Merge tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma

Pull rdma fixes from Jason Gunthorpe:
 "We had a fairly slow cycle on the rc side this time, here are the
  accumulated fixes, mostly in drivers:

   - irdma should not generate extra completions during flushing

   - Fix several memory leaks

   - Do not get confused in irdma's iwarp mode if IPv6 is present

   - Correct a link speed calculation in mlx5

   - Increase the EQ/WQ limits on erdma as they are too small for big
     applications

   - Use the right math for erdma's inline mtt feature

   - Make erdma probing more robust to boot time ordering differences

   - Fix a KMSAN crash in CMA due to uninitialized qkey"

* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma:
  RDMA/core: Fix GID entry ref leak when create_ah fails
  RDMA/cma: Allow UD qp_type to join multicast only
  RDMA/erdma: Defer probing if netdevice can not be found
  RDMA/erdma: Inline mtt entries into WQE if supported
  RDMA/erdma: Update default EQ depth to 4096 and max_send_wr to 8192
  RDMA/erdma: Fix some typos
  IB/mlx5: Add support for 400G_8X lane speed
  RDMA/irdma: Add ipv4 check to irdma_find_listener()
  RDMA/irdma: Increase iWARP CM default rexmit count
  RDMA/irdma: Fix memory leak of PBLE objects
  RDMA/irdma: Do not generate SW completions for NOPs

drivers/infiniband/core/cma.c

@@ -624,22 +624,11 @@ static inline unsigned short cma_family(struct rdma_id_private *id_priv)
 	return id_priv->id.route.addr.src_addr.ss_family;
 }
 
-static int cma_set_qkey(struct rdma_id_private *id_priv, u32 qkey)
+static int cma_set_default_qkey(struct rdma_id_private *id_priv)
 {
 	struct ib_sa_mcmember_rec rec;
 	int ret = 0;
 
-	if (id_priv->qkey) {
-		if (qkey && id_priv->qkey != qkey)
-			return -EINVAL;
-		return 0;
-	}
-
-	if (qkey) {
-		id_priv->qkey = qkey;
-		return 0;
-	}
-
 	switch (id_priv->id.ps) {
 	case RDMA_PS_UDP:
 	case RDMA_PS_IB:
@@ -659,6 +648,16 @@ static int cma_set_qkey(struct rdma_id_private *id_priv, u32 qkey)
 	return ret;
 }
 
+static int cma_set_qkey(struct rdma_id_private *id_priv, u32 qkey)
+{
+	if (!qkey ||
+	    (id_priv->qkey && (id_priv->qkey != qkey)))
+		return -EINVAL;
+
+	id_priv->qkey = qkey;
+	return 0;
+}
+
 static void cma_translate_ib(struct sockaddr_ib *sib, struct rdma_dev_addr *dev_addr)
 {
 	dev_addr->dev_type = ARPHRD_INFINIBAND;
@@ -1229,7 +1228,7 @@ static int cma_ib_init_qp_attr(struct rdma_id_private *id_priv,
 	*qp_attr_mask = IB_QP_STATE | IB_QP_PKEY_INDEX | IB_QP_PORT;
 
 	if (id_priv->id.qp_type == IB_QPT_UD) {
-		ret = cma_set_qkey(id_priv, 0);
+		ret = cma_set_default_qkey(id_priv);
 		if (ret)
 			return ret;
 
@@ -4569,7 +4568,10 @@ static int cma_send_sidr_rep(struct rdma_id_private *id_priv,
 	memset(&rep, 0, sizeof rep);
 	rep.status = status;
 	if (status == IB_SIDR_SUCCESS) {
-		ret = cma_set_qkey(id_priv, qkey);
+		if (qkey)
+			ret = cma_set_qkey(id_priv, qkey);
+		else
+			ret = cma_set_default_qkey(id_priv);
 		if (ret)
 			return ret;
 		rep.qp_num = id_priv->qp_num;
@@ -4774,9 +4776,7 @@ static void cma_make_mc_event(int status, struct rdma_id_private *id_priv,
 	enum ib_gid_type gid_type;
 	struct net_device *ndev;
 
-	if (!status)
-		status = cma_set_qkey(id_priv, be32_to_cpu(multicast->rec.qkey));
-	else
+	if (status)
 		pr_debug_ratelimited("RDMA CM: MULTICAST_ERROR: failed to join multicast. status %d\n",
 				     status);
 
