Merge tag 'rdma-for-3.6' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband

Pull InfiniBand/RDMA changes from Roland Dreier:
 - Updates to the qib low-level driver
 - First chunk of changes for SR-IOV support for mlx4 IB
 - RDMA CM support for IPv6-only binding
 - Other misc cleanups and fixes

Fix up some add-add conflicts in include/linux/mlx4/device.h and
drivers/net/ethernet/mellanox/mlx4/main.c

* tag 'rdma-for-3.6' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband: (30 commits)
  IB/qib: checkpatch fixes
  IB/qib: Add congestion control agent implementation
  IB/qib: Reduce sdma_lock contention
  IB/qib: Fix an incorrect log message
  IB/qib: Fix QP RCU sparse warnings
  mlx4: Put physical GID and P_Key table sizes in mlx4_phys_caps struct and paravirtualize them
  mlx4_core: Allow guests to have IB ports
  mlx4_core: Implement mechanism for reserved Q_Keys
  net/mlx4_core: Free ICM table in case of error
  IB/cm: Destroy idr as part of the module init error flow
  mlx4_core: Remove double function declarations
  IB/mlx4: Fill the masked_atomic_cap attribute in query device
  IB/mthca: Fill in sq_sig_type in query QP
  IB/mthca: Warning about event for non-existent QPs should show event type
  IB/qib: Fix sparse RCU warnings in qib_keys.c
  net/mlx4_core: Initialize IB port capabilities for all slaves
  mlx4: Use port management change event instead of smp_snoop
  IB/qib: RCU locking for MR validation
  IB/qib: Avoid returning EBUSY from MR deregister
  IB/qib: Fix UC MR refs for immediate operations
  ...

4 files changed, 182 insertions, 38 deletions

diff --git a/drivers/infiniband/hw/mlx4/mad.c b/drivers/infiniband/hw/mlx4/mad.c
index 259b0670b51..c27141fef1a 100644
--- a/drivers/infiniband/hw/mlx4/mad.c
+++ b/drivers/infiniband/hw/mlx4/mad.c
@@ -147,47 +147,51 @@ static void update_sm_ah(struct mlx4_ib_dev *dev, u8 port_num, u16 lid, u8 sl)
 }
 
 /*
- * Snoop SM MADs for port info and P_Key table sets, so we can
- * synthesize LID change and P_Key change events.
+ * Snoop SM MADs for port info, GUID info, and  P_Key table sets, so we can
+ * synthesize LID change, Client-Rereg, GID change, and P_Key change events.
  */
 static void smp_snoop(struct ib_device *ibdev, u8 port_num, struct ib_mad *mad,
-				u16 prev_lid)
+		      u16 prev_lid)
 {
-	struct ib_event event;
+	struct ib_port_info *pinfo;
+	u16 lid;
 
+	struct mlx4_ib_dev *dev = to_mdev(ibdev);
 	if ((mad->mad_hdr.mgmt_class == IB_MGMT_CLASS_SUBN_LID_ROUTED ||
 	     mad->mad_hdr.mgmt_class == IB_MGMT_CLASS_SUBN_DIRECTED_ROUTE) &&
-	    mad->mad_hdr.method == IB_MGMT_METHOD_SET) {
-		if (mad->mad_hdr.attr_id == IB_SMP_ATTR_PORT_INFO) {
-			struct ib_port_info *pinfo =
-				(struct ib_port_info *) ((struct ib_smp *) mad)->data;
-			u16 lid = be16_to_cpu(pinfo->lid);
+	    mad->mad_hdr.method == IB_MGMT_METHOD_SET)
+		switch (mad->mad_hdr.attr_id) {
+		case IB_SMP_ATTR_PORT_INFO:
+			pinfo = (struct ib_port_info *) ((struct ib_smp *) mad)->data;
+			lid = be16_to_cpu(pinfo->lid);
 
-			update_sm_ah(to_mdev(ibdev), port_num,
+			update_sm_ah(dev, port_num,
 				     be16_to_cpu(pinfo->sm_lid),
 				     pinfo->neighbormtu_mastersmsl & 0xf);
 
