aboutsummaryrefslogtreecommitdiff
path: root/drivers/infiniband/hw/mlx4/main.c
diff options
context:
space:
mode:
Diffstat (limited to 'drivers/infiniband/hw/mlx4/main.c')
-rw-r--r--drivers/infiniband/hw/mlx4/main.c174
1 files changed, 159 insertions, 15 deletions
diff --git a/drivers/infiniband/hw/mlx4/main.c b/drivers/infiniband/hw/mlx4/main.c
index 97d6878f9938..914bc98e753f 100644
--- a/drivers/infiniband/hw/mlx4/main.c
+++ b/drivers/infiniband/hw/mlx4/main.c
@@ -154,9 +154,9 @@ static struct net_device *mlx4_ib_get_netdev(struct ib_device *device, u8 port_n
return dev;
}
-static int mlx4_ib_update_gids(struct gid_entry *gids,
- struct mlx4_ib_dev *ibdev,
- u8 port_num)
+static int mlx4_ib_update_gids_v1(struct gid_entry *gids,
+ struct mlx4_ib_dev *ibdev,
+ u8 port_num)
{
struct mlx4_cmd_mailbox *mailbox;
int err;
@@ -187,6 +187,63 @@ static int mlx4_ib_update_gids(struct gid_entry *gids,
return err;
}
+static int mlx4_ib_update_gids_v1_v2(struct gid_entry *gids,
+ struct mlx4_ib_dev *ibdev,
+ u8 port_num)
+{
+ struct mlx4_cmd_mailbox *mailbox;
+ int err;
+ struct mlx4_dev *dev = ibdev->dev;
+ int i;
+ struct {
+ union ib_gid gid;
+ __be32 rsrvd1[2];
+ __be16 rsrvd2;
+ u8 type;
+ u8 version;
+ __be32 rsrvd3;
+ } *gid_tbl;
+
+ mailbox = mlx4_alloc_cmd_mailbox(dev);
+ if (IS_ERR(mailbox))
+ return -ENOMEM;
+
+ gid_tbl = mailbox->buf;
+ for (i = 0; i < MLX4_MAX_PORT_GIDS; ++i) {
+ memcpy(&gid_tbl[i].gid, &gids[i].gid, sizeof(union ib_gid));
+ if (gids[i].gid_type == IB_GID_TYPE_ROCE_UDP_ENCAP) {
+ gid_tbl[i].version = 2;
+ if (!ipv6_addr_v4mapped((struct in6_addr *)&gids[i].gid))
+ gid_tbl[i].type = 1;
+ else
+ memset(&gid_tbl[i].gid, 0, 12);
+ }
+ }
+
+ err = mlx4_cmd(dev, mailbox->dma,
+ MLX4_SET_PORT_ROCE_ADDR << 8 | port_num,
+ 1, MLX4_CMD_SET_PORT, MLX4_CMD_TIME_CLASS_B,
+ MLX4_CMD_WRAPPED);
+ if (mlx4_is_bonded(dev))
+ err += mlx4_cmd(dev, mailbox->dma,
+ MLX4_SET_PORT_ROCE_ADDR << 8 | 2,
+ 1, MLX4_CMD_SET_PORT, MLX4_CMD_TIME_CLASS_B,
+ MLX4_CMD_WRAPPED);
+
+ mlx4_free_cmd_mailbox(dev, mailbox);
+ return err;
+}
+
+static int mlx4_ib_update_gids(struct gid_entry *gids,
+ struct mlx4_ib_dev *ibdev,
+ u8 port_num)
+{
+ if (ibdev->dev->caps.flags2 & MLX4_DEV_CAP_FLAG2_ROCE_V1_V2)
+ return mlx4_ib_update_gids_v1_v2(gids, ibdev, port_num);
+
+ return mlx4_ib_update_gids_v1(gids, ibdev, port_num);
+}
+
static int mlx4_ib_add_gid(struct ib_device *device,
u8 port_num,
unsigned int index,
@@ -215,7 +272,8 @@ static int mlx4_ib_add_gid(struct ib_device *device,
port_gid_table = &iboe->gids[port_num - 1];
spin_lock_bh(&iboe->lock);
for (i = 0; i < MLX4_MAX_PORT_GIDS; ++i) {
- if (!memcmp(&port_gid_table->gids[i].gid, gid, sizeof(*gid))) {
+ if (!memcmp(&port_gid_table->gids[i].gid, gid, sizeof(*gid)) &&
+ (port_gid_table->gids[i].gid_type == attr->gid_type)) {
found = i;
break;
}
@@ -233,6 +291,7 @@ static int mlx4_ib_add_gid(struct ib_device *device,
} else {
