SWR	2
 
+#define ISM_RESERVED_VLANID	0x1FFF
+
 #define ISM_ERROR	0xFFFF
 
 struct smcd_event {
@@ -63,6 +65,8 @@ struct smcd_ops {
 	int (*move_data)(struct smcd_dev *dev, u64 dmb_tok, unsigned int idx,
 			 bool sf, unsigned int offset, void *data,
 			 unsigned int size);
+	void (*get_system_eid)(struct smcd_dev *dev, u8 **eid);
+	u16 (*get_chid)(struct smcd_dev *dev);
 };
 
 struct smcd_dev {

net/smc/af_smc.c

@@ -26,6 +26,7 @@
 #include <linux/sched/signal.h>
 #include <linux/if_vlan.h>
 #include <linux/rcupdate_wait.h>
+#include <linux/ctype.h>
 
 #include <net/sock.h>
 #include <net/tcp.h>
@@ -448,6 +449,16 @@ static void smcr_conn_save_peer_info(struct smc_sock *smc,
 	smc->conn.tx_off = bufsize * (smc->conn.peer_rmbe_idx - 1);
 }
 
+static bool smc_isascii(char *hostname)
+{
+	int i;
+
+	for (i = 0; i < SMC_MAX_HOSTNAME_LEN; i++)
+		if (!isascii(hostname[i]))
+			return false;
+	return true;
+}
+
 static void smcd_conn_save_peer_info(struct smc_sock *smc,
 				     struct smc_clc_msg_accept_confirm *clc)
 {
@@ -459,6 +470,22 @@ static void smcd_conn_save_peer_info(struct smc_sock *smc,
 	smc->conn.peer_rmbe_size = bufsize - sizeof(struct smcd_cdc_msg);
 	atomic_set(&smc->conn.peer_rmbe_space, smc->conn.peer_rmbe_size);
 	smc->conn.tx_off = bufsize * smc->conn.peer_rmbe_idx;
+	if (clc->hdr.version > SMC_V1 &&
+	    (clc->hdr.typev2 & SMC_FIRST_CONTACT_MASK)) {
+		struct smc_clc_msg_accept_confirm_v2 *clc_v2 =
+			(struct smc_clc_msg_accept_confirm_v2 *)clc;
+		struct smc_clc_first_contact_ext *fce =
+			(struct smc_clc_first_contact_ext *)
+				(((u8 *)clc_v2) + sizeof(*clc_v2));
+
+		memcpy(smc->conn.lgr->negotiated_eid, clc_v2->eid,
+		       SMC_MAX_EID_LEN);
+		smc->conn.lgr->peer_os = fce->os_type;
+		smc->conn.lgr->peer_smc_release = fce->release;
+		if (smc_isascii(fce->hostname))
+			memcpy(smc->conn.lgr->peer_hostname, fce->hostname,
+			       SMC_MAX_HOSTNAME_LEN);
+	}
 }
 
 static void smc_conn_save_peer_info(struct smc_sock *smc,
@@ -504,7 +531,8 @@ static int smc_connect_fallback(struct smc_sock *smc, int reason_code)
 }
 
 /* decline and fall back during connect */
-static int smc_connect_decline_fallback(struct smc_sock *smc, int reason_code)
+static int smc_connect_decline_fallback(struct smc_sock *smc, int reason_code,
+					u8 version)
 {
 	int rc;
 
@@ -514,7 +542,7 @@ static int smc_connect_decline_fallback(struct smc_sock *smc, int reason_code)
 		return reason_code;
 	}
 	if (reason_code != SMC_CLC_DECL_PEERDECL) {
-		rc = smc_clc_send_decline(smc, reason_code);
+		rc = smc_clc_send_decline(smc, reason_code, version);
 		if (rc < 0) {
 			if (smc->sk.sk_state == SMC_INIT)
 				sock_put(&smc->sk); /* passive closing */
@@ -564,47 +592,121 @@ static int smc_find_ism_device(struct smc_sock *smc, struct smc_init_info *ini)
 {
 	/* Find ISM device with same PNETID as connecting interface  */
 	smc_pnet_find_ism_resource(smc->clcsock->sk, ini);
-	if (!ini->ism_dev)
+	if (!ini->ism_dev[0])
 		return SMC_CLC_DECL_NOSMCDDEV;
+	else
+		ini->ism_chid[0] = smc_ism_get_chid(ini->ism_dev[0]);
 	return 0;
 }
 
+/* determine possible V2 ISM devices (either without PNETID or with PNETID plus
+ * PNETID matching net_device)
+ */
+static int smc_find_ism_v2_device_clnt(struct smc_sock *smc,
+				       struct smc_init_info *ini)
+{
+	int rc = SMC_CLC_DECL_NOSMCDDEV;
+	struct smcd_dev *smcd;
+	int i = 1;
+
+	if (smcd_indicated(ini->smc_type_v1))
+		rc = 0;		/* already initialized for V1 */
+	mutex_lock(&smcd_dev_list.mutex);
+	list_for_each_entry(smcd, &smcd_dev_list.list, list) {
+		if (smcd->going_away || smcd == ini->ism_dev[0])
+			continue;
+		if (!smc_pnet_is_pnetid_set(smcd->pnetid) ||
+		    smc_pnet_is_ndev_pnetid(sock_net(&smc->sk), smcd->pnetid)) {
+			ini->ism_dev[i] = smcd;
+			ini->ism_chid[i] = smc_ism_get_chid(ini->ism_dev[i]);
+			ini->is_smcd = true;
+			rc = 0;
+			i++;
+			if (i > SMC_MAX_ISM_DEVS)
+				break;
+		}
+	}
+	mutex_unlock(&smcd_dev_list.mutex);
+	ini->ism_offered_cnt = i - 1;
+	if (!ini->ism_dev[0] && !ini->ism_dev[1])
+		ini->smcd_version = 0;
+
+	return rc;
+}
+
 /* Check for VLAN ID and register it on ISM device just for CLC handshake */
 static int smc_connect_ism_vlan_setup(struct smc_sock *smc,
 				      struct smc_init_info *ini)
 {
-	if (ini->vlan_id && smc_ism_get_vlan(ini->ism_dev, ini->vlan_id))
+	if (ini->vlan_id && smc_ism_get_vlan(ini->ism_dev[0], ini->vlan_id))
 		return SMC_CLC_DECL_ISMVLANERR;
 	return 0;
 }
 
+static int smc_find_proposal_devices(struct smc_sock *smc,
+				     struct smc_init_info *ini)
+{
+	int rc = 0;
+
+	/* check if there is an ism device available */
+	if (ini->smcd_version & SMC_V1) {
+		if (smc_find_ism_device(smc, ini) ||
+		    smc_connect_ism_vlan_setup(smc, ini)) {
+			if (ini->smc_type_v1 == SMC_TYPE_B)
+				ini->smc_type_v1 = SMC_TYPE_R;
+			else
+				ini->smc_type_v1 = SMC_TYPE_N;
+		} /* else ISM V1 is supported for this connection */
+		if (smc_find_rdma_device(smc, ini)) {
+			if (ini->smc_type_v1 == SMC_TYPE_B)
+				ini->smc_type_v1 = SMC_TYPE_D;
+			else
+				ini->smc_type_v1 = SMC_TYPE_N;
+		} /* else RDMA is supported for this connection */
+	}
+	if (smc_ism_v2_capable && smc_find_ism_v2_device_clnt(smc, ini))
+		ini->smc_type_v2 = SMC_TYPE_N;
+
+	/* if neither ISM nor RDMA are supported, fallback */
+	if (!smcr_indicated(ini->smc_type_v1) &&
+	    ini->smc_type_v1 == SMC_TYPE_N && ini->smc_type_v2 == SMC_TYPE_N)
+		rc = SMC_CLC_DECL_NOSMCDEV;
+
+	return rc;
+}
+
 /* cleanup temporary VLAN ID registration used for CLC handshake. If ISM is
  * used, the VLAN ID will be registered again during the connection setup.
  */
-static int smc_connect_ism_vlan_cleanup(struct smc_sock *smc, bool is_smcd,
+static int smc_connect_ism_vlan_cleanup(struct smc_sock *smc,
 					struct smc_init_info *ini)
 {
-	if (!is_smcd)
+	if (!smcd_indicated(ini->smc_type_v1))
 		return 0;
-	if (ini->vlan_id && smc_ism_put_vlan(ini->ism_dev, ini->vlan_id))
+	if (ini->vlan_id && smc_ism_put_vlan(ini->ism_dev[0], ini->vlan_id))
 		return SMC_CLC_DECL_CNFERR;
 	return 0;
 }
 
+#define SMC_CLC_MAX_ACCEPT_LEN \
+	(sizeof(struct smc_clc_msg_accept_confirm_v2) + \
+	 sizeof(struct smc_clc_first_contact_ext) + \
+	 sizeof(struct smc_clc_msg_trail))
+
 /* CLC handshake during connect */
-static int smc_connect_clc(struct smc_sock *smc, int smc_type,
-			   struct smc_clc_msg_accept_confirm *aclc,
+static int smc_connect_clc(struct smc_sock *smc,
+			   struct smc_clc_msg_accept_confirm_v2 *aclc2,
 			   struct smc_init_info *ini)
 {
 	int rc = 0;
 
 	/* do inband token exchange */
-	rc = smc_clc_send_proposal(smc, smc_type, ini);
+	rc = smc_clc_send_proposal(smc, ini);
 	if (rc)
 		return rc;
 	/* receive SMC Accept CLC message */
-	return smc_clc_wait_msg(smc, aclc, sizeof(*aclc), SMC_CLC_ACCEPT,
-				CLC_WAIT_TIME);
+	return smc_clc_wait_msg(smc, aclc2, SMC_CLC_MAX_ACCEPT_LEN,
+				SMC_CLC_ACCEPT, CLC_WAIT_TIME);
 }
 
 /* setup for RDMA connection of client */
@@ -618,7 +720,7 @@ static int smc_connect_rdma(struct smc_sock *smc,
 	ini->is_smcd = false;
 	ini->ib_lcl = &aclc->r0.lcl;
 	ini->ib_clcqpn = ntoh24(aclc->r0.qpn);
-	ini->first_contact_peer = aclc->hdr.flag;
+	ini->first_contact_peer = aclc->hdr.typev2 & SMC_FIRST_CONTACT_MASK;
 
 	mutex_lock(&smc_client_lgr_pending);
 	reason_code = smc_conn_create(smc, ini);
@@ -678,7 +780,8 @@ static int smc_connect_rdma(struct smc_sock *smc,
 	}
 	smc_rmb_sync_sg_for_device(&smc->conn);
 
-	reason_code = smc_clc_send_confirm(smc);
+	reason_code = smc_clc_send_confirm(smc, ini->first_contact_local,
+					   SMC_V1);
 	if (reason_code)
 		return smc_connect_abort(smc, reason_code,
 					 ini->first_contact_local);
@@ -704,6 +807,25 @@ static int smc_connect_rdma(struct smc_sock *smc,
 	return 0;
 }
 
+/* The server has chosen one of the proposed ISM devices for the communication.
+ * Determine from the CHID of the received CLC ACCEPT the ISM device chosen.
+ */
+static int
+smc_v2_determine_accepted_chid(struct smc_clc_msg_accept_confirm_v2 *aclc,
+			       struct smc_init_info *ini)
+{
+	int i;
+
+	for (i = 0; i < ini->ism_offered_cnt + 1; i++) {
+		if (ini->ism_chid[i] == ntohs(aclc->chid)) {
+			ini->ism_selected = i;
+			return 0;
+		}
+	}
+
+	return -EPROTO;
+}
+
 /* setup for ISM connection of client */
 static int smc_connect_ism(struct smc_sock *smc,
 			   struct smc_clc_msg_accept_confirm *aclc,
@@ -712,8 +834,17 @@ static int smc_connect_ism(struct smc_sock *smc,
 	int rc = 0;
 
 	ini->is_smcd = true;
-	ini->ism_peer_gid = aclc->d0.gid;
-	ini->first_contact_peer = aclc->hdr.flag;
+	ini->first_contact_peer = aclc->hdr.typev2 & SMC_FIRST_CONTACT_MASK;
+
+	if (aclc->hdr.version == SMC_V2) {
+		struct smc_clc_msg_accept_confirm_v2 *aclc_v2 =
+			(struct smc_clc_msg_accept_confirm_v2 *)aclc;
+
+		rc = smc_v2_determine_accepted_chid(aclc_v2, ini);
+		if (rc)
+			return rc;
+	}
+	ini->ism_peer_gid[ini->ism_selected] = aclc->d0.gid;
 
 	/* there is only one lgr role for SMC-D; use server lock */
 	mutex_lock(&smc_server_lgr_pending);
@@ -736,7 +867,8 @@ static int smc_connect_ism(struct smc_sock *smc,
 	smc_rx_init(smc);
 	smc_tx_init(smc);
 