-			event.device	       = ibdev;
-			event.element.port_num = port_num;
+			if (pinfo->clientrereg_resv_subnetto & 0x80)
+				mlx4_ib_dispatch_event(dev, port_num,
+						       IB_EVENT_CLIENT_REREGISTER);
 
-			if (pinfo->clientrereg_resv_subnetto & 0x80) {
-				event.event    = IB_EVENT_CLIENT_REREGISTER;
-				ib_dispatch_event(&event);
-			}
+			if (prev_lid != lid)
+				mlx4_ib_dispatch_event(dev, port_num,
+						       IB_EVENT_LID_CHANGE);
+			break;
 
-			if (prev_lid != lid) {
-				event.event    = IB_EVENT_LID_CHANGE;
-				ib_dispatch_event(&event);
-			}
-		}
+		case IB_SMP_ATTR_PKEY_TABLE:
+			mlx4_ib_dispatch_event(dev, port_num,
+					       IB_EVENT_PKEY_CHANGE);
+			break;
 
-		if (mad->mad_hdr.attr_id == IB_SMP_ATTR_PKEY_TABLE) {
-			event.device	       = ibdev;
-			event.event	       = IB_EVENT_PKEY_CHANGE;
-			event.element.port_num = port_num;
-			ib_dispatch_event(&event);
+		case IB_SMP_ATTR_GUID_INFO:
+			/* paravirtualized master's guid is guid 0 -- does not change */
+			if (!mlx4_is_master(dev->dev))
+				mlx4_ib_dispatch_event(dev, port_num,
+						       IB_EVENT_GID_CHANGE);
+			break;
+		default:
+			break;
 		}
-	}
 }
 
 static void node_desc_override(struct ib_device *dev,
@@ -242,6 +246,25 @@ static int ib_process_mad(struct ib_device *ibdev, int mad_flags, u8 port_num,
 	int err;
 	struct ib_port_attr pattr;
 
+	if (in_wc && in_wc->qp->qp_num) {
+		pr_debug("received MAD: slid:%d sqpn:%d "
+			"dlid_bits:%d dqpn:%d wc_flags:0x%x, cls %x, mtd %x, atr %x\n",
+			in_wc->slid, in_wc->src_qp,
+			in_wc->dlid_path_bits,
+			in_wc->qp->qp_num,
+			in_wc->wc_flags,
+			in_mad->mad_hdr.mgmt_class, in_mad->mad_hdr.method,
+			be16_to_cpu(in_mad->mad_hdr.attr_id));
+		if (in_wc->wc_flags & IB_WC_GRH) {
+			pr_debug("sgid_hi:0x%016llx sgid_lo:0x%016llx\n",
+				 be64_to_cpu(in_grh->sgid.global.subnet_prefix),
+				 be64_to_cpu(in_grh->sgid.global.interface_id));
+			pr_debug("dgid_hi:0x%016llx dgid_lo:0x%016llx\n",
+				 be64_to_cpu(in_grh->dgid.global.subnet_prefix),
+				 be64_to_cpu(in_grh->dgid.global.interface_id));
+		}
+	}
+
 	slid = in_wc ? in_wc->slid : be16_to_cpu(IB_LID_PERMISSIVE);
 
 	if (in_mad->mad_hdr.method == IB_MGMT_METHOD_TRAP && slid == 0) {
@@ -286,7 +309,8 @@ static int ib_process_mad(struct ib_device *ibdev, int mad_flags, u8 port_num,
 		return IB_MAD_RESULT_FAILURE;
 
 	if (!out_mad->mad_hdr.status) {
-		smp_snoop(ibdev, port_num, in_mad, prev_lid);
+		if (!(to_mdev(ibdev)->dev->caps.flags & MLX4_DEV_CAP_FLAG_PORT_MNG_CHG_EV))
+			smp_snoop(ibdev, port_num, in_mad, prev_lid);
 		node_desc_override(ibdev, out_mad);
 	}
 