*context = port_gid_table->gids[free].ctx;
memcpy(&port_gid_table->gids[free].gid, gid, sizeof(*gid));
+ port_gid_table->gids[free].gid_type = attr->gid_type;
port_gid_table->gids[free].ctx->real_index = free;
port_gid_table->gids[free].ctx->refcount = 1;
hw_update = 1;
@@ -248,8 +307,10 @@ static int mlx4_ib_add_gid(struct ib_device *device,
if (!gids) {
ret = -ENOMEM;
} else {
- for (i = 0; i < MLX4_MAX_PORT_GIDS; i++)
+ for (i = 0; i < MLX4_MAX_PORT_GIDS; i++) {
memcpy(&gids[i].gid, &port_gid_table->gids[i].gid, sizeof(union ib_gid));
+ gids[i].gid_type = port_gid_table->gids[i].gid_type;
+ }
}
}
spin_unlock_bh(&iboe->lock);
@@ -325,6 +386,7 @@ int mlx4_ib_gid_index_to_real_index(struct mlx4_ib_dev *ibdev,
int i;
int ret;
unsigned long flags;
+ struct ib_gid_attr attr;
if (port_num > MLX4_MAX_PORTS)
return -EINVAL;
@@ -335,10 +397,13 @@ int mlx4_ib_gid_index_to_real_index(struct mlx4_ib_dev *ibdev,
if (!rdma_cap_roce_gid_table(&ibdev->ib_dev, port_num))
return index;
- ret = ib_get_cached_gid(&ibdev->ib_dev, port_num, index, &gid, NULL);
+ ret = ib_get_cached_gid(&ibdev->ib_dev, port_num, index, &gid, &attr);
if (ret)
return ret;
+ if (attr.ndev)
+ dev_put(attr.ndev);
+
if (!memcmp(&gid, &zgid, sizeof(gid)))
return -EINVAL;
@@ -346,7 +411,8 @@ int mlx4_ib_gid_index_to_real_index(struct mlx4_ib_dev *ibdev,
port_gid_table = &iboe->gids[port_num - 1];
for (i = 0; i < MLX4_MAX_PORT_GIDS; ++i)
- if (!memcmp(&port_gid_table->gids[i].gid, &gid, sizeof(gid))) {
+ if (!memcmp(&port_gid_table->gids[i].gid, &gid, sizeof(gid)) &&
+ attr.gid_type == port_gid_table->gids[i].gid_type) {
ctx = port_gid_table->gids[i].ctx;
break;
}
@@ -1577,6 +1643,56 @@ static int mlx4_ib_tunnel_steer_add(struct ib_qp *qp, struct ib_flow_attr *flow_
return err;
}
+static int mlx4_ib_add_dont_trap_rule(struct mlx4_dev *dev,
+ struct ib_flow_attr *flow_attr,
+ enum mlx4_net_trans_promisc_mode *type)
+{
+ int err = 0;
+
+ if (!(dev->caps.flags2 & MLX4_DEV_CAP_FLAG2_DMFS_UC_MC_SNIFFER) ||
+ (dev->caps.dmfs_high_steer_mode == MLX4_STEERING_DMFS_A0_STATIC) ||
+ (flow_attr->num_of_specs > 1) || (flow_attr->priority != 0)) {
+ return -EOPNOTSUPP;
+ }
+
+ if (flow_attr->num_of_specs == 0) {
+ type[0] = MLX4_FS_MC_SNIFFER;
+ type[1] = MLX4_FS_UC_SNIFFER;
+ } else {
+ union ib_flow_spec *ib_spec;
+
+ ib_spec = (union ib_flow_spec *)(flow_attr + 1);
+ if (ib_spec->type != IB_FLOW_SPEC_ETH)
+ return -EINVAL;
+
+ /* if all is zero than MC and UC */
+ if (is_zero_ether_addr(ib_spec->eth.mask.dst_mac)) {
+ type[0] = MLX4_FS_MC_SNIFFER;
+ type[1] = MLX4_FS_UC_SNIFFER;
+ } else {
+ u8 mac[ETH_ALEN] = {ib_spec->eth.mask.dst_mac[0] ^ 0x01,
+ ib_spec->eth.mask.dst_mac[1],
+ ib_spec->eth.mask.dst_mac[2],
+ ib_spec->eth.mask.dst_mac[3],
+ ib_spec->eth.mask.dst_mac[4],
+ ib_spec->eth.mask.dst_mac[5]};
+
+ /* Above xor was only on MC bit, non empty mask is valid
+ * only if this bit is set and rest are zero.