-	rc = smc_clc_send_confirm(smc);
+	rc = smc_clc_send_confirm(smc, ini->first_contact_local,
+				  aclc->hdr.version);
 	if (rc)
 		return smc_connect_abort(smc, rc, ini->first_contact_local);
 	mutex_unlock(&smc_server_lgr_pending);
@@ -749,13 +881,32 @@ static int smc_connect_ism(struct smc_sock *smc,
 	return 0;
 }
 
+/* check if received accept type and version matches a proposed one */
+static int smc_connect_check_aclc(struct smc_init_info *ini,
+				  struct smc_clc_msg_accept_confirm *aclc)
+{
+	if ((aclc->hdr.typev1 == SMC_TYPE_R &&
+	     !smcr_indicated(ini->smc_type_v1)) ||
+	    (aclc->hdr.typev1 == SMC_TYPE_D &&
+	     ((!smcd_indicated(ini->smc_type_v1) &&
+	       !smcd_indicated(ini->smc_type_v2)) ||
+	      (aclc->hdr.version == SMC_V1 &&
+	       !smcd_indicated(ini->smc_type_v1)) ||
+	      (aclc->hdr.version == SMC_V2 &&
+	       !smcd_indicated(ini->smc_type_v2)))))
+		return SMC_CLC_DECL_MODEUNSUPP;
+
+	return 0;
+}
+
 /* perform steps before actually connecting */
 static int __smc_connect(struct smc_sock *smc)
 {
-	bool ism_supported = false, rdma_supported = false;
-	struct smc_clc_msg_accept_confirm aclc;
-	struct smc_init_info ini = {0};
-	int smc_type;
+	u8 version = smc_ism_v2_capable ? SMC_V2 : SMC_V1;
+	struct smc_clc_msg_accept_confirm_v2 *aclc2;
+	struct smc_clc_msg_accept_confirm *aclc;
+	struct smc_init_info *ini = NULL;
+	u8 *buf = NULL;
 	int rc = 0;
 
 	if (smc->use_fallback)
@@ -765,58 +916,73 @@ static int __smc_connect(struct smc_sock *smc)
 	if (!tcp_sk(smc->clcsock->sk)->syn_smc)
 		return smc_connect_fallback(smc, SMC_CLC_DECL_PEERNOSMC);
 
-	/* IPSec connections opt out of SMC-R optimizations */
+	/* IPSec connections opt out of SMC optimizations */
 	if (using_ipsec(smc))
-		return smc_connect_decline_fallback(smc, SMC_CLC_DECL_IPSEC);
+		return smc_connect_decline_fallback(smc, SMC_CLC_DECL_IPSEC,
+						    version);
 
-	/* get vlan id from IP device */
-	if (smc_vlan_by_tcpsk(smc->clcsock, &ini))
-		return smc_connect_decline_fallback(smc,
-						    SMC_CLC_DECL_GETVLANERR);
+	ini = kzalloc(sizeof(*ini), GFP_KERNEL);
+	if (!ini)
+		return smc_connect_decline_fallback(smc, SMC_CLC_DECL_MEM,
+						    version);
 
-	/* check if there is an ism device available */
-	if (!smc_find_ism_device(smc, &ini) &&
-	    !smc_connect_ism_vlan_setup(smc, &ini)) {
-		/* ISM is supported for this connection */
-		ism_supported = true;
-		smc_type = SMC_TYPE_D;
-	}
-
-	/* check if there is a rdma device available */
-	if (!smc_find_rdma_device(smc, &ini)) {
-		/* RDMA is supported for this connection */
-		rdma_supported = true;
-		if (ism_supported)
-			smc_type = SMC_TYPE_B; /* both */
-		else
-			smc_type = SMC_TYPE_R; /* only RDMA */
+	ini->smcd_version = SMC_V1;
+	ini->smcd_version |= smc_ism_v2_capable ? SMC_V2 : 0;
+	ini->smc_type_v1 = SMC_TYPE_B;
+	ini->smc_type_v2 = smc_ism_v2_capable ? SMC_TYPE_D : SMC_TYPE_N;
+
+	/* get vlan id from IP device */
+	if (smc_vlan_by_tcpsk(smc->clcsock, ini)) {
+		ini->smcd_version &= ~SMC_V1;
+		ini->smc_type_v1 = SMC_TYPE_N;
+		if (!ini->smcd_version) {
+			rc = SMC_CLC_DECL_GETVLANERR;
+			goto fallback;
+		}
 	}
 
-	/* if neither ISM nor RDMA are supported, fallback */
-	if (!rdma_supported && !ism_supported)
-		return smc_connect_decline_fallback(smc, SMC_CLC_DECL_NOSMCDEV);
+	rc = smc_find_proposal_devices(smc, ini);
+	if (rc)
+		goto fallback;
 
-	/* perform CLC handshake */
-	rc = smc_connect_clc(smc, smc_type, &aclc, &ini);
-	if (rc) {
-		smc_connect_ism_vlan_cleanup(smc, ism_supported, &ini);
-		return smc_connect_decline_fallback(smc, rc);
+	buf = kzalloc(SMC_CLC_MAX_ACCEPT_LEN, GFP_KERNEL);
+	if (!buf) {
+		rc = SMC_CLC_DECL_MEM;
+		goto fallback;
 	}
+	aclc2 = (struct smc_clc_msg_accept_confirm_v2 *)buf;
+	aclc = (struct smc_clc_msg_accept_confirm *)aclc2;
+
+	/* perform CLC handshake */
+	rc = smc_connect_clc(smc, aclc2, ini);
+	if (rc)
+		goto vlan_cleanup;
+
+	/* check if smc modes and versions of CLC proposal and accept match */
+	rc = smc_connect_check_aclc(ini, aclc);
+	version = aclc->hdr.version == SMC_V1 ? SMC_V1 : version;
+	if (rc)
+		goto vlan_cleanup;
 
 	/* depending on previous steps, connect using rdma or ism */
-	if (rdma_supported && aclc.hdr.path == SMC_TYPE_R)
-		rc = smc_connect_rdma(smc, &aclc, &ini);
-	else if (ism_supported && aclc.hdr.path == SMC_TYPE_D)
-		rc = smc_connect_ism(smc, &aclc, &ini);
-	else
-		rc = SMC_CLC_DECL_MODEUNSUPP;
-	if (rc) {
-		smc_connect_ism_vlan_cleanup(smc, ism_supported, &ini);
-		return smc_connect_decline_fallback(smc, rc);
-	}
+	if (aclc->hdr.typev1 == SMC_TYPE_R)
+		rc = smc_connect_rdma(smc, aclc, ini);
+	else if (aclc->hdr.typev1 == SMC_TYPE_D)
+		rc = smc_connect_ism(smc, aclc, ini);
+	if (rc)
+		goto vlan_cleanup;
 
-	smc_connect_ism_vlan_cleanup(smc, ism_supported, &ini);
+	smc_connect_ism_vlan_cleanup(smc, ini);
+	kfree(buf);
+	kfree(ini);
 	return 0;
+
+vlan_cleanup:
+	smc_connect_ism_vlan_cleanup(smc, ini);
+	kfree(buf);
+fallback:
+	kfree(ini);
+	return smc_connect_decline_fallback(smc, rc, version);
 }
 
 static void smc_connect_work(struct work_struct *work)
@@ -1132,10 +1298,10 @@ static void smc_listen_out_err(struct smc_sock *new_smc)
 
 /* listen worker: decline and fall back if possible */
 static void smc_listen_decline(struct smc_sock *new_smc, int reason_code,
-			       bool local_first)
+			       struct smc_init_info *ini, u8 version)
 {
 	/* RDMA setup failed, switch back to TCP */
-	if (local_first)
+	if (ini->first_contact_local)
 		smc_lgr_cleanup_early(&new_smc->conn);
 	else
 		smc_conn_free(&new_smc->conn);
@@ -1146,7 +1312,7 @@ static void smc_listen_decline(struct smc_sock *new_smc, int reason_code,
 	smc_switch_to_fallback(new_smc);
 	new_smc->fallback_rsn = reason_code;
 	if (reason_code && reason_code != SMC_CLC_DECL_PEERDECL) {
-		if (smc_clc_send_decline(new_smc, reason_code) < 0) {
+		if (smc_clc_send_decline(new_smc, reason_code, version) < 0) {
 			smc_listen_out_err(new_smc);
 			return;
 		}
@@ -1154,6 +1320,47 @@ static void smc_listen_decline(struct smc_sock *new_smc, int reason_code,
 	smc_listen_out_connected(new_smc);
 }
 
+/* listen worker: version checking */
+static int smc_listen_v2_check(struct smc_sock *new_smc,
+			       struct smc_clc_msg_proposal *pclc,
+			       struct smc_init_info *ini)
+{
+	struct smc_clc_smcd_v2_extension *pclc_smcd_v2_ext;
+	struct smc_clc_v2_extension *pclc_v2_ext;
+
+	ini->smc_type_v1 = pclc->hdr.typev1;
+	ini->smc_type_v2 = pclc->hdr.typev2;
+	ini->smcd_version = ini->smc_type_v1 != SMC_TYPE_N ? SMC_V1 : 0;
+	if (pclc->hdr.version > SMC_V1)
+		ini->smcd_version |=
+				ini->smc_type_v2 != SMC_TYPE_N ? SMC_V2 : 0;
+	if (!smc_ism_v2_capable) {
+		ini->smcd_version &= ~SMC_V2;
+		goto out;
+	}
+	pclc_v2_ext = smc_get_clc_v2_ext(pclc);
+	if (!pclc_v2_ext) {
+		ini->smcd_version &= ~SMC_V2;
+		goto out;
+	}
+	pclc_smcd_v2_ext = smc_get_clc_smcd_v2_ext(pclc_v2_ext);
+	if (!pclc_smcd_v2_ext)
+		ini->smcd_version &= ~SMC_V2;
+
+out:
+	if (!ini->smcd_version) {
+		if (pclc->hdr.typev1 == SMC_TYPE_B ||
+		    pclc->hdr.typev2 == SMC_TYPE_B)
+			return SMC_CLC_DECL_NOSMCDEV;
+		if (pclc->hdr.typev1 == SMC_TYPE_D ||
+		    pclc->hdr.typev2 == SMC_TYPE_D)
+			return SMC_CLC_DECL_NOSMCDDEV;
+		return SMC_CLC_DECL_NOSMCRDEV;
+	}
+
+	return 0;
+}
+
 /* listen worker: check prefixes */
 static int smc_listen_prfx_check(struct smc_sock *new_smc,
 				 struct smc_clc_msg_proposal *pclc)
@@ -1161,6 +1368,8 @@ static int smc_listen_prfx_check(struct smc_sock *new_smc,
 	struct smc_clc_msg_proposal_prefix *pclc_prfx;
 	struct socket *newclcsock = new_smc->clcsock;
 
+	if (pclc->hdr.typev1 == SMC_TYPE_N)
+		return 0;
 	pclc_prfx = smc_clc_proposal_get_prefix(pclc);
 	if (smc_clc_prfx_match(newclcsock, pclc_prfx))
 		return SMC_CLC_DECL_DIFFPREFIX;
@@ -1188,7 +1397,6 @@ static int smc_listen_rdma_init(struct smc_sock *new_smc,
 
 /* listen worker: initialize connection and buffers for SMC-D */
 static int smc_listen_ism_init(struct smc_sock *new_smc,
-			       struct smc_clc_msg_proposal *pclc,
 			       struct smc_init_info *ini)
 {
 	int rc;
@@ -1211,6 +1419,125 @@ static int smc_listen_ism_init(struct smc_sock *new_smc,
 	return 0;
 }
 