@@ -427,3 +451,64 @@ void mlx4_ib_mad_cleanup(struct mlx4_ib_dev *dev)
 			ib_destroy_ah(dev->sm_ah[p]);
 	}
 }
+
+void handle_port_mgmt_change_event(struct work_struct *work)
+{
+	struct ib_event_work *ew = container_of(work, struct ib_event_work, work);
+	struct mlx4_ib_dev *dev = ew->ib_dev;
+	struct mlx4_eqe *eqe = &(ew->ib_eqe);
+	u8 port = eqe->event.port_mgmt_change.port;
+	u32 changed_attr;
+
+	switch (eqe->subtype) {
+	case MLX4_DEV_PMC_SUBTYPE_PORT_INFO:
+		changed_attr = be32_to_cpu(eqe->event.port_mgmt_change.params.port_info.changed_attr);
+
+		/* Update the SM ah - This should be done before handling
+		   the other changed attributes so that MADs can be sent to the SM */
+		if (changed_attr & MSTR_SM_CHANGE_MASK) {
+			u16 lid = be16_to_cpu(eqe->event.port_mgmt_change.params.port_info.mstr_sm_lid);
+			u8 sl = eqe->event.port_mgmt_change.params.port_info.mstr_sm_sl & 0xf;
+			update_sm_ah(dev, port, lid, sl);
+		}
+
+		/* Check if it is a lid change event */
+		if (changed_attr & MLX4_EQ_PORT_INFO_LID_CHANGE_MASK)
+			mlx4_ib_dispatch_event(dev, port, IB_EVENT_LID_CHANGE);
+
+		/* Generate GUID changed event */
+		if (changed_attr & MLX4_EQ_PORT_INFO_GID_PFX_CHANGE_MASK)
+			mlx4_ib_dispatch_event(dev, port, IB_EVENT_GID_CHANGE);
+
+		if (changed_attr & MLX4_EQ_PORT_INFO_CLIENT_REREG_MASK)
+			mlx4_ib_dispatch_event(dev, port,
+					       IB_EVENT_CLIENT_REREGISTER);
+		break;
+
+	case MLX4_DEV_PMC_SUBTYPE_PKEY_TABLE:
+		mlx4_ib_dispatch_event(dev, port, IB_EVENT_PKEY_CHANGE);
+		break;
+	case MLX4_DEV_PMC_SUBTYPE_GUID_INFO:
+		/* paravirtualized master's guid is guid 0 -- does not change */
+		if (!mlx4_is_master(dev->dev))
+			mlx4_ib_dispatch_event(dev, port, IB_EVENT_GID_CHANGE);
+		break;
+	default:
+		pr_warn("Unsupported subtype 0x%x for "
+			"Port Management Change event\n", eqe->subtype);
+	}
+
+	kfree(ew);
+}
+
+void mlx4_ib_dispatch_event(struct mlx4_ib_dev *dev, u8 port_num,
+			    enum ib_event_type type)
+{
+	struct ib_event event;
+
+	event.device		= &dev->ib_dev;
+	event.element.port_num	= port_num;
+	event.event		= type;
+
+	ib_dispatch_event(&event);
+}
diff --git a/drivers/infiniband/hw/mlx4/main.c b/drivers/infiniband/hw/mlx4/main.c
index a07b774e786..fe2088cfa6e 100644
--- a/drivers/infiniband/hw/mlx4/main.c
+++ b/drivers/infiniband/hw/mlx4/main.c
@@ -50,7 +50,7 @@
 #include "mlx4_ib.h"
 #include "user.h"
 
-#define DRV_NAME	"mlx4_ib"
+#define DRV_NAME	MLX4_IB_DRV_NAME
 #define DRV_VERSION	"1.0"
 #define DRV_RELDATE	"April 4, 2008"
 
@@ -157,7 +157,7 @@ static int mlx4_ib_query_device(struct ib_device *ibdev,
 	props->local_ca_ack_delay  = dev->dev->caps.local_ca_ack_delay;
 	props->atomic_cap	   = dev->dev->caps.flags & MLX4_DEV_CAP_FLAG_ATOMIC ?
 		IB_ATOMIC_HCA : IB_ATOMIC_NONE;
-	props->masked_atomic_cap   = IB_ATOMIC_HCA;
+	props->masked_atomic_cap   = props->atomic_cap;
 	props->max_pkeys	   = dev->dev->caps.pkey_table_len[1];
 	props->max_mcast_grp	   = dev->dev->caps.num_mgms + dev->dev->caps.num_amgms;
 	props->max_mcast_qp_attach = dev->dev->caps.num_qp_per_mgm;
@@ -946,7 +946,6 @@ static void update_gids_task(struct work_struct *work)
 	union ib_gid *gids;
 	int err;
 	struct mlx4_dev	*dev = gw->dev->dev;
-	struct ib_event event;
 