@@ -4804,7 +4804,7 @@ static void cma_make_mc_event(int status, struct rdma_id_private *id_priv,
 	}
 
 	event->param.ud.qp_num = 0xFFFFFF;
-	event->param.ud.qkey = be32_to_cpu(multicast->rec.qkey);
+	event->param.ud.qkey = id_priv->qkey;
 
 out:
 	if (ndev)
@@ -4823,8 +4823,11 @@ static int cma_ib_mc_handler(int status, struct ib_sa_multicast *multicast)
 	    READ_ONCE(id_priv->state) == RDMA_CM_DESTROYING)
 		goto out;
 
-	cma_make_mc_event(status, id_priv, multicast, &event, mc);
-	ret = cma_cm_event_handler(id_priv, &event);
+	ret = cma_set_qkey(id_priv, be32_to_cpu(multicast->rec.qkey));
+	if (!ret) {
+		cma_make_mc_event(status, id_priv, multicast, &event, mc);
+		ret = cma_cm_event_handler(id_priv, &event);
+	}
 	rdma_destroy_ah_attr(&event.param.ud.ah_attr);
 	WARN_ON(ret);
 
@@ -4877,9 +4880,11 @@ static int cma_join_ib_multicast(struct rdma_id_private *id_priv,
 	if (ret)
 		return ret;
 
-	ret = cma_set_qkey(id_priv, 0);
-	if (ret)
-		return ret;
+	if (!id_priv->qkey) {
+		ret = cma_set_default_qkey(id_priv);
+		if (ret)
+			return ret;
+	}
 
 	cma_set_mgid(id_priv, (struct sockaddr *) &mc->addr, &rec.mgid);
 	rec.qkey = cpu_to_be32(id_priv->qkey);
@@ -4956,9 +4961,6 @@ static int cma_iboe_join_multicast(struct rdma_id_private *id_priv,
 	cma_iboe_set_mgid(addr, &ib.rec.mgid, gid_type);
 
 	ib.rec.pkey = cpu_to_be16(0xffff);
-	if (id_priv->id.ps == RDMA_PS_UDP)
-		ib.rec.qkey = cpu_to_be32(RDMA_UDP_QKEY);
-
 	if (dev_addr->bound_dev_if)
 		ndev = dev_get_by_index(dev_addr->net, dev_addr->bound_dev_if);
 	if (!ndev)
@@ -4984,6 +4986,9 @@ static int cma_iboe_join_multicast(struct rdma_id_private *id_priv,
 	if (err || !ib.rec.mtu)
 		return err ?: -EINVAL;
 
+	if (!id_priv->qkey)
+		cma_set_default_qkey(id_priv);
+
 	rdma_ip2gid((struct sockaddr *)&id_priv->id.route.addr.src_addr,
 		    &ib.rec.port_gid);
 	INIT_WORK(&mc->iboe_join.work, cma_iboe_join_work_handler);
@@ -5009,6 +5014,9 @@ int rdma_join_multicast(struct rdma_cm_id *id, struct sockaddr *addr,
 			    READ_ONCE(id_priv->state) != RDMA_CM_ADDR_RESOLVED))
 		return -EINVAL;
 
+	if (id_priv->id.qp_type != IB_QPT_UD)
+		return -EINVAL;
+
 	mc = kzalloc(sizeof(*mc), GFP_KERNEL);
 	if (!mc)
 		return -ENOMEM;

drivers/infiniband/core/verbs.c

@@ -532,6 +532,8 @@ static struct ib_ah *_rdma_create_ah(struct ib_pd *pd,
 	else
 		ret = device->ops.create_ah(ah, &init_attr, NULL);
 	if (ret) {
+		if (ah->sgid_attr)
+			rdma_put_gid_attr(ah->sgid_attr);
 		kfree(ah);
 		return ERR_PTR(ret);
 	}

drivers/infiniband/hw/erdma/erdma_cq.c

@@ -65,7 +65,7 @@ static const enum ib_wc_opcode wc_mapping_table[ERDMA_NUM_OPCODES] = {
 	[ERDMA_OP_LOCAL_INV] = IB_WC_LOCAL_INV,
 	[ERDMA_OP_READ_WITH_INV] = IB_WC_RDMA_READ,
 	[ERDMA_OP_ATOMIC_CAS] = IB_WC_COMP_SWAP,
-	[ERDMA_OP_ATOMIC_FAD] = IB_WC_FETCH_ADD,
+	[ERDMA_OP_ATOMIC_FAA] = IB_WC_FETCH_ADD,
 };
 
 static const struct {

drivers/infiniband/hw/erdma/erdma_hw.h

@@ -441,7 +441,7 @@ struct erdma_reg_mr_sqe {
 };
 
 /* EQ related. */
-#define ERDMA_DEFAULT_EQ_DEPTH 256
+#define ERDMA_DEFAULT_EQ_DEPTH 4096
 