+static bool smc_is_already_selected(struct smcd_dev *smcd,
+				    struct smc_init_info *ini,
+				    int matches)
+{
+	int i;
+
+	for (i = 0; i < matches; i++)
+		if (smcd == ini->ism_dev[i])
+			return true;
+
+	return false;
+}
+
+/* check for ISM devices matching proposed ISM devices */
+static void smc_check_ism_v2_match(struct smc_init_info *ini,
+				   u16 proposed_chid, u64 proposed_gid,
+				   unsigned int *matches)
+{
+	struct smcd_dev *smcd;
+
+	list_for_each_entry(smcd, &smcd_dev_list.list, list) {
+		if (smcd->going_away)
+			continue;
+		if (smc_is_already_selected(smcd, ini, *matches))
+			continue;
+		if (smc_ism_get_chid(smcd) == proposed_chid &&
+		    !smc_ism_cantalk(proposed_gid, ISM_RESERVED_VLANID, smcd)) {
+			ini->ism_peer_gid[*matches] = proposed_gid;
+			ini->ism_dev[*matches] = smcd;
+			(*matches)++;
+			break;
+		}
+	}
+}
+
+static void smc_find_ism_v2_device_serv(struct smc_sock *new_smc,
+					struct smc_clc_msg_proposal *pclc,
+					struct smc_init_info *ini)
+{
+	struct smc_clc_smcd_v2_extension *smcd_v2_ext;
+	struct smc_clc_v2_extension *smc_v2_ext;
+	struct smc_clc_msg_smcd *pclc_smcd;
+	unsigned int matches = 0;
+	u8 *eid = NULL;
+	int i;
+
+	if (!(ini->smcd_version & SMC_V2) || !smcd_indicated(ini->smc_type_v2))
+		return;
+
+	pclc_smcd = smc_get_clc_msg_smcd(pclc);
+	smc_v2_ext = smc_get_clc_v2_ext(pclc);
+	smcd_v2_ext = smc_get_clc_smcd_v2_ext(smc_v2_ext);
+	if (!smcd_v2_ext ||
+	    !smc_v2_ext->hdr.flag.seid) /* no system EID support for SMCD */
+		goto not_found;
+
+	mutex_lock(&smcd_dev_list.mutex);
+	if (pclc_smcd->ism.chid)
+		/* check for ISM device matching proposed native ISM device */
+		smc_check_ism_v2_match(ini, ntohs(pclc_smcd->ism.chid),
+				       ntohll(pclc_smcd->ism.gid), &matches);
+	for (i = 1; i <= smc_v2_ext->hdr.ism_gid_cnt; i++) {
+		/* check for ISM devices matching proposed non-native ISM
+		 * devices
+		 */
+		smc_check_ism_v2_match(ini,
+				       ntohs(smcd_v2_ext->gidchid[i - 1].chid),
+				       ntohll(smcd_v2_ext->gidchid[i - 1].gid),
+				       &matches);
+	}
+	mutex_unlock(&smcd_dev_list.mutex);
+
+	if (ini->ism_dev[0]) {
+		smc_ism_get_system_eid(ini->ism_dev[0], &eid);
+		if (memcmp(eid, smcd_v2_ext->system_eid, SMC_MAX_EID_LEN))
+			goto not_found;
+	} else {
+		goto not_found;
+	}
+
+	/* separate - outside the smcd_dev_list.lock */
+	for (i = 0; i < matches; i++) {
+		ini->smcd_version = SMC_V2;
+		ini->is_smcd = true;
+		ini->ism_selected = i;
+		if (smc_listen_ism_init(new_smc, ini))
+			/* try next active ISM device */
+			continue;
+		return; /* matching and usable V2 ISM device found */
+	}
+
+not_found:
+	ini->smcd_version &= ~SMC_V2;
+	ini->ism_dev[0] = NULL;
+	ini->is_smcd = false;
+}
+
+static void smc_find_ism_v1_device_serv(struct smc_sock *new_smc,
+					struct smc_clc_msg_proposal *pclc,
+					struct smc_init_info *ini)
+{
+	struct smc_clc_msg_smcd *pclc_smcd = smc_get_clc_msg_smcd(pclc);
+
+	/* check if ISM V1 is available */
+	if (!(ini->smcd_version & SMC_V1) || !smcd_indicated(ini->smc_type_v1))
+		goto not_found;
+	ini->is_smcd = true; /* prepare ISM check */
+	ini->ism_peer_gid[0] = ntohll(pclc_smcd->ism.gid);
+	if (smc_find_ism_device(new_smc, ini))
+		goto not_found;
+	ini->ism_selected = 0;
+	if (!smc_listen_ism_init(new_smc, ini))
+		return;		/* V1 ISM device found */
+
+not_found:
+	ini->ism_dev[0] = NULL;
+	ini->is_smcd = false;
+}
+
 /* listen worker: register buffers */
 static int smc_listen_rdma_reg(struct smc_sock *new_smc, bool local_first)
 {
@@ -1225,6 +1552,67 @@ static int smc_listen_rdma_reg(struct smc_sock *new_smc, bool local_first)
 	return 0;
 }
 
+static int smc_find_rdma_v1_device_serv(struct smc_sock *new_smc,
+					struct smc_clc_msg_proposal *pclc,
+					struct smc_init_info *ini)
+{
+	int rc;
+
+	if (!smcr_indicated(ini->smc_type_v1))
+		return SMC_CLC_DECL_NOSMCDEV;
+
+	/* prepare RDMA check */
+	ini->ib_lcl = &pclc->lcl;
+	rc = smc_find_rdma_device(new_smc, ini);
+	if (rc) {
+		/* no RDMA device found */
+		if (ini->smc_type_v1 == SMC_TYPE_B)
+			/* neither ISM nor RDMA device found */
+			rc = SMC_CLC_DECL_NOSMCDEV;
+		return rc;
+	}
+	rc = smc_listen_rdma_init(new_smc, ini);
+	if (rc)
+		return rc;
+	return smc_listen_rdma_reg(new_smc, ini->first_contact_local);
+}
+
+/* determine the local device matching to proposal */
+static int smc_listen_find_device(struct smc_sock *new_smc,
+				  struct smc_clc_msg_proposal *pclc,
+				  struct smc_init_info *ini)
+{
+	int rc;
+
+	/* check for ISM device matching V2 proposed device */
+	smc_find_ism_v2_device_serv(new_smc, pclc, ini);
+	if (ini->ism_dev[0])
+		return 0;
+
+	if (!(ini->smcd_version & SMC_V1))
+		return SMC_CLC_DECL_NOSMCDEV;
+
+	/* check for matching IP prefix and subnet length */
+	rc = smc_listen_prfx_check(new_smc, pclc);
+	if (rc)
+		return rc;
+
+	/* get vlan id from IP device */
+	if (smc_vlan_by_tcpsk(new_smc->clcsock, ini))
+		return SMC_CLC_DECL_GETVLANERR;
+
+	/* check for ISM device matching V1 proposed device */
+	smc_find_ism_v1_device_serv(new_smc, pclc, ini);
+	if (ini->ism_dev[0])
+		return 0;
+
+	if (pclc->hdr.typev1 == SMC_TYPE_D)
+		return SMC_CLC_DECL_NOSMCDDEV; /* skip RDMA and decline */
+
+	/* check if RDMA is available */
+	return smc_find_rdma_v1_device_serv(new_smc, pclc, ini);
+}
+
 /* listen worker: finish RDMA setup */
 static int smc_listen_rdma_finish(struct smc_sock *new_smc,
 				  struct smc_clc_msg_accept_confirm *cclc,
@@ -1250,17 +1638,18 @@ static int smc_listen_rdma_finish(struct smc_sock *new_smc,
 	return reason_code;
 }
 
-/* setup for RDMA connection of server */
+/* setup for connection of server */
 static void smc_listen_work(struct work_struct *work)
 {
 	struct smc_sock *new_smc = container_of(work, struct smc_sock,
 						smc_listen_work);
+	u8 version = smc_ism_v2_capable ? SMC_V2 : SMC_V1;
 	struct socket *newclcsock = new_smc->clcsock;
-	struct smc_clc_msg_accept_confirm cclc;
+	struct smc_clc_msg_accept_confirm_v2 *cclc2;
+	struct smc_clc_msg_accept_confirm *cclc;
 	struct smc_clc_msg_proposal_area *buf;
 	struct smc_clc_msg_proposal *pclc;
-	struct smc_init_info ini = {0};
-	bool ism_supported = false;
+	struct smc_init_info *ini = NULL;
 	int rc = 0;
 
 	if (new_smc->listen_smc->sk.sk_state != SMC_LISTEN)
@@ -1292,101 +1681,76 @@ static void smc_listen_work(struct work_struct *work)
 			      SMC_CLC_PROPOSAL, CLC_WAIT_TIME);
 	if (rc)
 		goto out_decl;
+	version = pclc->hdr.version == SMC_V1 ? SMC_V1 : version;
 
-	/* IPSec connections opt out of SMC-R optimizations */
+	/* IPSec connections opt out of SMC optimizations */
 	if (using_ipsec(new_smc)) {
 		rc = SMC_CLC_DECL_IPSEC;
 		goto out_decl;
 	}
 
-	/* check for matching IP prefix and subnet length */
-	rc = smc_listen_prfx_check(new_smc, pclc);
-	if (rc)
+	ini = kzalloc(sizeof(*ini), GFP_KERNEL);
+	if (!ini) {
+		rc = SMC_CLC_DECL_MEM;
 		goto out_decl;
+	}
 
-	/* get vlan id from IP device */
-	if (smc_vlan_by_tcpsk(new_smc->clcsock, &ini)) {
-		rc = SMC_CLC_DECL_GETVLANERR;
+	/* initial version checking */
+	rc = smc_listen_v2_check(new_smc, pclc, ini);
+	if (rc)
 		goto out_decl;
-	}
 
 	mutex_lock(&smc_server_lgr_pending);
 	smc_close_init(new_smc);
 	smc_rx_init(new_smc);
 	smc_tx_init(new_smc);
 
-	/* check if ISM is available */
-	if (pclc->hdr.path == SMC_TYPE_D || pclc->hdr.path == SMC_TYPE_B) {
-		struct smc_clc_msg_smcd *pclc_smcd = smc_get_clc_msg_smcd(pclc);
-
-		ini.is_smcd = true; /* prepare ISM check */
-		ini.ism_peer_gid = pclc_smcd->gid;
-		rc = smc_find_ism_device(new_smc, &ini);
-		if (!rc)
-			rc = smc_listen_ism_init(new_smc, pclc, &ini);
-		if (!rc)
-			ism_supported = true;
-		else if (pclc->hdr.path == SMC_TYPE_D)
-			goto out_unlock; /* skip RDMA and decline */
-	}
-
-	/* check if RDMA is available */
-	if (!ism_supported) { /* SMC_TYPE_R or SMC_TYPE_B */
-		/* prepare RDMA check */
-		ini.is_smcd = false;
-		ini.ism_dev = NULL;
-		ini.ib_lcl = &pclc->lcl;
-		rc = smc_find_rdma_device(new_smc, &ini);
-		if (rc) {
-			/* no RDMA device found */
-			if (pclc->hdr.path == SMC_TYPE_B)
-				/* neither ISM nor RDMA device found */
-				rc = SMC_CLC_DECL_NOSMCDEV;
-			goto out_unlock;
-		}
-		rc = smc_listen_rdma_init(new_smc, &ini);
-		if (rc)
-			goto out_unlock;
-		rc = smc_listen_rdma_reg(new_smc, ini.first_contact_local);
-		if (rc)
-			goto out_unlock;
-	}
+	/* determine ISM or RoCE device used for connection */
+	rc = smc_listen_find_device(new_smc, pclc, ini);
+	if (rc)
+		goto out_unlock;
 
 	/* send SMC Accept CLC message */
-	rc = smc_clc_send_accept(new_smc, ini.first_contact_local);
+	rc = smc_clc_send_accept(new_smc, ini->first_contact_local,
+				 ini->smcd_version == SMC_V2 ? SMC_V2 : SMC_V1);
 	if (rc)
 		goto out_unlock;
 
 	/* SMC-D does not need this lock any more */
-	if (ism_supported)
+	if (ini->is_smcd)
 		mutex_unlock(&smc_server_lgr_pending);
 
 	/* receive SMC Confirm CLC message */
-	rc = smc_clc_wait_msg(new_smc, &cclc, sizeof(cclc),
+	cclc2 = (struct smc_clc_msg_accept_confirm_v2 *)buf;
+	cclc = (struct smc_clc_msg_accept_confirm *)cclc2;
+	memset(buf, 0, sizeof(struct smc_clc_msg_proposal_area));
+	rc = smc_clc_wait_msg(new_smc, cclc2,
+			      sizeof(struct smc_clc_msg_proposal_area),
 			      SMC_CLC_CONFIRM, CLC_WAIT_TIME);
 	if (rc) {
-		if (!ism_supported)
+		if (!ini->is_smcd)
 			goto out_unlock;
 		goto out_decl;
 	}
 
 	/* finish worker */
-	if (!ism_supported) {
-		rc = smc_listen_rdma_finish(new_smc, &cclc,
-					    ini.first_contact_local);
+	if (!ini->is_smcd) {
+		rc = smc_listen_rdma_finish(new_smc, cclc,
+					    ini->first_contact_local);
 		if (rc)
 			goto out_unlock;
 		mutex_unlock(&smc_server_lgr_pending);
 	}
-	smc_conn_save_peer_info(new_smc, &cclc);
+	smc_conn_save_peer_info(new_smc, cclc);
 	smc_listen_out_connected(new_smc);
 	goto out_free;
 
 out_unlock:
 	mutex_unlock(&smc_server_lgr_pending);
 out_decl:
-	smc_listen_decline(new_smc, rc, ini.first_contact_local);
+	smc_listen_decline(new_smc, rc, ini, version);
 out_free:
+	kfree(ini);
 	kfree(buf);
 }
 
@@ -2092,6 +2456,9 @@ static int __init smc_init(void)
 	if (rc)
 		return rc;
 
+	smc_ism_init();
+	smc_clc_init();
+
 	rc = smc_pnet_init();
 	if (rc)
 		goto out_pernet_subsys;

net/smc/smc.h

@@ -19,10 +19,19 @@
 #include "smc_ib.h"
 
