IB/mlx4: Add support for IBoE

Add support for IBoE to mlx4_ib.  The bulk of the code is handling the
new address vector fields; mlx4 needs the MAC address of a remote node
to include it in a WQE (for datagrams) or in the QP context (for
connected QPs).  Address resolution is done by assuming all unicast
GIDs are either link-local IPv6 addresses.

Multicast group attach/detach needs to update the NIC's multicast
filters; but since attaching a QP to a multicast group can be done
before the QP is bound to a port, for IBoE we need to keep track of
all multicast groups that a QP is attached too before it transitions
from INIT to RTR (since it does not have a port in the INIT state).

Signed-off-by: Eli Cohen <eli@mellanox.co.il>

[ Many things cleaned up and otherwise monkeyed with; hope I didn't
  introduce too many bugs.  - Roland ]

Signed-off-by: Roland Dreier <rolandd@cisco.com>

1 files changed, 120 insertions, 33 deletions

diff --git a/drivers/infiniband/hw/mlx4/ah.c b/drivers/infiniband/hw/mlx4/ah.c
index 11a236f8d884..3bf3544c0aa0 100644
--- a/drivers/infiniband/hw/mlx4/ah.c
+++ b/drivers/infiniband/hw/mlx4/ah.c
@@ -30,66 +30,153 @@
  * SOFTWARE.
  */
 
+#include <rdma/ib_addr.h>
+
 #include <linux/slab.h>
+#include <linux/inet.h>
+#include <linux/string.h>
 
 #include "mlx4_ib.h"
 
-struct ib_ah *mlx4_ib_create_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr)
+int mlx4_ib_resolve_grh(struct mlx4_ib_dev *dev, const struct ib_ah_attr *ah_attr,
+			u8 *mac, int *is_mcast, u8 port)
 {
-	struct mlx4_dev *dev = to_mdev(pd->device)->dev;
-	struct mlx4_ib_ah *ah;
+	struct in6_addr in6;
 
-	ah = kmalloc(sizeof *ah, GFP_ATOMIC);
-	if (!ah)
-		return ERR_PTR(-ENOMEM);
+	*is_mcast = 0;
 
-	memset(&ah->av, 0, sizeof ah->av);
+	memcpy(&in6, ah_attr->grh.dgid.raw, sizeof in6);
+	if (rdma_link_local_addr(&in6))
+		rdma_get_ll_mac(&in6, mac);
+	else if (rdma_is_multicast_addr(&in6)) {
+		rdma_get_mcast_mac(&in6, mac);
+		*is_mcast = 1;
+	} else
+		return -EINVAL;
 
-	ah->av.port_pd = cpu_to_be32(to_mpd(pd)->pdn | (ah_attr->port_num << 24));
-	ah->av.g_slid  = ah_attr->src_path_bits;
-	ah->av.dlid    = cpu_to_be16(ah_attr->dlid);
-	if (ah_attr->static_rate) {
-		ah->av.stat_rate = ah_attr->static_rate + MLX4_STAT_RATE_OFFSET;
-		while (ah->av.stat_rate > IB_RATE_2_5_GBPS + MLX4_STAT_RATE_OFFSET &&
-		       !(1 << ah->av.stat_rate & dev->caps.stat_rate_support))
-			--ah->av.stat_rate;
-	}
-	ah->av.sl_tclass_flowlabel = cpu_to_be32(ah_attr->sl << 28);
+	return 0;
+}
+
+static struct ib_ah *create_ib_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr,
+				  struct mlx4_ib_ah *ah)
+{
+	struct mlx4_dev *dev = to_mdev(pd->device)->dev;
+
+	ah->av.ib.port_pd = cpu_to_be32(to_mpd(pd)->pdn | (ah_attr->port_num << 24));
+	ah->av.ib.g_slid  = ah_attr->src_path_bits;
 	if (ah_attr->ah_flags & IB_AH_GRH) {
-		ah->av.g_slid   |= 0x80;
-		ah->av.gid_index = ah_attr->grh.sgid_index;
-		ah->av.hop_limit = ah_attr->grh.hop_limit;
-		ah->av.sl_tclass_flowlabel |=
+		ah->av.ib.g_slid   |= 0x80;
+		ah->av.ib.gid_index = ah_attr->grh.sgid_index;
+		ah->av.ib.hop_limit = ah_attr->grh.hop_limit;
+		ah->av.ib.sl_tclass_flowlabel |=
 			cpu_to_be32((ah_attr->grh.traffic_class << 20) |
 				    ah_attr->grh.flow_label);
-		memcpy(ah->av.dgid, ah_attr->grh.dgid.raw, 16);
+		memcpy(ah->av.ib.dgid, ah_attr->grh.dgid.raw, 16);
+	}
+
+	ah->av.ib.dlid    = cpu_to_be16(ah_attr->dlid);
+	if (ah_attr->static_rate) {
+		ah->av.ib.stat_rate = ah_attr->static_rate + MLX4_STAT_RATE_OFFSET;
+		while (ah->av.ib.stat_rate > IB_RATE_2_5_GBPS + MLX4_STAT_RATE_OFFSET &&
+		       !(1 << ah->av.ib.stat_rate & dev->caps.stat_rate_support))
+			--ah->av.ib.stat_rate;
 	}
+	ah->av.ib.sl_tclass_flowlabel = cpu_to_be32(ah_attr->sl << 28);
 