 	mailbox = mlx4_alloc_cmd_mailbox(dev);
 	if (IS_ERR(mailbox)) {
@@ -964,10 +963,7 @@ static void update_gids_task(struct work_struct *work)
 		pr_warn("set port command failed\n");
 	else {
 		memcpy(gw->dev->iboe.gid_table[gw->port - 1], gw->gids, sizeof gw->gids);
-		event.device = &gw->dev->ib_dev;
-		event.element.port_num = gw->port;
-		event.event    = IB_EVENT_GID_CHANGE;
-		ib_dispatch_event(&event);
+		mlx4_ib_dispatch_event(gw->dev, gw->port, IB_EVENT_GID_CHANGE);
 	}
 
 	mlx4_free_cmd_mailbox(dev, mailbox);
@@ -1432,10 +1428,18 @@ static void mlx4_ib_remove(struct mlx4_dev *dev, void *ibdev_ptr)
 }
 
 static void mlx4_ib_event(struct mlx4_dev *dev, void *ibdev_ptr,
-			  enum mlx4_dev_event event, int port)
+			  enum mlx4_dev_event event, unsigned long param)
 {
 	struct ib_event ibev;
 	struct mlx4_ib_dev *ibdev = to_mdev((struct ib_device *) ibdev_ptr);
+	struct mlx4_eqe *eqe = NULL;
+	struct ib_event_work *ew;
+	int port = 0;
+
+	if (event == MLX4_DEV_EVENT_PORT_MGMT_CHANGE)
+		eqe = (struct mlx4_eqe *)param;
+	else
+		port = (u8)param;
 
 	if (port > ibdev->num_ports)
 		return;
@@ -1454,6 +1458,19 @@ static void mlx4_ib_event(struct mlx4_dev *dev, void *ibdev_ptr,
 		ibev.event = IB_EVENT_DEVICE_FATAL;
 		break;
 
+	case MLX4_DEV_EVENT_PORT_MGMT_CHANGE:
+		ew = kmalloc(sizeof *ew, GFP_ATOMIC);
+		if (!ew) {
+			pr_err("failed to allocate memory for events work\n");
+			break;
+		}
+
+		INIT_WORK(&ew->work, handle_port_mgmt_change_event);
+		memcpy(&ew->ib_eqe, eqe, sizeof *eqe);
+		ew->ib_dev = ibdev;
+		handle_port_mgmt_change_event(&ew->work);
+		return;
+
 	default:
 		return;
 	}
diff --git a/drivers/infiniband/hw/mlx4/mlx4_ib.h b/drivers/infiniband/hw/mlx4/mlx4_ib.h
index 42df4f7a6a5..c136bb618e2 100644
--- a/drivers/infiniband/hw/mlx4/mlx4_ib.h
+++ b/drivers/infiniband/hw/mlx4/mlx4_ib.h
@@ -44,6 +44,16 @@
 #include <linux/mlx4/device.h>
 #include <linux/mlx4/doorbell.h>
 
+#define MLX4_IB_DRV_NAME	"mlx4_ib"
+
+#ifdef pr_fmt
+#undef pr_fmt
+#endif
+#define pr_fmt(fmt)	"<" MLX4_IB_DRV_NAME "> %s: " fmt, __func__
+
+#define mlx4_ib_warn(ibdev, format, arg...) \
+	dev_warn((ibdev)->dma_device, MLX4_IB_DRV_NAME ": " format, ## arg)
+
 enum {
 	MLX4_IB_SQ_MIN_WQE_SHIFT = 6,
 	MLX4_IB_MAX_HEADROOM	 = 2048
@@ -215,6 +225,12 @@ struct mlx4_ib_dev {
 	int			eq_added;
 };
 