 #define SMC_V1		1		/* SMC version V1 */
+#define SMC_V2		2		/* SMC version V2 */
+#define SMC_RELEASE	0
 
 #define SMCPROTO_SMC		0	/* SMC protocol, IPv4 */
 #define SMCPROTO_SMC6		1	/* SMC protocol, IPv6 */
 
+#define SMC_MAX_ISM_DEVS	8	/* max # of proposed non-native ISM
+					 * devices
+					 */
+
+#define SMC_MAX_HOSTNAME_LEN	32
+#define SMC_MAX_EID_LEN		32
+
 extern struct proto smc_proto;
 extern struct proto smc_proto6;
 
@@ -246,6 +255,9 @@ extern struct workqueue_struct	*smc_close_wq;	/* wq for close work */
 
 extern u8	local_systemid[SMC_SYSTEMID_LEN]; /* unique system identifier */
 
+#define ntohll(x) be64_to_cpu(x)
+#define htonll(x) cpu_to_be64(x)
+
 /* convert an u32 value into network byte order, store it into a 3 byte field */
 static inline void hton24(u8 *net, u32 host)
 {

net/smc/smc_clc.c

@@ -14,6 +14,8 @@
 #include <linux/inetdevice.h>
 #include <linux/if_ether.h>
 #include <linux/sched/signal.h>
+#include <linux/utsname.h>
+#include <linux/ctype.h>
 
 #include <net/addrconf.h>
 #include <net/sock.h>
@@ -27,6 +29,7 @@
 
 #define SMCR_CLC_ACCEPT_CONFIRM_LEN 68
 #define SMCD_CLC_ACCEPT_CONFIRM_LEN 48
+#define SMCD_CLC_ACCEPT_CONFIRM_LEN_V2 78
 #define SMC_CLC_RECV_BUF_LEN	100
 
 /* eye catcher "SMCR" EBCDIC for CLC messages */
@@ -34,13 +37,88 @@ static const char SMC_EYECATCHER[4] = {'\xe2', '\xd4', '\xc3', '\xd9'};
 /* eye catcher "SMCD" EBCDIC for CLC messages */
 static const char SMCD_EYECATCHER[4] = {'\xe2', '\xd4', '\xc3', '\xc4'};
 
+static u8 smc_hostname[SMC_MAX_HOSTNAME_LEN];
+
+/* check arriving CLC proposal */
+static bool smc_clc_msg_prop_valid(struct smc_clc_msg_proposal *pclc)
+{
+	struct smc_clc_msg_proposal_prefix *pclc_prfx;
+	struct smc_clc_smcd_v2_extension *smcd_v2_ext;
+	struct smc_clc_msg_hdr *hdr = &pclc->hdr;
+	struct smc_clc_v2_extension *v2_ext;
+
+	v2_ext = smc_get_clc_v2_ext(pclc);
+	pclc_prfx = smc_clc_proposal_get_prefix(pclc);
+	if (hdr->version == SMC_V1) {
+		if (hdr->typev1 == SMC_TYPE_N)
+			return false;
+		if (ntohs(hdr->length) !=
+			sizeof(*pclc) + ntohs(pclc->iparea_offset) +
+			sizeof(*pclc_prfx) +
+			pclc_prfx->ipv6_prefixes_cnt *
+				sizeof(struct smc_clc_ipv6_prefix) +
+			sizeof(struct smc_clc_msg_trail))
+			return false;
+	} else {
+		if (ntohs(hdr->length) !=
+			sizeof(*pclc) +
+			sizeof(struct smc_clc_msg_smcd) +
+			(hdr->typev1 != SMC_TYPE_N ?
+				sizeof(*pclc_prfx) +
+				pclc_prfx->ipv6_prefixes_cnt *
+				sizeof(struct smc_clc_ipv6_prefix) : 0) +
+			(hdr->typev2 != SMC_TYPE_N ?
+				sizeof(*v2_ext) +
+				v2_ext->hdr.eid_cnt * SMC_MAX_EID_LEN : 0) +
+			(smcd_indicated(hdr->typev2) ?
+				sizeof(*smcd_v2_ext) + v2_ext->hdr.ism_gid_cnt *
+					sizeof(struct smc_clc_smcd_gid_chid) :
+				0) +
+			sizeof(struct smc_clc_msg_trail))
+			return false;
+	}
+	return true;
+}
+
+/* check arriving CLC accept or confirm */
+static bool
+smc_clc_msg_acc_conf_valid(struct smc_clc_msg_accept_confirm_v2 *clc_v2)
+{
+	struct smc_clc_msg_hdr *hdr = &clc_v2->hdr;
+
+	if (hdr->typev1 != SMC_TYPE_R && hdr->typev1 != SMC_TYPE_D)
+		return false;
+	if (hdr->version == SMC_V1) {
+		if ((hdr->typev1 == SMC_TYPE_R &&
+		     ntohs(hdr->length) != SMCR_CLC_ACCEPT_CONFIRM_LEN) ||
+		    (hdr->typev1 == SMC_TYPE_D &&
+		     ntohs(hdr->length) != SMCD_CLC_ACCEPT_CONFIRM_LEN))
+			return false;
+	} else {
+		if (hdr->typev1 == SMC_TYPE_D &&
+		    ntohs(hdr->length) != SMCD_CLC_ACCEPT_CONFIRM_LEN_V2 &&
+		    (ntohs(hdr->length) != SMCD_CLC_ACCEPT_CONFIRM_LEN_V2 +
+				sizeof(struct smc_clc_first_contact_ext)))
+			return false;
+	}
+	return true;
+}
+
+static void smc_clc_fill_fce(struct smc_clc_first_contact_ext *fce, int *len)
+{
+	memset(fce, 0, sizeof(*fce));
+	fce->os_type = SMC_CLC_OS_LINUX;
+	fce->release = SMC_RELEASE;
+	memcpy(fce->hostname, smc_hostname, sizeof(smc_hostname));
+	(*len) += sizeof(*fce);
+}
+
 /* check if received message has a correct header length and contains valid
  * heading and trailing eyecatchers
  */
 static bool smc_clc_msg_hdr_valid(struct smc_clc_msg_hdr *clcm, bool check_trl)
 {
-	struct smc_clc_msg_proposal_prefix *pclc_prfx;
-	struct smc_clc_msg_accept_confirm *clc;
+	struct smc_clc_msg_accept_confirm_v2 *clc_v2;
 	struct smc_clc_msg_proposal *pclc;
 	struct smc_clc_msg_decline *dclc;
 	struct smc_clc_msg_trail *trl;
@@ -51,29 +129,19 @@ static bool smc_clc_msg_hdr_valid(struct smc_clc_msg_hdr *clcm, bool check_trl)
 	switch (clcm->type) {
 	case SMC_CLC_PROPOSAL:
 		pclc = (struct smc_clc_msg_proposal *)clcm;
-		pclc_prfx = smc_clc_proposal_get_prefix(pclc);
-		if (ntohs(pclc->hdr.length) <
-			sizeof(*pclc) + ntohs(pclc->iparea_offset) +
-			sizeof(*pclc_prfx) +
-			pclc_prfx->ipv6_prefixes_cnt *
-				sizeof(struct smc_clc_ipv6_prefix) +
-			sizeof(*trl))
+		if (!smc_clc_msg_prop_valid(pclc))
 			return false;
 		trl = (struct smc_clc_msg_trail *)
 			((u8 *)pclc + ntohs(pclc->hdr.length) - sizeof(*trl));
 		break;
 	case SMC_CLC_ACCEPT:
 	case SMC_CLC_CONFIRM:
-		if (clcm->path != SMC_TYPE_R && clcm->path != SMC_TYPE_D)
-			return false;
-		clc = (struct smc_clc_msg_accept_confirm *)clcm;
-		if ((clcm->path == SMC_TYPE_R &&
-		     ntohs(clc->hdr.length) != SMCR_CLC_ACCEPT_CONFIRM_LEN) ||
-		    (clcm->path == SMC_TYPE_D &&
-		     ntohs(clc->hdr.length) != SMCD_CLC_ACCEPT_CONFIRM_LEN))
+		clc_v2 = (struct smc_clc_msg_accept_confirm_v2 *)clcm;
+		if (!smc_clc_msg_acc_conf_valid(clc_v2))
 			return false;
 		trl = (struct smc_clc_msg_trail *)
-			((u8 *)clc + ntohs(clc->hdr.length) - sizeof(*trl));
+			((u8 *)clc_v2 + ntohs(clc_v2->hdr.length) -
+							sizeof(*trl));
 		break;
 	case SMC_CLC_DECLINE:
 		dclc = (struct smc_clc_msg_decline *)clcm;
@@ -327,9 +395,6 @@ int smc_clc_wait_msg(struct smc_sock *smc, void *buf, int buflen,
 		goto out;
 	}
 
-	if (clcm->type == SMC_CLC_PROPOSAL && clcm->path == SMC_TYPE_N)
-		reason_code = SMC_CLC_DECL_VERSMISMAT; /* just V2 offered */
-
 	/* receive the complete CLC message */
 	memset(&msg, 0, sizeof(struct msghdr));
 	if (datlen > buflen) {
@@ -365,7 +430,8 @@ int smc_clc_wait_msg(struct smc_sock *smc, void *buf, int buflen,
 		dclc = (struct smc_clc_msg_decline *)clcm;
 		reason_code = SMC_CLC_DECL_PEERDECL;
 		smc->peer_diagnosis = ntohl(dclc->peer_diagnosis);
-		if (((struct smc_clc_msg_decline *)buf)->hdr.flag) {
+		if (((struct smc_clc_msg_decline *)buf)->hdr.typev2 &
+						SMC_FIRST_CONTACT_MASK) {
 			smc->conn.lgr->sync_err = 1;
 			smc_lgr_terminate_sched(smc->conn.lgr);
 		}
@@ -377,7 +443,7 @@ int smc_clc_wait_msg(struct smc_sock *smc, void *buf, int buflen,
 }
 
 /* send CLC DECLINE message across internal TCP socket */
-int smc_clc_send_decline(struct smc_sock *smc, u32 peer_diag_info)
+int smc_clc_send_decline(struct smc_sock *smc, u32 peer_diag_info, u8 version)
 {
 	struct smc_clc_msg_decline dclc;
 	struct msghdr msg;
@@ -388,8 +454,10 @@ int smc_clc_send_decline(struct smc_sock *smc, u32 peer_diag_info)
 	memcpy(dclc.hdr.eyecatcher, SMC_EYECATCHER, sizeof(SMC_EYECATCHER));
 	dclc.hdr.type = SMC_CLC_DECLINE;
 	dclc.hdr.length = htons(sizeof(struct smc_clc_msg_decline));
-	dclc.hdr.version = SMC_V1;
-	dclc.hdr.flag = (peer_diag_info == SMC_CLC_DECL_SYNCERR) ? 1 : 0;
+	dclc.hdr.version = version;
+	dclc.os_type = version == SMC_V1 ? 0 : SMC_CLC_OS_LINUX;
+	dclc.hdr.typev2 = (peer_diag_info == SMC_CLC_DECL_SYNCERR) ?
+						SMC_FIRST_CONTACT_MASK : 0;
 	if ((!smc->conn.lgr || !smc->conn.lgr->is_smcd) &&
 	    smc_ib_is_valid_local_systemid())
 		memcpy(dclc.id_for_peer, local_systemid,
@@ -408,18 +476,20 @@ int smc_clc_send_decline(struct smc_sock *smc, u32 peer_diag_info)
 }
 
 /* send CLC PROPOSAL message across internal TCP socket */
-int smc_clc_send_proposal(struct smc_sock *smc, int smc_type,
-			  struct smc_init_info *ini)
+int smc_clc_send_proposal(struct smc_sock *smc, struct smc_init_info *ini)
 {
+	struct smc_clc_smcd_v2_extension *smcd_v2_ext;
 	struct smc_clc_msg_proposal_prefix *pclc_prfx;
 	struct smc_clc_msg_proposal *pclc_base;
+	struct smc_clc_smcd_gid_chid *gidchids;
 	struct smc_clc_msg_proposal_area *pclc;
 	struct smc_clc_ipv6_prefix *ipv6_prfx;
+	struct smc_clc_v2_extension *v2_ext;
 	struct smc_clc_msg_smcd *pclc_smcd;
 	struct smc_clc_msg_trail *trl;
 	int len, i, plen, rc;
 	int reason_code = 0;
-	struct kvec vec[5];
+	struct kvec vec[8];
 	struct msghdr msg;
 
 	pclc = kzalloc(sizeof(*pclc), GFP_KERNEL);
@@ -430,56 +500,121 @@ int smc_clc_send_proposal(struct smc_sock *smc, int smc_type,
 	pclc_smcd = &pclc->pclc_smcd;
 	pclc_prfx = &pclc->pclc_prfx;
 	ipv6_prfx = pclc->pclc_prfx_ipv6;
+	v2_ext = &pclc->pclc_v2_ext;
+	smcd_v2_ext = &pclc->pclc_smcd_v2_ext;
+	gidchids = pclc->pclc_gidchids;
 	trl = &pclc->pclc_trl;
 