+ */
+ if (!is_zero_ether_addr(&mac[0]))
+ return -EINVAL;
+
+ if (is_multicast_ether_addr(ib_spec->eth.val.dst_mac))
+ type[0] = MLX4_FS_MC_SNIFFER;
+ else
+ type[0] = MLX4_FS_UC_SNIFFER;
+ }
+ }
+
+ return err;
+}
+
static struct ib_flow *mlx4_ib_create_flow(struct ib_qp *qp,
struct ib_flow_attr *flow_attr,
int domain)
@@ -1587,6 +1703,10 @@ static struct ib_flow *mlx4_ib_create_flow(struct ib_qp *qp,
struct mlx4_dev *dev = (to_mdev(qp->device))->dev;
int is_bonded = mlx4_is_bonded(dev);
+ if ((flow_attr->flags & IB_FLOW_ATTR_FLAGS_DONT_TRAP) &&
+ (flow_attr->type != IB_FLOW_ATTR_NORMAL))
+ return ERR_PTR(-EOPNOTSUPP);
+
memset(type, 0, sizeof(type));
mflow = kzalloc(sizeof(*mflow), GFP_KERNEL);
@@ -1597,7 +1717,19 @@ static struct ib_flow *mlx4_ib_create_flow(struct ib_qp *qp,
switch (flow_attr->type) {
case IB_FLOW_ATTR_NORMAL:
- type[0] = MLX4_FS_REGULAR;
+ /* If dont trap flag (continue match) is set, under specific
+ * condition traffic be replicated to given qp,
+ * without stealing it
+ */
+ if (unlikely(flow_attr->flags & IB_FLOW_ATTR_FLAGS_DONT_TRAP)) {
+ err = mlx4_ib_add_dont_trap_rule(dev,
+ flow_attr,
+ type);
+ if (err)
+ goto err_free;
+ } else {
+ type[0] = MLX4_FS_REGULAR;
+ }
break;
case IB_FLOW_ATTR_ALL_DEFAULT:
@@ -1609,8 +1741,8 @@ static struct ib_flow *mlx4_ib_create_flow(struct ib_qp *qp,
break;
case IB_FLOW_ATTR_SNIFFER:
- type[0] = MLX4_FS_UC_SNIFFER;
- type[1] = MLX4_FS_MC_SNIFFER;
+ type[0] = MLX4_FS_MIRROR_RX_PORT;
+ type[1] = MLX4_FS_MIRROR_SX_PORT;
break;
default:
@@ -2119,6 +2251,7 @@ static int mlx4_port_immutable(struct ib_device *ibdev, u8 port_num,
struct ib_port_immutable *immutable)
{
struct ib_port_attr attr;
+ struct mlx4_ib_dev *mdev = to_mdev(ibdev);
int err;
err = mlx4_ib_query_port(ibdev, port_num, &attr);
@@ -2128,10 +2261,15 @@ static int mlx4_port_immutable(struct ib_device *ibdev, u8 port_num,
immutable->pkey_tbl_len = attr.pkey_tbl_len;
immutable->gid_tbl_len = attr.gid_tbl_len;
- if (mlx4_ib_port_link_layer(ibdev, port_num) == IB_LINK_LAYER_INFINIBAND)
+ if (mlx4_ib_port_link_layer(ibdev, port_num) == IB_LINK_LAYER_INFINIBAND) {
immutable->core_cap_flags = RDMA_CORE_PORT_IBA_IB;
- else
- immutable->core_cap_flags = RDMA_CORE_PORT_IBA_ROCE;
+ } else {
+ if (mdev->dev->caps.flags & MLX4_DEV_CAP_FLAG_IBOE)
+ immutable->core_cap_flags = RDMA_CORE_PORT_IBA_ROCE;
+ if (mdev->dev->caps.flags2 & MLX4_DEV_CAP_FLAG2_ROCE_V1_V2)
+ immutable->core_cap_flags = RDMA_CORE_PORT_IBA_ROCE |
+ RDMA_CORE_PORT_IBA_ROCE_UDP_ENCAP;
+ }
immutable->max_mad_size = IB_MGMT_MAD_SIZE;
@@ -2283,7 +2421,6 @@ static void *mlx4_ib_add(struct mlx4_dev *dev)
if (dev->caps.flags & MLX4_DEV_CAP_FLAG_MEM_WINDOW ||
dev->caps.bmme_flags & MLX4_BMME_FLAG_TYPE_2_WIN) {
ibdev->ib_dev.alloc_mw = mlx4_ib_alloc_mw;
- ibdev->ib_dev.bind_mw = mlx4_ib_bind_mw;
ibdev->ib_dev.dealloc_mw = mlx4_ib_dealloc_mw;
ibdev->ib_dev.uverbs_cmd_mask |=
@@ -2423,7 +2560,8 @@ static void *mlx4_ib_add(struct mlx4_dev *dev)
if (mlx4_ib_init_sriov(ibdev))
goto err_mad;
- if (dev->caps.flags & MLX4_DEV_CAP_FLAG_IBOE) {
+ if (dev->caps.flags & MLX4_DEV_CAP_FLAG_IBOE ||
+ dev->caps.flags2 & MLX4_DEV_CAP_FLAG2_ROCE_V1_V2) {
if (!iboe->nb.notifier_call) {
iboe->nb.notifier_call = mlx4_ib_netdev_event;
err = register_netdevice_notifier(&iboe->nb);
@@ -2432,6 +2570,12 @@ static void *mlx4_ib_add(struct mlx4_dev *dev)
goto err_notif;
}
}
+ if (dev->caps.flags2 & MLX4_DEV_CAP_FLAG2_ROCE_V1_V2) {
+ err = mlx4_config_roce_v2_port(dev, ROCE_V2_UDP_DPORT);
+ if (err) {
+ goto err_notif;
+ }
+ }
}
for (j = 0; j < ARRAY_SIZE(mlx4_class_attributes); ++j) {
generated by cgit (git 2.34.1) at 2025-06-01 00:38:39 +0000