 	return &ah->ibah;
 }
 
+static struct ib_ah *create_iboe_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr,
+				    struct mlx4_ib_ah *ah)
+{
+	struct mlx4_ib_dev *ibdev = to_mdev(pd->device);
+	struct mlx4_dev *dev = ibdev->dev;
+	u8 mac[6];
+	int err;
+	int is_mcast;
+
+	err = mlx4_ib_resolve_grh(ibdev, ah_attr, mac, &is_mcast, ah_attr->port_num);
+	if (err)
+		return ERR_PTR(err);
+
+	memcpy(ah->av.eth.mac, mac, 6);
+	ah->av.eth.port_pd = cpu_to_be32(to_mpd(pd)->pdn | (ah_attr->port_num << 24));
+	ah->av.eth.gid_index = ah_attr->grh.sgid_index;
+	if (ah_attr->static_rate) {
+		ah->av.eth.stat_rate = ah_attr->static_rate + MLX4_STAT_RATE_OFFSET;
+		while (ah->av.eth.stat_rate > IB_RATE_2_5_GBPS + MLX4_STAT_RATE_OFFSET &&
+		       !(1 << ah->av.eth.stat_rate & dev->caps.stat_rate_support))
+			--ah->av.eth.stat_rate;
+	}
+
+	/*
+	 * HW requires multicast LID so we just choose one.
+	 */
+	if (is_mcast)
+		ah->av.ib.dlid = cpu_to_be16(0xc000);
+
+	memcpy(ah->av.eth.dgid, ah_attr->grh.dgid.raw, 16);
+	ah->av.eth.sl_tclass_flowlabel = cpu_to_be32(ah_attr->sl << 28);
+
+	return &ah->ibah;
+}
+
+struct ib_ah *mlx4_ib_create_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr)
+{
+	struct mlx4_ib_ah *ah;
+	struct ib_ah *ret;
+
+	ah = kzalloc(sizeof *ah, GFP_ATOMIC);
+	if (!ah)
+		return ERR_PTR(-ENOMEM);
+
+	if (rdma_port_get_link_layer(pd->device, ah_attr->port_num) == IB_LINK_LAYER_ETHERNET) {
+		if (!(ah_attr->ah_flags & IB_AH_GRH)) {
+			ret = ERR_PTR(-EINVAL);
+		} else {
+			/*
+			 * TBD: need to handle the case when we get
+			 * called in an atomic context and there we
+			 * might sleep.  We don't expect this
+			 * currently since we're working with link
+			 * local addresses which we can translate
+			 * without going to sleep.
+			 */
+			ret = create_iboe_ah(pd, ah_attr, ah);
+		}
+
+		if (IS_ERR(ret))
+			kfree(ah);
+
+		return ret;
+	} else
+		return create_ib_ah(pd, ah_attr, ah); /* never fails */
+}
+
 int mlx4_ib_query_ah(struct ib_ah *ibah, struct ib_ah_attr *ah_attr)
 {
 	struct mlx4_ib_ah *ah = to_mah(ibah);
+	enum rdma_link_layer ll;
 
 	memset(ah_attr, 0, sizeof *ah_attr);
-	ah_attr->dlid	       = be16_to_cpu(ah->av.dlid);
-	ah_attr->sl	       = be32_to_cpu(ah->av.sl_tclass_flowlabel) >> 28;
-	ah_attr->port_num      = be32_to_cpu(ah->av.port_pd) >> 24;
-	if (ah->av.stat_rate)
-		ah_attr->static_rate = ah->av.stat_rate - MLX4_STAT_RATE_OFFSET;
-	ah_attr->src_path_bits = ah->av.g_slid & 0x7F;
+	ah_attr->sl = be32_to_cpu(ah->av.ib.sl_tclass_flowlabel) >> 28;
+	ah_attr->port_num = be32_to_cpu(ah->av.ib.port_pd) >> 24;
+	ll = rdma_port_get_link_layer(ibah->device, ah_attr->port_num);
+	ah_attr->dlid = ll == IB_LINK_LAYER_INFINIBAND ? be16_to_cpu(ah->av.ib.dlid) : 0;
+	if (ah->av.ib.stat_rate)
+		ah_attr->static_rate = ah->av.ib.stat_rate - MLX4_STAT_RATE_OFFSET;
+	ah_attr->src_path_bits = ah->av.ib.g_slid & 0x7F;
 
 	if (mlx4_ib_ah_grh_present(ah)) {
 		ah_attr->ah_flags = IB_AH_GRH;
 
 		ah_attr->grh.traffic_class =
-			be32_to_cpu(ah->av.sl_tclass_flowlabel) >> 20;
+			be32_to_cpu(ah->av.ib.sl_tclass_flowlabel) >> 20;
 		ah_attr->grh.flow_label =
-			be32_to_cpu(ah->av.sl_tclass_flowlabel) & 0xfffff;
-		ah_attr->grh.hop_limit  = ah->av.hop_limit;
-		ah_attr->grh.sgid_index = ah->av.gid_index;
-		memcpy(ah_attr->grh.dgid.raw, ah->av.dgid, 16);
+			be32_to_cpu(ah->av.ib.sl_tclass_flowlabel) & 0xfffff;
+		ah_attr->grh.hop_limit  = ah->av.ib.hop_limit;
+		ah_attr->grh.sgid_index = ah->av.ib.gid_index;
+		memcpy(ah_attr->grh.dgid.raw, ah->av.ib.dgid, 16);
 	}
 
 	return 0;