+	pclc_base->hdr.version = SMC_V2;
+	pclc_base->hdr.typev1 = ini->smc_type_v1;
+	pclc_base->hdr.typev2 = ini->smc_type_v2;
+	plen = sizeof(*pclc_base) + sizeof(*pclc_smcd) + sizeof(*trl);
+
 	/* retrieve ip prefixes for CLC proposal msg */
-	rc = smc_clc_prfx_set(smc->clcsock, pclc_prfx, ipv6_prfx);
-	if (rc) {
-		kfree(pclc);
-		return SMC_CLC_DECL_CNFERR; /* configuration error */
+	if (ini->smc_type_v1 != SMC_TYPE_N) {
+		rc = smc_clc_prfx_set(smc->clcsock, pclc_prfx, ipv6_prfx);
+		if (rc) {
+			if (ini->smc_type_v2 == SMC_TYPE_N) {
+				kfree(pclc);
+				return SMC_CLC_DECL_CNFERR;
+			}
+			pclc_base->hdr.typev1 = SMC_TYPE_N;
+		} else {
+			pclc_base->iparea_offset = htons(sizeof(*pclc_smcd));
+			plen += sizeof(*pclc_prfx) +
+					pclc_prfx->ipv6_prefixes_cnt *
+					sizeof(ipv6_prfx[0]);
+		}
 	}
 
-	/* send SMC Proposal CLC message */
-	plen = sizeof(*pclc_base) + sizeof(*pclc_prfx) +
-	       (pclc_prfx->ipv6_prefixes_cnt * sizeof(ipv6_prfx[0])) +
-	       sizeof(*trl);
+	/* build SMC Proposal CLC message */
 	memcpy(pclc_base->hdr.eyecatcher, SMC_EYECATCHER,
 	       sizeof(SMC_EYECATCHER));
 	pclc_base->hdr.type = SMC_CLC_PROPOSAL;
-	pclc_base->hdr.version = SMC_V1;		/* SMC version */
-	pclc_base->hdr.path = smc_type;
-	if (smc_type == SMC_TYPE_R || smc_type == SMC_TYPE_B) {
+	if (smcr_indicated(ini->smc_type_v1)) {
 		/* add SMC-R specifics */
 		memcpy(pclc_base->lcl.id_for_peer, local_systemid,
 		       sizeof(local_systemid));
 		memcpy(pclc_base->lcl.gid, ini->ib_gid, SMC_GID_SIZE);
 		memcpy(pclc_base->lcl.mac, &ini->ib_dev->mac[ini->ib_port - 1],
 		       ETH_ALEN);
-		pclc_base->iparea_offset = htons(0);
 	}
-	if (smc_type == SMC_TYPE_D || smc_type == SMC_TYPE_B) {
+	if (smcd_indicated(ini->smc_type_v1)) {
 		/* add SMC-D specifics */
-		plen += sizeof(*pclc_smcd);
-		pclc_base->iparea_offset = htons(sizeof(*pclc_smcd));
-		pclc_smcd->gid = ini->ism_dev->local_gid;
+		if (ini->ism_dev[0]) {
+			pclc_smcd->ism.gid = htonll(ini->ism_dev[0]->local_gid);
+			pclc_smcd->ism.chid =
+				htons(smc_ism_get_chid(ini->ism_dev[0]));
+		}
+	}
+	if (ini->smc_type_v2 == SMC_TYPE_N) {
+		pclc_smcd->v2_ext_offset = 0;
+	} else {
+		u16 v2_ext_offset;
+		u8 *eid = NULL;
+
+		v2_ext_offset = sizeof(*pclc_smcd) -
+			offsetofend(struct smc_clc_msg_smcd, v2_ext_offset);
+		if (ini->smc_type_v1 != SMC_TYPE_N)
+			v2_ext_offset += sizeof(*pclc_prfx) +
+						pclc_prfx->ipv6_prefixes_cnt *
+						sizeof(ipv6_prfx[0]);
+		pclc_smcd->v2_ext_offset = htons(v2_ext_offset);
+		v2_ext->hdr.eid_cnt = 0;
+		v2_ext->hdr.ism_gid_cnt = ini->ism_offered_cnt;
+		v2_ext->hdr.flag.release = SMC_RELEASE;
+		v2_ext->hdr.flag.seid = 1;
+		v2_ext->hdr.smcd_v2_ext_offset = htons(sizeof(*v2_ext) -
+				offsetofend(struct smc_clnt_opts_area_hdr,
+					    smcd_v2_ext_offset) +
+				v2_ext->hdr.eid_cnt * SMC_MAX_EID_LEN);
+		if (ini->ism_dev[0])
+			smc_ism_get_system_eid(ini->ism_dev[0], &eid);
+		else
+			smc_ism_get_system_eid(ini->ism_dev[1], &eid);
+		if (eid)
+			memcpy(smcd_v2_ext->system_eid, eid, SMC_MAX_EID_LEN);
+		plen += sizeof(*v2_ext) + sizeof(*smcd_v2_ext);
+		if (ini->ism_offered_cnt) {
+			for (i = 1; i <= ini->ism_offered_cnt; i++) {
+				gidchids[i - 1].gid =
+					htonll(ini->ism_dev[i]->local_gid);
+				gidchids[i - 1].chid =
+					htons(smc_ism_get_chid(ini->ism_dev[i]));
+			}
+			plen += ini->ism_offered_cnt *
+				sizeof(struct smc_clc_smcd_gid_chid);
+		}
 	}
 	pclc_base->hdr.length = htons(plen);
-
 	memcpy(trl->eyecatcher, SMC_EYECATCHER, sizeof(SMC_EYECATCHER));
+
+	/* send SMC Proposal CLC message */
 	memset(&msg, 0, sizeof(msg));
 	i = 0;
 	vec[i].iov_base = pclc_base;
 	vec[i++].iov_len = sizeof(*pclc_base);
-	if (smc_type == SMC_TYPE_D || smc_type == SMC_TYPE_B) {
-		vec[i].iov_base = pclc_smcd;
-		vec[i++].iov_len = sizeof(*pclc_smcd);
+	vec[i].iov_base = pclc_smcd;
+	vec[i++].iov_len = sizeof(*pclc_smcd);
+	if (ini->smc_type_v1 != SMC_TYPE_N) {
+		vec[i].iov_base = pclc_prfx;
+		vec[i++].iov_len = sizeof(*pclc_prfx);
+		if (pclc_prfx->ipv6_prefixes_cnt > 0) {
+			vec[i].iov_base = ipv6_prfx;
+			vec[i++].iov_len = pclc_prfx->ipv6_prefixes_cnt *
+					   sizeof(ipv6_prfx[0]);
+		}
 	}
-	vec[i].iov_base = pclc_prfx;
-	vec[i++].iov_len = sizeof(*pclc_prfx);
-	if (pclc_prfx->ipv6_prefixes_cnt > 0) {
-		vec[i].iov_base = ipv6_prfx;
-		vec[i++].iov_len = pclc_prfx->ipv6_prefixes_cnt *
-				   sizeof(ipv6_prfx[0]);
+	if (ini->smc_type_v2 != SMC_TYPE_N) {
+		vec[i].iov_base = v2_ext;
+		vec[i++].iov_len = sizeof(*v2_ext);
+		vec[i].iov_base = smcd_v2_ext;
+		vec[i++].iov_len = sizeof(*smcd_v2_ext);
+		if (ini->ism_offered_cnt) {
+			vec[i].iov_base = gidchids;
+			vec[i++].iov_len = ini->ism_offered_cnt *
+					sizeof(struct smc_clc_smcd_gid_chid);
+		}
 	}
 	vec[i].iov_base = trl;
 	vec[i++].iov_len = sizeof(*trl);
@@ -499,29 +634,47 @@ int smc_clc_send_proposal(struct smc_sock *smc, int smc_type,
 
 /* build and send CLC CONFIRM / ACCEPT message */
 static int smc_clc_send_confirm_accept(struct smc_sock *smc,
-				       struct smc_clc_msg_accept_confirm *clc,
-				       int first_contact)
+				       struct smc_clc_msg_accept_confirm_v2 *clc_v2,
+				       int first_contact, u8 version)
 {
 	struct smc_connection *conn = &smc->conn;
+	struct smc_clc_msg_accept_confirm *clc;
+	struct smc_clc_first_contact_ext fce;
+	struct smc_clc_msg_trail trl;
+	struct kvec vec[3];
 	struct msghdr msg;
-	struct kvec vec;
+	int i, len;
 
 	/* send SMC Confirm CLC msg */
-	clc->hdr.version = SMC_V1;		/* SMC version */
+	clc = (struct smc_clc_msg_accept_confirm *)clc_v2;
+	clc->hdr.version = version;	/* SMC version */
 	if (first_contact)
-		clc->hdr.flag = 1;
+		clc->hdr.typev2 |= SMC_FIRST_CONTACT_MASK;
 	if (conn->lgr->is_smcd) {
 		/* SMC-D specific settings */
 		memcpy(clc->hdr.eyecatcher, SMCD_EYECATCHER,
 		       sizeof(SMCD_EYECATCHER));
-		clc->hdr.path = SMC_TYPE_D;
-		clc->hdr.length = htons(SMCD_CLC_ACCEPT_CONFIRM_LEN);
+		clc->hdr.typev1 = SMC_TYPE_D;
 		clc->d0.gid = conn->lgr->smcd->local_gid;
 		clc->d0.token = conn->rmb_desc->token;
 		clc->d0.dmbe_size = conn->rmbe_size_short;
 		clc->d0.dmbe_idx = 0;
 		memcpy(&clc->d0.linkid, conn->lgr->id, SMC_LGR_ID_SIZE);
-		memcpy(clc->d0.smcd_trl.eyecatcher, SMCD_EYECATCHER,
+		if (version == SMC_V1) {
+			clc->hdr.length = htons(SMCD_CLC_ACCEPT_CONFIRM_LEN);
+		} else {
+			u8 *eid = NULL;
+
+			clc_v2->chid = htons(smc_ism_get_chid(conn->lgr->smcd));
+			smc_ism_get_system_eid(conn->lgr->smcd, &eid);
+			if (eid)
+				memcpy(clc_v2->eid, eid, SMC_MAX_EID_LEN);
+			len = SMCD_CLC_ACCEPT_CONFIRM_LEN_V2;
+			if (first_contact)
+				smc_clc_fill_fce(&fce, &len);
+			clc_v2->hdr.length = htons(len);
+		}
+		memcpy(trl.eyecatcher, SMCD_EYECATCHER,
 		       sizeof(SMCD_EYECATCHER));
 	} else {
 		struct smc_link *link = conn->lnk;
@@ -530,7 +683,7 @@ static int smc_clc_send_confirm_accept(struct smc_sock *smc,
 		link = conn->lnk;
 		memcpy(clc->hdr.eyecatcher, SMC_EYECATCHER,
 		       sizeof(SMC_EYECATCHER));
-		clc->hdr.path = SMC_TYPE_R;
+		clc->hdr.typev1 = SMC_TYPE_R;
 		clc->hdr.length = htons(SMCR_CLC_ACCEPT_CONFIRM_LEN);
 		memcpy(clc->r0.lcl.id_for_peer, local_systemid,
 		       sizeof(local_systemid));
@@ -554,29 +707,43 @@ static int smc_clc_send_confirm_accept(struct smc_sock *smc,
 		clc->r0.rmb_dma_addr = cpu_to_be64((u64)sg_dma_address
 				(conn->rmb_desc->sgt[link->link_idx].sgl));
 		hton24(clc->r0.psn, link->psn_initial);
-		memcpy(clc->r0.smcr_trl.eyecatcher, SMC_EYECATCHER,
-		       sizeof(SMC_EYECATCHER));
+		memcpy(trl.eyecatcher, SMC_EYECATCHER, sizeof(SMC_EYECATCHER));
 	}
 
 	memset(&msg, 0, sizeof(msg));
-	vec.iov_base = clc;
-	vec.iov_len = ntohs(clc->hdr.length);
-	return kernel_sendmsg(smc->clcsock, &msg, &vec, 1,
+	i = 0;
+	vec[i].iov_base = clc_v2;
+	if (version > SMC_V1)
+		vec[i++].iov_len = SMCD_CLC_ACCEPT_CONFIRM_LEN_V2 - sizeof(trl);
+	else
+		vec[i++].iov_len = (clc->hdr.typev1 == SMC_TYPE_D ?
+						SMCD_CLC_ACCEPT_CONFIRM_LEN :
+						SMCR_CLC_ACCEPT_CONFIRM_LEN) -
+				   sizeof(trl);
+	if (version > SMC_V1 && first_contact) {
+		vec[i].iov_base = &fce;
+		vec[i++].iov_len = sizeof(fce);
+	}
+	vec[i].iov_base = &trl;
+	vec[i++].iov_len = sizeof(trl);
+	return kernel_sendmsg(smc->clcsock, &msg, vec, 1,
 			      ntohs(clc->hdr.length));
 }
 