 /* ceqe */
 #define ERDMA_CEQE_HDR_DB_MASK BIT_ULL(63)
@@ -491,7 +491,7 @@ enum erdma_opcode {
 	ERDMA_OP_LOCAL_INV = 15,
 	ERDMA_OP_READ_WITH_INV = 16,
 	ERDMA_OP_ATOMIC_CAS = 17,
-	ERDMA_OP_ATOMIC_FAD = 18,
+	ERDMA_OP_ATOMIC_FAA = 18,
 	ERDMA_NUM_OPCODES = 19,
 	ERDMA_OP_INVALID = ERDMA_NUM_OPCODES + 1
 };

drivers/infiniband/hw/erdma/erdma_main.c

@@ -56,7 +56,7 @@ static int erdma_netdev_event(struct notifier_block *nb, unsigned long event,
 static int erdma_enum_and_get_netdev(struct erdma_dev *dev)
 {
 	struct net_device *netdev;
-	int ret = -ENODEV;
+	int ret = -EPROBE_DEFER;
 
 	/* Already binded to a net_device, so we skip. */
 	if (dev->netdev)

drivers/infiniband/hw/erdma/erdma_qp.c

@@ -405,7 +405,7 @@ static int erdma_push_one_sqe(struct erdma_qp *qp, u16 *pi,
 			FIELD_PREP(ERDMA_SQE_MR_MTT_CNT_MASK,
 				   mr->mem.mtt_nents);
 
-		if (mr->mem.mtt_nents < ERDMA_MAX_INLINE_MTT_ENTRIES) {
+		if (mr->mem.mtt_nents <= ERDMA_MAX_INLINE_MTT_ENTRIES) {
 			attrs |= FIELD_PREP(ERDMA_SQE_MR_MTT_TYPE_MASK, 0);
 			/* Copy SGLs to SQE content to accelerate */
 			memcpy(get_queue_entry(qp->kern_qp.sq_buf, idx + 1,
@@ -439,7 +439,7 @@ static int erdma_push_one_sqe(struct erdma_qp *qp, u16 *pi,
 				cpu_to_le64(atomic_wr(send_wr)->compare_add);
 		} else {
 			wqe_hdr |= FIELD_PREP(ERDMA_SQE_HDR_OPCODE_MASK,
-					      ERDMA_OP_ATOMIC_FAD);
+					      ERDMA_OP_ATOMIC_FAA);
 			atomic_sqe->fetchadd_swap_data =
 				cpu_to_le64(atomic_wr(send_wr)->compare_add);
 		}

drivers/infiniband/hw/erdma/erdma_verbs.h

@@ -11,7 +11,7 @@
 
 /* RDMA Capability. */
 #define ERDMA_MAX_PD (128 * 1024)
-#define ERDMA_MAX_SEND_WR 4096
+#define ERDMA_MAX_SEND_WR 8192
 #define ERDMA_MAX_ORD 128
 #define ERDMA_MAX_IRD 128
 #define ERDMA_MAX_SGE_RD 1

drivers/infiniband/hw/irdma/cm.c

@@ -1458,13 +1458,15 @@ static int irdma_send_fin(struct irdma_cm_node *cm_node)
  * irdma_find_listener - find a cm node listening on this addr-port pair
  * @cm_core: cm's core
  * @dst_addr: listener ip addr
+ * @ipv4: flag indicating IPv4 when true
  * @dst_port: listener tcp port num
  * @vlan_id: virtual LAN ID
  * @listener_state: state to match with listen node's
  */
 static struct irdma_cm_listener *
-irdma_find_listener(struct irdma_cm_core *cm_core, u32 *dst_addr, u16 dst_port,
-		    u16 vlan_id, enum irdma_cm_listener_state listener_state)
+irdma_find_listener(struct irdma_cm_core *cm_core, u32 *dst_addr, bool ipv4,
+		    u16 dst_port, u16 vlan_id,
+		    enum irdma_cm_listener_state listener_state)
 {
 	struct irdma_cm_listener *listen_node;
 	static const u32 ip_zero[4] = { 0, 0, 0, 0 };
@@ -1477,7 +1479,7 @@ irdma_find_listener(struct irdma_cm_core *cm_core, u32 *dst_addr, u16 dst_port,
 	list_for_each_entry (listen_node, &cm_core->listen_list, list) {
 		memcpy(listen_addr, listen_node->loc_addr, sizeof(listen_addr));
 		listen_port = listen_node->loc_port;
-		if (listen_port != dst_port ||
+		if (listen_node->ipv4 != ipv4 || listen_port != dst_port ||
 		    !(listener_state & listen_node->listener_state))
 			continue;
 		/* compare node pair, return node handle if a match */
@@ -2902,9 +2904,10 @@ irdma_make_listen_node(struct irdma_cm_core *cm_core,
 	unsigned long flags;
 