+struct ib_event_work {
+	struct work_struct	work;
+	struct mlx4_ib_dev	*ib_dev;
+	struct mlx4_eqe		ib_eqe;
+};
+
 static inline struct mlx4_ib_dev *to_mdev(struct ib_device *ibdev)
 {
 	return container_of(ibdev, struct mlx4_ib_dev, ib_dev);
@@ -372,4 +388,7 @@ static inline int mlx4_ib_ah_grh_present(struct mlx4_ib_ah *ah)
 int mlx4_ib_add_mc(struct mlx4_ib_dev *mdev, struct mlx4_ib_qp *mqp,
 		   union ib_gid *gid);
 
+void mlx4_ib_dispatch_event(struct mlx4_ib_dev *dev, u8 port_num,
+			    enum ib_event_type type);
+
 #endif /* MLX4_IB_H */
diff --git a/drivers/infiniband/hw/mlx4/qp.c b/drivers/infiniband/hw/mlx4/qp.c
index 6af19f6c2b1..a6d8ea060ea 100644
--- a/drivers/infiniband/hw/mlx4/qp.c
+++ b/drivers/infiniband/hw/mlx4/qp.c
@@ -1336,11 +1336,21 @@ int mlx4_ib_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr,
 	cur_state = attr_mask & IB_QP_CUR_STATE ? attr->cur_qp_state : qp->state;
 	new_state = attr_mask & IB_QP_STATE ? attr->qp_state : cur_state;
 
-	if (!ib_modify_qp_is_ok(cur_state, new_state, ibqp->qp_type, attr_mask))
+	if (!ib_modify_qp_is_ok(cur_state, new_state, ibqp->qp_type, attr_mask)) {
+		pr_debug("qpn 0x%x: invalid attribute mask specified "
+			 "for transition %d to %d. qp_type %d,"
+			 " attr_mask 0x%x\n",
+			 ibqp->qp_num, cur_state, new_state,
+			 ibqp->qp_type, attr_mask);
 		goto out;
+	}
 
 	if ((attr_mask & IB_QP_PORT) &&
 	    (attr->port_num == 0 || attr->port_num > dev->dev->caps.num_ports)) {
+		pr_debug("qpn 0x%x: invalid port number (%d) specified "
+			 "for transition %d to %d. qp_type %d\n",
+			 ibqp->qp_num, attr->port_num, cur_state,
+			 new_state, ibqp->qp_type);
 		goto out;
 	}
 
@@ -1351,17 +1361,30 @@ int mlx4_ib_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr,
 
 	if (attr_mask & IB_QP_PKEY_INDEX) {
 		int p = attr_mask & IB_QP_PORT ? attr->port_num : qp->port;
-		if (attr->pkey_index >= dev->dev->caps.pkey_table_len[p])
+		if (attr->pkey_index >= dev->dev->caps.pkey_table_len[p]) {
+			pr_debug("qpn 0x%x: invalid pkey index (%d) specified "
+				 "for transition %d to %d. qp_type %d\n",
+				 ibqp->qp_num, attr->pkey_index, cur_state,
+				 new_state, ibqp->qp_type);
 			goto out;
+		}
 	}
 
 	if (attr_mask & IB_QP_MAX_QP_RD_ATOMIC &&
 	    attr->max_rd_atomic > dev->dev->caps.max_qp_init_rdma) {
+		pr_debug("qpn 0x%x: max_rd_atomic (%d) too large. "
+			 "Transition %d to %d. qp_type %d\n",
+			 ibqp->qp_num, attr->max_rd_atomic, cur_state,
+			 new_state, ibqp->qp_type);
 		goto out;
 	}
 
 	if (attr_mask & IB_QP_MAX_DEST_RD_ATOMIC &&
 	    attr->max_dest_rd_atomic > dev->dev->caps.max_qp_dest_rdma) {
+		pr_debug("qpn 0x%x: max_dest_rd_atomic (%d) too large. "
+			 "Transition %d to %d. qp_type %d\n",
+			 ibqp->qp_num, attr->max_dest_rd_atomic, cur_state,
+			 new_state, ibqp->qp_type);
 		goto out;
 	}