 /* send CLC CONFIRM message across internal TCP socket */
-int smc_clc_send_confirm(struct smc_sock *smc)
+int smc_clc_send_confirm(struct smc_sock *smc, bool clnt_first_contact,
+			 u8 version)
 {
-	struct smc_clc_msg_accept_confirm cclc;
+	struct smc_clc_msg_accept_confirm_v2 cclc_v2;
 	int reason_code = 0;
 	int len;
 
 	/* send SMC Confirm CLC msg */
-	memset(&cclc, 0, sizeof(cclc));
-	cclc.hdr.type = SMC_CLC_CONFIRM;
-	len = smc_clc_send_confirm_accept(smc, &cclc, 0);
-	if (len < ntohs(cclc.hdr.length)) {
+	memset(&cclc_v2, 0, sizeof(cclc_v2));
+	cclc_v2.hdr.type = SMC_CLC_CONFIRM;
+	len = smc_clc_send_confirm_accept(smc, &cclc_v2, clnt_first_contact,
+					  version);
+	if (len < ntohs(cclc_v2.hdr.length)) {
 		if (len >= 0) {
 			reason_code = -ENETUNREACH;
 			smc->sk.sk_err = -reason_code;
@@ -589,16 +756,28 @@ int smc_clc_send_confirm(struct smc_sock *smc)
 }
 
 /* send CLC ACCEPT message across internal TCP socket */
-int smc_clc_send_accept(struct smc_sock *new_smc, bool srv_first_contact)
+int smc_clc_send_accept(struct smc_sock *new_smc, bool srv_first_contact,
+			u8 version)
 {
-	struct smc_clc_msg_accept_confirm aclc;
+	struct smc_clc_msg_accept_confirm_v2 aclc_v2;
 	int len;
 
-	memset(&aclc, 0, sizeof(aclc));
-	aclc.hdr.type = SMC_CLC_ACCEPT;
-	len = smc_clc_send_confirm_accept(new_smc, &aclc, srv_first_contact);
-	if (len < ntohs(aclc.hdr.length))
+	memset(&aclc_v2, 0, sizeof(aclc_v2));
+	aclc_v2.hdr.type = SMC_CLC_ACCEPT;
+	len = smc_clc_send_confirm_accept(new_smc, &aclc_v2, srv_first_contact,
+					  version);
+	if (len < ntohs(aclc_v2.hdr.length))
 		len = len >= 0 ? -EPROTO : -new_smc->clcsock->sk->sk_err;
 
 	return len > 0 ? 0 : len;
 }
+
+void __init smc_clc_init(void)
+{
+	struct new_utsname *u;
+
+	memset(smc_hostname, _S, sizeof(smc_hostname)); /* ASCII blanks */
+	u = utsname();
+	memcpy(smc_hostname, u->nodename,
+	       min_t(size_t, strlen(u->nodename), sizeof(smc_hostname)));
+}

net/smc/smc_clc.h

@@ -54,19 +54,19 @@
 #define SMC_CLC_DECL_ERR_RDYLNK	0x09990002  /*	 ib ready link failed	      */
 #define SMC_CLC_DECL_ERR_REGRMB	0x09990003  /*	 reg rmb failed		      */
 
+#define SMC_FIRST_CONTACT_MASK	0b10	/* first contact bit within typev2 */
+
 struct smc_clc_msg_hdr {	/* header1 of clc messages */
 	u8 eyecatcher[4];	/* eye catcher */
 	u8 type;		/* proposal / accept / confirm / decline */
 	__be16 length;
 #if defined(__BIG_ENDIAN_BITFIELD)
 	u8 version : 4,
-	   flag    : 1,
-	   rsvd	   : 1,
-	   path	   : 2;
+	   typev2  : 2,
+	   typev1  : 2;
 #elif defined(__LITTLE_ENDIAN_BITFIELD)
-	u8 path    : 2,
-	   rsvd    : 1,
-	   flag    : 1,
+	u8 typev1  : 2,
+	   typev2  : 2,
 	   version : 4;
 #endif
 } __packed;			/* format defined in RFC7609 */
@@ -81,8 +81,6 @@ struct smc_clc_msg_local {	/* header2 of clc messages */
 	u8 mac[6];		/* mac of ib_device port */
 };
 
-#define SMC_CLC_MAX_V6_PREFIX	8
-
 /* Struct would be 4 byte aligned, but it is used in an array that is sent
  * to peers and must conform to RFC7609, hence we need to use packed here.
  */
@@ -91,6 +89,44 @@ struct smc_clc_ipv6_prefix {
 	u8 prefix_len;
 } __packed;			/* format defined in RFC7609 */
 
+#if defined(__BIG_ENDIAN_BITFIELD)
+struct smc_clc_v2_flag {
+	u8 release : 4,
+	   rsvd    : 3,
+	   seid    : 1;
+};
+#elif defined(__LITTLE_ENDIAN_BITFIELD)
+struct smc_clc_v2_flag {
+	u8 seid   : 1,
+	rsvd      : 3,
+	release   : 4;
+};
+#endif
+
+struct smc_clnt_opts_area_hdr {
+	u8 eid_cnt;		/* number of user defined EIDs */
+	u8 ism_gid_cnt;		/* number of ISMv2 GIDs */
+	u8 reserved1;
+	struct smc_clc_v2_flag flag;
+	u8 reserved2[2];
+	__be16 smcd_v2_ext_offset; /* SMC-Dv2 Extension Offset */
+};
+
+struct smc_clc_smcd_gid_chid {
+	__be64 gid;		/* ISM GID */
+	__be16 chid;		/* ISMv2 CHID */
+} __packed;		/* format defined in
+			 * IBM Shared Memory Communications Version 2
+			 * (https://www.ibm.com/support/pages/node/6326337)
+			 */
+
+struct smc_clc_v2_extension {
+	struct smc_clnt_opts_area_hdr hdr;
+	u8 roce[16];		/* RoCEv2 GID */
+	u8 reserved[16];
+	u8 user_eids[0][SMC_MAX_EID_LEN];
+};
+
 struct smc_clc_msg_proposal_prefix {	/* prefix part of clc proposal message*/
 	__be32 outgoing_subnet;	/* subnet mask */
 	u8 prefix_len;		/* number of significant bits in mask */
@@ -99,8 +135,15 @@ struct smc_clc_msg_proposal_prefix {	/* prefix part of clc proposal message*/
 } __aligned(4);
 
 struct smc_clc_msg_smcd {	/* SMC-D GID information */
-	u64 gid;		/* ISM GID of requestor */
-	u8 res[32];
+	struct smc_clc_smcd_gid_chid ism; /* ISM native GID+CHID of requestor */
+	__be16 v2_ext_offset;	/* SMC Version 2 Extension Offset */
+	u8 reserved[28];
+};
+
+struct smc_clc_smcd_v2_extension {
+	u8 system_eid[SMC_MAX_EID_LEN];
+	u8 reserved[16];
+	struct smc_clc_smcd_gid_chid gidchid[0];
 };
 
 struct smc_clc_msg_proposal {	/* clc proposal message sent by Linux */
@@ -109,11 +152,16 @@ struct smc_clc_msg_proposal {	/* clc proposal message sent by Linux */
 	__be16 iparea_offset;	/* offset to IP address information area */
 } __aligned(4);
 
+#define SMC_CLC_MAX_V6_PREFIX		8
+
 struct smc_clc_msg_proposal_area {
 	struct smc_clc_msg_proposal		pclc_base;
 	struct smc_clc_msg_smcd			pclc_smcd;
 	struct smc_clc_msg_proposal_prefix	pclc_prfx;
 	struct smc_clc_ipv6_prefix	pclc_prfx_ipv6[SMC_CLC_MAX_V6_PREFIX];
+	struct smc_clc_v2_extension		pclc_v2_ext;
+	struct smc_clc_smcd_v2_extension	pclc_smcd_v2_ext;
+	struct smc_clc_smcd_gid_chid		pclc_gidchids[SMC_MAX_ISM_DEVS];
 	struct smc_clc_msg_trail		pclc_trl;
 };
 
@@ -134,11 +182,9 @@ struct smcr_clc_msg_accept_confirm {	/* SMCR accept/confirm */
 	__be64 rmb_dma_addr;	/* RMB virtual address */
 	u8 reserved2;
 	u8 psn[3];		/* packet sequence number */
-	struct smc_clc_msg_trail smcr_trl;
-				/* eye catcher "SMCR" EBCDIC */
 } __packed;
 
-struct smcd_clc_msg_accept_confirm {	/* SMCD accept/confirm */
+struct smcd_clc_msg_accept_confirm_common {	/* SMCD accept/confirm */
 	u64 gid;		/* Sender GID */
 	u64 token;		/* DMB token */
 	u8 dmbe_idx;		/* DMBE index */
@@ -150,26 +196,63 @@ struct smcd_clc_msg_accept_confirm {	/* SMCD accept/confirm */
 	   dmbe_size : 4;
 #endif
 	u16 reserved4;
-	u32 linkid;		/* Link identifier */
-	u32 reserved5[3];
-	struct smc_clc_msg_trail smcd_trl;
-				/* eye catcher "SMCD" EBCDIC */
+	__be32 linkid;		/* Link identifier */
 } __packed;
 
+#define SMC_CLC_OS_ZOS		1
+#define SMC_CLC_OS_LINUX	2
+#define SMC_CLC_OS_AIX		3
+
+struct smc_clc_first_contact_ext {
+	u8 reserved1;
+#if defined(__BIG_ENDIAN_BITFIELD)
+	u8 os_type : 4,
+	   release : 4;
+#elif defined(__LITTLE_ENDIAN_BITFIELD)
+	u8 release : 4,
+	   os_type : 4;
+#endif
+	u8 reserved2[2];
+	u8 hostname[SMC_MAX_HOSTNAME_LEN];
+};
+
 struct smc_clc_msg_accept_confirm {	/* clc accept / confirm message */
 	struct smc_clc_msg_hdr hdr;
 	union {
 		struct smcr_clc_msg_accept_confirm r0; /* SMC-R */
-		struct smcd_clc_msg_accept_confirm d0; /* SMC-D */
+		struct { /* SMC-D */
+			struct smcd_clc_msg_accept_confirm_common d0;
+			u32 reserved5[3];
+		};
 	};
 } __packed;			/* format defined in RFC7609 */
 
+struct smc_clc_msg_accept_confirm_v2 {	/* clc accept / confirm message */
+	struct smc_clc_msg_hdr hdr;
+	union {
+		struct smcr_clc_msg_accept_confirm r0; /* SMC-R */
+		struct { /* SMC-D */
+			struct smcd_clc_msg_accept_confirm_common d0;
+			__be16 chid;
+			u8 eid[SMC_MAX_EID_LEN];
+			u8 reserved5[8];
+		};
+	};
+};
+
 struct smc_clc_msg_decline {	/* clc decline message */
 	struct smc_clc_msg_hdr hdr;
 	u8 id_for_peer[SMC_SYSTEMID_LEN]; /* sender peer_id */
 	__be32 peer_diagnosis;	/* diagnosis information */
-	u8 reserved2[4];
-	struct smc_clc_msg_trail trl; /* eye catcher "SMCR" EBCDIC */
+#if defined(__BIG_ENDIAN_BITFIELD)
+	u8 os_type  : 4,
+	   reserved : 4;
+#elif defined(__LITTLE_ENDIAN_BITFIELD)
+	u8 reserved : 4,
+	   os_type  : 4;
+#endif
+	u8 reserved2[3];
+	struct smc_clc_msg_trail trl; /* eye catcher "SMCD" or "SMCR" EBCDIC */
 } __aligned(4);
 
 /* determine start of the prefix area within the proposal message */
@@ -180,16 +263,58 @@ smc_clc_proposal_get_prefix(struct smc_clc_msg_proposal *pclc)
 	       ((u8 *)pclc + sizeof(*pclc) + ntohs(pclc->iparea_offset));
 }
 