 	/* cannot have multiple matching listeners */
-	listener = irdma_find_listener(cm_core, cm_info->loc_addr,
-				       cm_info->loc_port, cm_info->vlan_id,
-				       IRDMA_CM_LISTENER_EITHER_STATE);
+	listener =
+		irdma_find_listener(cm_core, cm_info->loc_addr, cm_info->ipv4,
+				    cm_info->loc_port, cm_info->vlan_id,
+				    IRDMA_CM_LISTENER_EITHER_STATE);
 	if (listener &&
 	    listener->listener_state == IRDMA_CM_LISTENER_ACTIVE_STATE) {
 		refcount_dec(&listener->refcnt);
@@ -3153,6 +3156,7 @@ void irdma_receive_ilq(struct irdma_sc_vsi *vsi, struct irdma_puda_buf *rbuf)
 
 		listener = irdma_find_listener(cm_core,
 					       cm_info.loc_addr,
+					       cm_info.ipv4,
 					       cm_info.loc_port,
 					       cm_info.vlan_id,
 					       IRDMA_CM_LISTENER_ACTIVE_STATE);

drivers/infiniband/hw/irdma/cm.h

@@ -41,7 +41,7 @@
 #define TCP_OPTIONS_PADDING	3
 
 #define IRDMA_DEFAULT_RETRYS	64
-#define IRDMA_DEFAULT_RETRANS	8
+#define IRDMA_DEFAULT_RETRANS	32
 #define IRDMA_DEFAULT_TTL		0x40
 #define IRDMA_DEFAULT_RTT_VAR		6
 #define IRDMA_DEFAULT_SS_THRESH		0x3fffffff

drivers/infiniband/hw/irdma/hw.c

@@ -41,6 +41,7 @@ static enum irdma_hmc_rsrc_type iw_hmc_obj_types[] = {
 	IRDMA_HMC_IW_XFFL,
 	IRDMA_HMC_IW_Q1,
 	IRDMA_HMC_IW_Q1FL,
+	IRDMA_HMC_IW_PBLE,
 	IRDMA_HMC_IW_TIMER,
 	IRDMA_HMC_IW_FSIMC,
 	IRDMA_HMC_IW_FSIAV,
@@ -827,6 +828,8 @@ static int irdma_create_hmc_objs(struct irdma_pci_f *rf, bool privileged,
 	info.entry_type = rf->sd_type;
 
 	for (i = 0; i < IW_HMC_OBJ_TYPE_NUM; i++) {
+		if (iw_hmc_obj_types[i] == IRDMA_HMC_IW_PBLE)
+			continue;
 		if (dev->hmc_info->hmc_obj[iw_hmc_obj_types[i]].cnt) {
 			info.rsrc_type = iw_hmc_obj_types[i];
 			info.count = dev->hmc_info->hmc_obj[info.rsrc_type].cnt;

drivers/infiniband/hw/irdma/utils.c

@@ -2595,7 +2595,10 @@ void irdma_generate_flush_completions(struct irdma_qp *iwqp)
 			/* remove the SQ WR by moving SQ tail*/
 			IRDMA_RING_SET_TAIL(*sq_ring,
 				sq_ring->tail + qp->sq_wrtrk_array[sq_ring->tail].quanta);
-
+			if (cmpl->cpi.op_type == IRDMAQP_OP_NOP) {
+				kfree(cmpl);
+				continue;
+			}
 			ibdev_dbg(iwqp->iwscq->ibcq.device,
 				  "DEV: %s: adding wr_id = 0x%llx SQ Completion to list qp_id=%d\n",
 				  __func__, cmpl->cpi.wr_id, qp->qp_id);

drivers/infiniband/hw/mlx5/main.c

@@ -442,6 +442,10 @@ static int translate_eth_ext_proto_oper(u32 eth_proto_oper, u16 *active_speed,
 		*active_width = IB_WIDTH_2X;
 		*active_speed = IB_SPEED_NDR;
 		break;
+	case MLX5E_PROT_MASK(MLX5E_400GAUI_8):
+		*active_width = IB_WIDTH_8X;
+		*active_speed = IB_SPEED_HDR;
+		break;
 	case MLX5E_PROT_MASK(MLX5E_400GAUI_4_400GBASE_CR4_KR4):
 		*active_width = IB_WIDTH_4X;
 		*active_speed = IB_SPEED_NDR;