+static inline bool smcr_indicated(int smc_type)
+{
+	return smc_type == SMC_TYPE_R || smc_type == SMC_TYPE_B;
+}
+
+static inline bool smcd_indicated(int smc_type)
+{
+	return smc_type == SMC_TYPE_D || smc_type == SMC_TYPE_B;
+}
+
 /* get SMC-D info from proposal message */
 static inline struct smc_clc_msg_smcd *
 smc_get_clc_msg_smcd(struct smc_clc_msg_proposal *prop)
 {
-	if (ntohs(prop->iparea_offset) != sizeof(struct smc_clc_msg_smcd))
+	if (smcd_indicated(prop->hdr.typev1) &&
+	    ntohs(prop->iparea_offset) != sizeof(struct smc_clc_msg_smcd))
 		return NULL;
 
 	return (struct smc_clc_msg_smcd *)(prop + 1);
 }
 
+static inline struct smc_clc_v2_extension *
+smc_get_clc_v2_ext(struct smc_clc_msg_proposal *prop)
+{
+	struct smc_clc_msg_smcd *prop_smcd = smc_get_clc_msg_smcd(prop);
+
+	if (!prop_smcd || !ntohs(prop_smcd->v2_ext_offset))
+		return NULL;
+
+	return (struct smc_clc_v2_extension *)
+	       ((u8 *)prop_smcd +
+	       offsetof(struct smc_clc_msg_smcd, v2_ext_offset) +
+	       sizeof(prop_smcd->v2_ext_offset) +
+	       ntohs(prop_smcd->v2_ext_offset));
+}
+
+static inline struct smc_clc_smcd_v2_extension *
+smc_get_clc_smcd_v2_ext(struct smc_clc_v2_extension *prop_v2ext)
+{
+	if (!prop_v2ext)
+		return NULL;
+	if (!ntohs(prop_v2ext->hdr.smcd_v2_ext_offset))
+		return NULL;
+
+	return (struct smc_clc_smcd_v2_extension *)
+		((u8 *)prop_v2ext +
+		 offsetof(struct smc_clc_v2_extension, hdr) +
+		 offsetof(struct smc_clnt_opts_area_hdr, smcd_v2_ext_offset) +
+		 sizeof(prop_v2ext->hdr.smcd_v2_ext_offset) +
+		 ntohs(prop_v2ext->hdr.smcd_v2_ext_offset));
+}
+
 struct smcd_dev;
 struct smc_init_info;
 
@@ -197,10 +322,12 @@ int smc_clc_prfx_match(struct socket *clcsock,
 		       struct smc_clc_msg_proposal_prefix *prop);
 int smc_clc_wait_msg(struct smc_sock *smc, void *buf, int buflen,
 		     u8 expected_type, unsigned long timeout);
-int smc_clc_send_decline(struct smc_sock *smc, u32 peer_diag_info);
-int smc_clc_send_proposal(struct smc_sock *smc, int smc_type,
-			  struct smc_init_info *ini);
-int smc_clc_send_confirm(struct smc_sock *smc);
-int smc_clc_send_accept(struct smc_sock *smc, bool srv_first_contact);
+int smc_clc_send_decline(struct smc_sock *smc, u32 peer_diag_info, u8 version);
+int smc_clc_send_proposal(struct smc_sock *smc, struct smc_init_info *ini);
+int smc_clc_send_confirm(struct smc_sock *smc, bool clnt_first_contact,
+			 u8 version);
+int smc_clc_send_accept(struct smc_sock *smc, bool srv_first_contact,
+			u8 version);
+void smc_clc_init(void) __init;
 
 #endif

net/smc/smc_core.c

@@ -375,7 +375,8 @@ static int smc_lgr_create(struct smc_sock *smc, struct smc_init_info *ini)
 	int i;
 
 	if (ini->is_smcd && ini->vlan_id) {
-		if (smc_ism_get_vlan(ini->ism_dev, ini->vlan_id)) {
+		if (smc_ism_get_vlan(ini->ism_dev[ini->ism_selected],
+				     ini->vlan_id)) {
 			rc = SMC_CLC_DECL_ISMVLANERR;
 			goto out;
 		}
@@ -412,13 +413,14 @@ static int smc_lgr_create(struct smc_sock *smc, struct smc_init_info *ini)
 	lgr->conns_all = RB_ROOT;
 	if (ini->is_smcd) {
 		/* SMC-D specific settings */
-		get_device(&ini->ism_dev->dev);
-		lgr->peer_gid = ini->ism_peer_gid;
-		lgr->smcd = ini->ism_dev;
-		lgr_list = &ini->ism_dev->lgr_list;
+		get_device(&ini->ism_dev[ini->ism_selected]->dev);
+		lgr->peer_gid = ini->ism_peer_gid[ini->ism_selected];
+		lgr->smcd = ini->ism_dev[ini->ism_selected];
+		lgr_list = &ini->ism_dev[ini->ism_selected]->lgr_list;
 		lgr_lock = &lgr->smcd->lgr_lock;
+		lgr->smc_version = ini->smcd_version;
 		lgr->peer_shutdown = 0;
-		atomic_inc(&ini->ism_dev->lgr_cnt);
+		atomic_inc(&ini->ism_dev[ini->ism_selected]->lgr_cnt);
 	} else {
 		/* SMC-R specific settings */
 		lgr->role = smc->listen_smc ? SMC_SERV : SMC_CLNT;
@@ -449,7 +451,7 @@ static int smc_lgr_create(struct smc_sock *smc, struct smc_init_info *ini)
 	kfree(lgr);
 ism_put_vlan:
 	if (ini->is_smcd && ini->vlan_id)
-		smc_ism_put_vlan(ini->ism_dev, ini->vlan_id);
+		smc_ism_put_vlan(ini->ism_dev[ini->ism_selected], ini->vlan_id);
 out:
 	if (rc < 0) {
 		if (rc == -ENOMEM)
@@ -1288,8 +1290,10 @@ int smc_conn_create(struct smc_sock *smc, struct smc_init_info *ini)
 	spinlock_t *lgr_lock;
 	int rc = 0;
 
-	lgr_list = ini->is_smcd ? &ini->ism_dev->lgr_list : &smc_lgr_list.list;
-	lgr_lock = ini->is_smcd ? &ini->ism_dev->lgr_lock : &smc_lgr_list.lock;
+	lgr_list = ini->is_smcd ? &ini->ism_dev[ini->ism_selected]->lgr_list :
+				  &smc_lgr_list.list;
+	lgr_lock = ini->is_smcd ? &ini->ism_dev[ini->ism_selected]->lgr_lock :
+				  &smc_lgr_list.lock;
 	ini->first_contact_local = 1;
 	role = smc->listen_smc ? SMC_SERV : SMC_CLNT;
 	if (role == SMC_CLNT && ini->first_contact_peer)
@@ -1301,7 +1305,8 @@ int smc_conn_create(struct smc_sock *smc, struct smc_init_info *ini)
 	list_for_each_entry(lgr, lgr_list, list) {
 		write_lock_bh(&lgr->conns_lock);
 		if ((ini->is_smcd ?
-		     smcd_lgr_match(lgr, ini->ism_dev, ini->ism_peer_gid) :
+		     smcd_lgr_match(lgr, ini->ism_dev[ini->ism_selected],
+				    ini->ism_peer_gid[ini->ism_selected]) :
 		     smcr_lgr_match(lgr, ini->ib_lcl, role, ini->ib_clcqpn)) &&
 		    !lgr->sync_err &&
 		    lgr->vlan_id == ini->vlan_id &&

net/smc/smc_core.h

@@ -231,6 +231,11 @@ struct smc_link_group {
 	u8			freeing : 1;	/* lgr is being freed */
 
 	bool			is_smcd;	/* SMC-R or SMC-D */
+	u8			smc_version;
+	u8			negotiated_eid[SMC_MAX_EID_LEN];
+	u8			peer_os;	/* peer operating system */
+	u8			peer_smc_release;
+	u8			peer_hostname[SMC_MAX_HOSTNAME_LEN];
 	union {
 		struct { /* SMC-R */
 			enum smc_lgr_role	role;
@@ -291,6 +296,8 @@ struct smc_clc_msg_local;
 
 struct smc_init_info {
 	u8			is_smcd;
+	u8			smc_type_v1;
+	u8			smc_type_v2;
 	u8			first_contact_peer;
 	u8			first_contact_local;
 	unsigned short		vlan_id;
@@ -301,8 +308,12 @@ struct smc_init_info {
 	u8			ib_port;
 	u32			ib_clcqpn;
 	/* SMC-D */
-	u64			ism_peer_gid;
-	struct smcd_dev		*ism_dev;
+	u64			ism_peer_gid[SMC_MAX_ISM_DEVS + 1];
+	struct smcd_dev		*ism_dev[SMC_MAX_ISM_DEVS + 1];
+	u16			ism_chid[SMC_MAX_ISM_DEVS + 1];
+	u8			ism_offered_cnt; /* # of ISM devices offered */
+	u8			ism_selected;    /* index of selected ISM dev*/
+	u8			smcd_version;
 };
 
 /* Find the connection associated with the given alert token in the link group.

net/smc/smc_ism.c

@@ -21,7 +21,9 @@ struct smcd_dev_list smcd_dev_list = {
 	.mutex = __MUTEX_INITIALIZER(smcd_dev_list.mutex)
 };
 
-/* Test if an ISM communication is possible. */
+bool smc_ism_v2_capable;
+
+/* Test if an ISM communication is possible - same CPC */
 int smc_ism_cantalk(u64 peer_gid, unsigned short vlan_id, struct smcd_dev *smcd)
 {
 	return smcd->ops->query_remote_gid(smcd, peer_gid, vlan_id ? 1 : 0,
@@ -39,6 +41,16 @@ int smc_ism_write(struct smcd_dev *smcd, const struct smc_ism_position *pos,
 	return rc < 0 ? rc : 0;
 }
 
+void smc_ism_get_system_eid(struct smcd_dev *smcd, u8 **eid)
+{
+	smcd->ops->get_system_eid(smcd, eid);
+}
+
+u16 smc_ism_get_chid(struct smcd_dev *smcd)
+{
+	return smcd->ops->get_chid(smcd);
+}
+
 /* Set a connection using this DMBE. */
 void smc_ism_set_conn(struct smc_connection *conn)
 {
@@ -319,7 +331,18 @@ EXPORT_SYMBOL_GPL(smcd_alloc_dev);
 int smcd_register_dev(struct smcd_dev *smcd)
 {
 	mutex_lock(&smcd_dev_list.mutex);
-	list_add_tail(&smcd->list, &smcd_dev_list.list);
+	if (list_empty(&smcd_dev_list.list)) {
+		u8 *system_eid = NULL;
+
+		smc_ism_get_system_eid(smcd, &system_eid);
+		if ((*system_eid) + 24 != '0' || (*system_eid) + 28 != '0')
+			smc_ism_v2_capable = true;
+	}
+	/* sort list: devices without pnetid before devices with pnetid */
+	if (smcd->pnetid[0])
+		list_add_tail(&smcd->list, &smcd_dev_list.list);
+	else
+		list_add(&smcd->list, &smcd_dev_list.list);
 	mutex_unlock(&smcd_dev_list.mutex);
 
 	pr_warn_ratelimited("smc: adding smcd device %s with pnetid %.16s%s\n",
@@ -399,3 +422,8 @@ void smcd_handle_irq(struct smcd_dev *smcd, unsigned int dmbno)
 	spin_unlock_irqrestore(&smcd->lock, flags);
 }
 EXPORT_SYMBOL_GPL(smcd_handle_irq);
+
+void __init smc_ism_init(void)
+{
+	smc_ism_v2_capable = false;
+}

net/smc/smc_ism.h

@@ -19,7 +19,10 @@ struct smcd_dev_list {	/* List of SMCD devices */
 	struct mutex mutex;	/* Protects list of devices */
 };
 
-extern struct smcd_dev_list	smcd_dev_list; /* list of smcd devices */
+extern struct smcd_dev_list	smcd_dev_list;	/* list of smcd devices */
+extern bool	smc_ism_v2_capable;	/* HW supports ISM V2 and thus
+					 * System EID is defined
+					 */
 
 struct smc_ism_vlanid {			/* VLAN id set on ISM device */
 	struct list_head list;
@@ -47,4 +50,7 @@ int smc_ism_unregister_dmb(struct smcd_dev *dev, struct smc_buf_desc *dmb_desc);
 int smc_ism_write(struct smcd_dev *dev, const struct smc_ism_position *pos,
 		  void *data, size_t len);
 int smc_ism_signal_shutdown(struct smc_link_group *lgr);
+void smc_ism_get_system_eid(struct smcd_dev *dev, u8 **eid);
+u16 smc_ism_get_chid(struct smcd_dev *dev);
+void smc_ism_init(void);
 #endif

net/smc/smc_netns.h

@@ -16,5 +16,6 @@ extern unsigned int smc_net_id;
 /* per-network namespace private data */
 struct smc_net {
 	struct smc_pnettable pnettable;
+	struct smc_pnetids_ndev pnetids_ndev;
 };
 #endif

net/smc/smc_pnet.c

@@ -29,8 +29,7 @@
 #include "smc_ism.h"
 #include "smc_core.h"
 
-#define SMC_ASCII_BLANK 32
-
+static struct net_device *__pnet_find_base_ndev(struct net_device *ndev);
 static struct net_device *pnet_find_base_ndev(struct net_device *ndev);
 
 static const struct nla_policy smc_pnet_policy[SMC_PNETID_MAX + 1] = {
@@ -73,14 +72,22 @@ struct smc_pnetentry {
 	};
 };
 
+/* Check if the pnetid is set */
+bool smc_pnet_is_pnetid_set(u8 *pnetid)
+{
+	if (pnetid[0] == 0 || pnetid[0] == _S)
+		return false;
+	return true;
+}
+
 /* Check if two given pnetids match */
 static bool smc_pnet_match(u8 *pnetid1, u8 *pnetid2)
 {
 	int i;
 
 	for (i = 0; i < SMC_MAX_PNETID_LEN; i++) {
-		if ((pnetid1[i] == 0 || pnetid1[i] == SMC_ASCII_BLANK) &&
-		    (pnetid2[i] == 0 || pnetid2[i] == SMC_ASCII_BLANK))
+		if ((pnetid1[i] == 0 || pnetid1[i] == _S) &&
+		    (pnetid2[i] == 0 || pnetid2[i] == _S))
 			break;
 		if (pnetid1[i] != pnetid2[i])
 			return false;
@@ -238,11 +245,10 @@ static int smc_pnet_remove_by_ndev(struct net_device *ndev)
 static bool smc_pnet_apply_ib(struct smc_ib_device *ib_dev, u8 ib_port,
 			      char *pnet_name)
 {
-	u8 pnet_null[SMC_MAX_PNETID_LEN] = {0};
 	bool applied = false;
 
 	mutex_lock(&smc_ib_devices.mutex);
-	if (smc_pnet_match(ib_dev->pnetid[ib_port - 1], pnet_null)) {
+	if (!smc_pnet_is_pnetid_set(ib_dev->pnetid[ib_port - 1])) {
 		memcpy(ib_dev->pnetid[ib_port - 1], pnet_name,
 		       SMC_MAX_PNETID_LEN);
 		ib_dev->pnetid_by_user[ib_port - 1] = true;
@@ -256,11 +262,10 @@ static bool smc_pnet_apply_ib(struct smc_ib_device *ib_dev, u8 ib_port,
  */
 static bool smc_pnet_apply_smcd(struct smcd_dev *smcd_dev, char *pnet_name)
 {
-	u8 pnet_null[SMC_MAX_PNETID_LEN] = {0};
 	bool applied = false;
 
 	mutex_lock(&smcd_dev_list.mutex);
-	if (smc_pnet_match(smcd_dev->pnetid, pnet_null)) {
+	if (!smc_pnet_is_pnetid_set(smcd_dev->pnetid)) {
 		memcpy(smcd_dev->pnetid, pnet_name, SMC_MAX_PNETID_LEN);
 		smcd_dev->pnetid_by_user = true;
 		applied = true;
@@ -708,10 +713,115 @@ static struct genl_family smc_pnet_nl_family __ro_after_init = {
 	.n_ops =  ARRAY_SIZE(smc_pnet_ops)
 };
 
+bool smc_pnet_is_ndev_pnetid(struct net *net, u8 *pnetid)
+{
+	struct smc_net *sn = net_generic(net, smc_net_id);
+	struct smc_pnetids_ndev_entry *pe;
+	bool rc = false;
+
+	read_lock(&sn->pnetids_ndev.lock);
+	list_for_each_entry(pe, &sn->pnetids_ndev.list, list) {
+		if (smc_pnet_match(pnetid, pe->pnetid)) {
+			rc = true;
+			goto unlock;
+		}
+	}
+
+unlock:
+	read_unlock(&sn->pnetids_ndev.lock);
+	return rc;
+}
+
+static int smc_pnet_add_pnetid(struct net *net, u8 *pnetid)
+{
+	struct smc_net *sn = net_generic(net, smc_net_id);
+	struct smc_pnetids_ndev_entry *pe, *pi;
+
+	pe = kzalloc(sizeof(*pe), GFP_KERNEL);
+	if (!pe)
+		return -ENOMEM;
+
+	write_lock(&sn->pnetids_ndev.lock);
+	list_for_each_entry(pi, &sn->pnetids_ndev.list, list) {
+		if (smc_pnet_match(pnetid, pe->pnetid)) {
+			refcount_inc(&pi->refcnt);
+			kfree(pe);
+			goto unlock;
+		}
+	}
+	refcount_set(&pe->refcnt, 1);
+	memcpy(pe->pnetid, pnetid, SMC_MAX_PNETID_LEN);
+	list_add_tail(&pe->list, &sn->pnetids_ndev.list);
+
+unlock:
+	write_unlock(&sn->pnetids_ndev.lock);
+	return 0;
+}
+
+static void smc_pnet_remove_pnetid(struct net *net, u8 *pnetid)
+{
+	struct smc_net *sn = net_generic(net, smc_net_id);
+	struct smc_pnetids_ndev_entry *pe, *pe2;
+
+	write_lock(&sn->pnetids_ndev.lock);
+	list_for_each_entry_safe(pe, pe2, &sn->pnetids_ndev.list, list) {
+		if (smc_pnet_match(pnetid, pe->pnetid)) {
+			if (refcount_dec_and_test(&pe->refcnt)) {
+				list_del(&pe->list);
+				kfree(pe);
+			}
+			break;
+		}
+	}
+	write_unlock(&sn->pnetids_ndev.lock);
+}
+
+static void smc_pnet_add_base_pnetid(struct net *net, struct net_device *dev,
+				     u8 *ndev_pnetid)
+{
+	struct net_device *base_dev;
+
+	base_dev = __pnet_find_base_ndev(dev);
+	if (base_dev->flags & IFF_UP &&
+	    !smc_pnetid_by_dev_port(base_dev->dev.parent, base_dev->dev_port,
+				    ndev_pnetid)) {
+		/* add to PNETIDs list */
+		smc_pnet_add_pnetid(net, ndev_pnetid);
+	}
+}
+
+/* create initial list of netdevice pnetids */
+static void smc_pnet_create_pnetids_list(struct net *net)
+{
+	u8 ndev_pnetid[SMC_MAX_PNETID_LEN];
+	struct net_device *dev;
+
+	rtnl_lock();
+	for_each_netdev(net, dev)
+		smc_pnet_add_base_pnetid(net, dev, ndev_pnetid);
+	rtnl_unlock();
+}
+
+/* clean up list of netdevice pnetids */
+static void smc_pnet_destroy_pnetids_list(struct net *net)
+{
+	struct smc_net *sn = net_generic(net, smc_net_id);
+	struct smc_pnetids_ndev_entry *pe, *temp_pe;
+
+	write_lock(&sn->pnetids_ndev.lock);
+	list_for_each_entry_safe(pe, temp_pe, &sn->pnetids_ndev.list, list) {
+		list_del(&pe->list);
+		kfree(pe);
+	}
+	write_unlock(&sn->pnetids_ndev.lock);
+}
+
 static int smc_pnet_netdev_event(struct notifier_block *this,
 				 unsigned long event, void *ptr)
 {
 	struct net_device *event_dev = netdev_notifier_info_to_dev(ptr);
+	struct net *net = dev_net(event_dev);
+	u8 ndev_pnetid[SMC_MAX_PNETID_LEN];
 
 	switch (event) {
 	case NETDEV_REBOOT:
@@ -721,6 +831,17 @@ static int smc_pnet_netdev_event(struct notifier_block *this,
 	case NETDEV_REGISTER:
 		smc_pnet_add_by_ndev(event_dev);
 		return NOTIFY_OK;
+	case NETDEV_UP:
+		smc_pnet_add_base_pnetid(net, event_dev, ndev_pnetid);
+		return NOTIFY_OK;
+	case NETDEV_DOWN:
+		event_dev = __pnet_find_base_ndev(event_dev);
+		if (!smc_pnetid_by_dev_port(event_dev->dev.parent,
+					    event_dev->dev_port, ndev_pnetid)) {
+			/* remove from PNETIDs list */
+			smc_pnet_remove_pnetid(net, ndev_pnetid);
+		}
+		return NOTIFY_OK;
 	default:
 		return NOTIFY_DONE;
 	}
@@ -735,9 +856,14 @@ int smc_pnet_net_init(struct net *net)
 {
 	struct smc_net *sn = net_generic(net, smc_net_id);
 	struct smc_pnettable *pnettable = &sn->pnettable;
+	struct smc_pnetids_ndev *pnetids_ndev = &sn->pnetids_ndev;
 
 	INIT_LIST_HEAD(&pnettable->pnetlist);
 	rwlock_init(&pnettable->lock);
+	INIT_LIST_HEAD(&pnetids_ndev->list);
+	rwlock_init(&pnetids_ndev->lock);
+
+	smc_pnet_create_pnetids_list(net);
 
 	return 0;
 }
@@ -752,6 +878,7 @@ int __init smc_pnet_init(void)
 	rc = register_netdevice_notifier(&smc_netdev_notifier);
 	if (rc)
 		genl_unregister_family(&smc_pnet_nl_family);
+
 	return rc;
 }
 
@@ -760,6 +887,7 @@ void smc_pnet_net_exit(struct net *net)
 {
 	/* flush pnet table */
 	smc_pnet_remove_by_pnetid(net, NULL);
+	smc_pnet_destroy_pnetids_list(net);
 }
 
 void smc_pnet_exit(void)
@@ -768,16 +896,11 @@ void smc_pnet_exit(void)
 	genl_unregister_family(&smc_pnet_nl_family);
 }
 
-/* Determine one base device for stacked net devices.
- * If the lower device level contains more than one devices
- * (for instance with bonding slaves), just the first device
- * is used to reach a base device.
- */
-static struct net_device *pnet_find_base_ndev(struct net_device *ndev)
+static struct net_device *__pnet_find_base_ndev(struct net_device *ndev)
 {
 	int i, nest_lvl;
 
-	rtnl_lock();
+	ASSERT_RTNL();
 	nest_lvl = ndev->lower_level;
 	for (i = 0; i < nest_lvl; i++) {
 		struct list_head *lower = &ndev->adj_list.lower;
@@ -787,6 +910,18 @@ static struct net_device *pnet_find_base_ndev(struct net_device *ndev)
 		lower = lower->next;
 		ndev = netdev_lower_get_next(ndev, &lower);
 	}
+	return ndev;
+}
+
+/* Determine one base device for stacked net devices.
+ * If the lower device level contains more than one devices
+ * (for instance with bonding slaves), just the first device
+ * is used to reach a base device.
+ */
+static struct net_device *pnet_find_base_ndev(struct net_device *ndev)
+{
+	rtnl_lock();
+	ndev = __pnet_find_base_ndev(ndev);
 	rtnl_unlock();
 	return ndev;
 }
@@ -929,10 +1064,10 @@ static void smc_pnet_find_ism_by_pnetid(struct net_device *ndev,
 	list_for_each_entry(ismdev, &smcd_dev_list.list, list) {
 		if (smc_pnet_match(ismdev->pnetid, ndev_pnetid) &&
 		    !ismdev->going_away &&
-		    (!ini->ism_peer_gid ||
-		     !smc_ism_cantalk(ini->ism_peer_gid, ini->vlan_id,
+		    (!ini->ism_peer_gid[0] ||
+		     !smc_ism_cantalk(ini->ism_peer_gid[0], ini->vlan_id,
 				      ismdev))) {
-			ini->ism_dev = ismdev;
+			ini->ism_dev[0] = ismdev;
 			break;
 		}
 	}
@@ -966,7 +1101,7 @@ void smc_pnet_find_ism_resource(struct sock *sk, struct smc_init_info *ini)
 {
 	struct dst_entry *dst = sk_dst_get(sk);
 
-	ini->ism_dev = NULL;
+	ini->ism_dev[0] = NULL;
 	if (!dst)
 		goto out;
 	if (!dst->dev)

net/smc/smc_pnet.h

@@ -12,6 +12,8 @@
 #ifndef _SMC_PNET_H
 #define _SMC_PNET_H
 
+#include <net/smc.h>
+
 #if IS_ENABLED(CONFIG_HAVE_PNETID)
 #include <asm/pnet.h>
 #endif
@@ -31,6 +33,17 @@ struct smc_pnettable {
 	struct list_head pnetlist;
 };
 
+struct smc_pnetids_ndev {	/* list of pnetids for net devices in UP state*/
+	struct list_head	list;
+	rwlock_t		lock;
+};
+
+struct smc_pnetids_ndev_entry {
+	struct list_head	list;
+	u8			pnetid[SMC_MAX_PNETID_LEN];
+	refcount_t		refcnt;
+};
+
 static inline int smc_pnetid_by_dev_port(struct device *dev,
 					 unsigned short port, u8 *pnetid)
 {
@@ -52,4 +65,6 @@ int smc_pnetid_by_table_smcd(struct smcd_dev *smcd);
 void smc_pnet_find_alt_roce(struct smc_link_group *lgr,
 			    struct smc_init_info *ini,
 			    struct smc_ib_device *known_dev);
+bool smc_pnet_is_ndev_pnetid(struct net *net, u8 *pnetid);
+bool smc_pnet_is_pnetid_set(u8 *pnetid);
 #endif